The hype of the Internet of things (IoT) is marching forward supported by many analysts and IT vendors. The aspiration is that IoT nourished by the flow of wearable devices emerging first in the fitness and wellness areas coupled with mobile applications will boost patients’ and, in general, citizens’ engagement in their own care.

A running girl

The reality, however, is different. Despite the optimistic prognoses that nearly 100 million wearable monitoring devices will ship over the next five years spurred by consumer interest and increasing awareness of how mHealth can enhance healthcare-related activities, the actual usability of such devices was relatively small [1].

The nationwide US study designed by Technology Advice Research in September 18-19, 2014 interrogated over 900 adults in regard to their general fitness tracking habits. Among them, 419 adults were surveyed on their specific reasons for not using tracking devices or apps. The results showed that nearly 74.9% were not using mHealth fitness apps or devices [2].

Should the discovery be surprising?

Until recently, consumers were presented with multiple mostly fitness-related devices measuring vital signs that create pools of personal health records (PHR) connected to cloud storages such as the MS Health Vault or the coming Apple Health Kit. But the crucial question remained: if I collect my health data, what do I do next?

Doctor watching PHR

What was missing is the intelligence that would help consumers not only to collect, but also to interpret the correlations between various health parameters. In short, consumers were looking for personal assistants that would prompt them what to do to manage their health better.

It seems that large IT vendors are gradually realizing such concerns. Both Apple SIRI, Google Now and Microsoft Cortana products are promising consumers personal virtual assistance (PVA) services for their mobile devices. PVA will use a natural language user interface to answer questions, make recommendations and perform actions by delegating requests to a set of Web services. The responses will be sent automatically to consumer mobile devices. So far, the “services menu” is quite general and rather rudimentary. One, for example, can give a voice command to his/her PVA to book a restaurant, a flight ticket or receive a voice reminder that a flight was delayed and there is no need to hurry.

But the development is going on. The Microsoft product Cortana is said to add new features almost every two weeks. So far available for Windows Phone 8.1, the technology is expected to be added to Windows 10 and XBox One sometimes in 2015. It seems that Cortana is looking far beyond a simple “voice search” with the ambition to step into the area of artificial intelligence. According to Peter Lee, the Head of Microsoft Research, “the aim is to produce systems that can see, hear and understand, by focusing on correlations in large amounts of data” [3].

For example, by analyzing data coming from multiple applications such as email, calendar and favorite locations stored on your devices or personal cloud, the intelligent service may predict when, e.g., you should get out of your house to reach the bus station on time, provided that there is a traffic jam; or by tracking how frequently you visit some of the popular musical sites, remind you that your favorite jazz band is coming the next month and it is time to book tickets.

What if such personal virtual assistant would be adapted to mHealth applications  empowered by intelligent services through a cloud platform e.g. Microsoft Health Vault or the likes? In this case one can easily envisage the following scenarios:

  • Medication Management

To remind you to take a medication based on your calendar is not a big thing. One can do it with SMS already now. But how about tracking the efficiency of medications?

Medication reminder

What if your virtual assistant after comparing your blood pressure with the dosage of your beta blockers or ACE inhibitors will let you know that there is almost no positive progress? On top of that, for example, the levels of aspartate aminotransferase (AST, also known as serum glutamic oxaloacetic transaminase, SGOT) and alanine aminotransferase (ALT or SGPT) are above the norm indicating potential liver damage. It could be a good start to discuss with your physician a possible medication change. Some hospitals are already installing such systems for their patients.

For instance, doctors at Vanderbilt University Medical Center (Vanderbilt University Medical Center ) in Nashville and St. Jude’s Medical Center in Memphis are getting pop-up notifications within individual patients’ electronic medical records. The alerts tell them, in particular, when a drug might not work for a patient with certain genetic traits [4].

  • Early Alerts

Many people are able now to track and store their lab results. However one single parameter, e.g., Low Density Cholesterol (LDC) will not show you the full picture of your health and may even create a false alarm. Often a combination of multiple parameters such as HDC, LDC, Triglycerides or C-reactive protein (CRP) is needed to get the real picture of your, in this case, heart condition. After comparing multiple correlated parameters, the alert could be sent if the lab results are bridging the norm.

  • Patient Monitoring

One of my friends, an experienced cardiologist, once told me: it is not enough to read the EMR, I need to “hear” a patient. Our voice pitch and tempo could tell much more about our conditions, pains or moods. Besides, in putting comments in a PHR manually could be a cumbersome procedure. By using a natural language process (NLP), a person could simply dictate what he or she is feeling at the given moment. This voice message can “tell” a physician a lot more about his patient’s health.

  • Appointment Booking

Those services are already available for multiple devices by various companies operating in mHealth area. For example, Spanish  MedCitas, a centralized appointment system developed by NetBoss eHealth for patients and physicians. By analyzing multiple calendars and registrations it helps to book the nearest appointment with physicians based on their profiles and specific expertise. Using a bit of imagination one can foresee that by analyzing personal data coming from PHR through multiple wearable devices a person will receive an alert that he or she needs to visit e.g. a gastroenterologist or cardiologist located in the nearest town.

If these services are easily available to patients and consumers, they will drastically stimulate the interest in wearable devices and related mobile applications. So far, it may be considered a vison. But it is more than a dream. The almost unlimited pool of personal health data are on the way to be transformed into intelligent services helping consumers and physicians to manage Healthcare faster and more efficiently.


  1. Report:100m-mhealth-monitoring-wearables-will-ship-2020
  2. Study: wearable technology for preventative-healthcare
  3. Beyond Cortana: What artificial intelligence means for Microsoft?



Continuous medical education for patients : three basic steps to succeed

While watching the avalanche of articles describing the benefits of “patients’ empowerment” boosted by personal health records, mobile applications, wearable devices, etc., I am thinking about my local GP. Like many other “Hausärzte” (home doctors) in Germany he sits in his small practice with two nurses, both assuming also the role of a receptionist. One probably could find similar practices in many places across Europe. Usually, there are 5-6 people sitting in Dr. Fischer’s waiting room. The queue is forming not because the patients are coming at random. All of them know their appointment time, but often the 12-15 minutes allotted per patient are not enough. Dr. Fischer is a diligent doctor: he nearly always measures blood pressure; if necessary, he takes ECG, performs ultrasound tests, makes injections – and all by himself.

Now imagine that at least half of our 12.000 village inhabitants will get mobile tools like PHR to store their medical data. What will they discover there?

It would be a truism to state that the medical terminology is remarkably weird and mainly serves the purpose to hide facts from a patient. This is reflected in the very beginning of clinical procedures: providing a diagnosis. The latter is denoted as dx or Dx, which is already confusing enough for uninitiated patients. Diagnoses are usually written in specific jargon (for example, “mono“ should be understood as mononucleosis) or using acronyms and initialisms such as, e.g., GCA (giant cell arteritis) which is to denote an inflammation of blood vessels, SLE (systemic lupus erythematosus), an autoimmune connective tissue damage, PE (pulmonary embolism), DVT (deep vein thrombosis) or VTE (venous thromboembolism), etc. It is interesting that medical acronyms related to the same disease are different in different languages (e.g., ТЭЛА stands in Russian for “тромбоэмболия легочной артерии“) which is basically the same as PE. This language-specific difference in acronyms creates additional difficulties. Besides, medical acronyms are notoriously non-unique (in mathematical sense); thus, what can one, for example, understand when meeting “BE“ – bacterial encephalitis, bacterial endocarditis, bacterial endarteritis or something else, depending on the concrete medical context. If a patient sees the term “hypercoagulability” in her/his diagnosis, what will be the reaction? In this case I can imagine my doctor Fischer receiving incessant calls and requests for appointments from patients concerned about their lab results.

A study  conducted by the University of Michigan screening 1,800 adults ages 40-70 discovered that “only slightly more than half of the patients, on average, were able to decipher electronic lab test results on their own”. As a result, contrary to the general belief that online health services (PHR being one of them) will substantially reduce costs by eliminating unnecessary visits, the number of calls and visits to physicians are growing [1]. The observation was proved by the recent Jama study. It was discovered that after using online clinical services there was a significant increase in the per member rates of office visits (0.7 per member per year; 95% CI – confidence interval) and telephone encounters (0.3 per member per year; 95% CI). There was also a significant increase in per 1000 member rates of after-hours clinic visits (18.7 per 1000 members per year; 95% CI) [2]. Information Technologies (IT) seem to be a natural remedy to increase patient’s health knowledge basics. During the recent years one sees good and bad examples of countless mobile and Web applications coming out. Unfortunately, it is still not easy to find relevant apps addressing the needs of consumers in a clear way.

There are at least 3 basic requirements consumers are expecting from developers and professional medical groups:  

Give what we need, not what you can give

The best way to understand what patients need is to follow medical procedures most of us frequently have to go through. For example, a complete blood count test (CBC) is one of the first checks your doctor usually requires you to do. To many people the ranges of their red blood cell proportion or hematocrit (hct) is a puzzle. It is not easy to relate these numbers to the actual risks their deviation from the “norm” can bring to individual. Of course there are some very good online resources like, e.g., National Heart, Lung and Blood Institute [3], that explains the correlations between the measured entities and health symptoms, but a corresponding and convenient consumer application is hard to find. BloodTest2 The top ones that came from the search (blood test, CBC test) were Nokia Blood test, a complete blood count test, and iPhone Blood tracker that records your laboratory values presenting them in a chronological sequence. Unfortunately, even the graphical representation of, say, one’s sodium concentration (needed for blood pressure control) or bilirubin level variations hardly explains how relevant it is for individual health conditions. As for the iTune record Blood Test Pro that contains over 140 laboratory values, it is clearly stated that “only a physician can judge the individual significance of laboratory values for the user’s physical health”. Patients on the other hand are offered numerous applications that explain the differences between various A, B, O blood groups which could be useful for donors, but are of a more academic interest for the majority of people keen to know what it means when, e.g., their homocysteine ratio is above the level of 15 micromole/l.

Present information in a clear way.

According to the NHC guidelines US National Health Council, patients “should receive complete and easily understood information about their condition and treatment options” [4]. This recommendation must be a necessity for consumer medical applications. There are some excellent examples of applications such as, e.g., NHS Choices which I often use to get updated information on medications, verify some symptoms or just to find quick answers to health-related questions that concern me at the moment.MedicalEorrorNHS The information I receive is mostly precise, crisp and clear. Unfortunately, it is not often the case with many other mobile health applications that are abundantly landing on the consumer market. Surfing on MS Store, I recently came across a MedWhat application that offered consumers an opportunity to ask medicine-related questions either by voice or typing it directly into the app. That looked promising, and I immediately typed: “what is the difference between Ramipril and Valsartan”, the two medications that are often prescribed to patients suffering from hypertension. Instead of explaining the difference between Ramipril, a popular ACE inhibitor, and Valsartan, an angiotensin II receptor blocker (ARB), I have received the following: “Ramipril or Valsartan significantly preserved the peritubular capillaries as well as renal function (p. <0.01). Tubulointerstitial hypoxia and tubular TGF-beta expression were noted well before the development of tubulointerstitial damage”. One can only imagine a perplex expression on a face of a lay person who is checking on medications to reduce his/her blood pressure.

Provide health analytics vs. data storage

The majority of applications like Personal Health Records (PHR) are tracking and storing the abundance of medical data (vital signs, laboratory data, medications or doctors’ visits) with no correlations between each other. I can track my blood and sugar levels every day and even each hour, but it is not easy to understand how the new medication I am taking can be compared to the one I had before in terms of effectiveness. The same applies to many samples of consumer medical devices. The typical example of disconnected applications is blood pressure monitors. The majority of them contain only 3 fields: “systolic”, ‘diastolic” and “pulls”. At best, one can put manual comments. You can see what is going on, but with no idea why. For example, if a person having elevated blood pressure switches from Ramipril, an angiotensin-converting enzyme (ACE) inhibitor, to Valsartan? A solution would be a simple backend analysis application, e.g., for a period of three months providing valuable information on blood pressure and medication interaction, preparing a person for a thoughtful discussion with a physician on the impact of the prescribed medication. Unfortunately, such multidimensional analytical consumer enabling applications are hard or impossible to find. So what could be a temporary solution for now? Just put your data in a familiar Excel application and it will make you colorful graphics showing how multiple parameters, e.g., your vital signs, intensity of physical exercises, medications or even weather can correlate with each other and impact your condition.

Call to action

The conclusion one can draw from multiple studies: governments, physicians, developers, etc. can spend a lot of money on creating sophisticated applications and PHRs for patients to access their data, but as long as the majority of people do not understand what those data mean, the effect of such efforts is limited. Educating patients in regard to medical basics is critical for establishing a meaningful dialogue and trustful collaboration between patients and physicians to protect the latter from the burden of unnecessary and time consuming requests from worried patients and for the overall improvement of the quality of care. A continuous medical education for citizens and patients may look like a chain of systematic online medical courses provided by local communities with the support of medical experts (e.g., retired doctors). Such courses will describe, e.g., the functioning of the body organs and the correlations between the main body subsystems: nervous system, circulatory system, respiratory system, digestive system, excretory system, metabolic system in general, etc., together with the pertinent groups of medications, also helping people to read their laboratory tests. Important is that the information is presented in a very clear way, understandable for non-medical community. Health-related consumer applications could be a great help to enrich a consumer’s knowledge in regard to her/his personal health, but these applications should be transformed into a flamboyant multidimensional form emphasizing salient features – a personal analytical health instrument vs. a pack of boring online data copied or transferred from the traditional medical documentation. References:


Have you ever doubt your doctor’s decision? If so, you are not the only one. Even doctors nowadays doubt the efficiency of healthcare system advising patients not to trust physicians “any more than you trust your stockbroker (if you are foolish enough to have one) [1].
This is a warning sign. If there is no trust between patients and physicians the positive outcomes of care are highly questionable.
The majority of patients do want to trust their physicians – actually doctors are their last resort for relief and comfort. I have also little doubt that physicians predominantly want to do their work well. That is to cure people and make a difference in their lives. That is what physicians were taught to do during many years of training. Unfortunately, care and cure are not the only things doctors have to consider under the requirements of the modern healthcare system. There are hundreds of KPIs they have to report on: the patient’s waiting list, time spent on each patient, average days of hospitalization, readmission, complications and – not the last one – costs.

Patients have only one KPI to care: that is to feel better. But this KPI encompasses all that physicians ultimately want to achieve. If this is the case, patients are the best allies to physicians to improve the efficiency of healthcare as a system. That is why patients’ complaints have to be heard and taken seriously.

What concerns us patients?

When patients are coming to see their physician they crave empathy. They expect doctors to listen to them, no matter how confusing their story might be. Instead quite often patients are treated like impersonal objects who are disturbing doctors with their dilettantish questions. Maybe this is the reason why many doctors are complaining on their patient’s behavior accusing them of being “difficult, nasty, obnoxious or disruptive”, placing “unrealistic responsibility on their doctors”. [2].

I vividly remember a patient who has visited a doctor with an eczema on his hand. He tried to show the burning spot and explain when exactly it has appeared. The doctor, possibly fighting with her computer, was barely looking at the damaged skin absorbed with the new software that was guiding her through various dermatological symptoms. After the long series of clicks the dermatologist received a suggested medication which was immediately prescribed. A man walked out with a sigh. Later, I have learned that the prescribed medication did not help. On the contrary: the eczema was spreading. A patient simply had to try at random several balms from the nearest drug store. He still did not know what the reason of his eczema was, but apparently one of the ointments he had chosen successfully worked.

2.Formalistic approach

No doubt health requires regulations and approved medical practices. Unfortunately those practices do not work for everybody. Modern healthcare is not ready for exceptions. The vivid example is our approach to medications. For centuries the traditional healthcare system was addressing a “typical” patient with “typical” symptoms and a “typical” reaction to medications. This approach appears to be methodologically wrong since it is based on an implicit assumption of the Gauss distribution underlying medical statistics.

The statistical results, e.g. produced in clinical trials, in fact do not answer the crucial question: what is the best strategy for a given patient. The patients’ reaction is simple: 50% of the patients are not adhering to prescribed medications [3]. A typical example: after taking Ramipril, a popular ACE inhibitor, for two weeks a patient complained on unpleasant heart pains and sleeping disorders. The physician’s response was classical: “It is not possible. These side effects are not common for Ramipril.” Only due to the patient’s persistency, Ramipril was substituted by Valsartan, an angiotensin II receptor blocker (ARB). Pains were gone. This was possibly an atypical case, and the doctors tend to disregard such cases offhand.


Misdiagnosis is one of the major threatening issues in Healthcare. According to Kaiser Health News at least 10 to 20 percent of cases are misdiagnosed. One report found that 28 percent of 583 diagnostic mistakes were life threatening or had resulted in death or permanent disability. Another study estimated that fatal diagnostic errors in the U.S. intensive care units equal the number of breast cancer deaths each year i.e. about 40,500 [4]. Quite often misdiagnosis is associated with indifference and formalistic approach to patients.

A mother brought her four-year daughter for a regular checkup. Part of it was the kidney ultrasonic investigation. After 40 minutes of examination, a doctor shook his head showing a worried mother a blurred picture which she could barely decipher. Instead of providing explanations, a doctor insisted on immediate hospitalization. “Surgery will show”, he consoled her. Fortunately a mother took a second opinion at the Pediatric Center. After studying attentively the child’s previous medical history and making his own ultrasound analysis, a young surgeon made his verdict: “You can take your child now”. “To the admission room?”- mumbled the mother who was half-fainted at this moment. “No, just go home”, retorted the doctor. “There is nothing serious”.

One need not only a good equipment, but the ability to understand the picture it provides. The child was perfectly healthy, there was just an inborn peculiar shape of a kidney pelvis – a poorly visible dilation. Just imagine the consequences if a healthy four-year child would have been operated. But the first doctor did not care.

4.Perceiving health as a business


Whenever nowadays we hear discussions about healthcare, it is always about costs. Often we are prescribed medications that are covered by insurances but not necessarily the ones that are needed. We receive inadequate hospital care allegedly because there is not enough money to hire a professional nurse.

Who would expect that a male patient in the intensive care unit (ICU) right after a complex open-heart surgery (five bypasses) in one of the well-known German hospitals (Krankenhaus Bogenhausen in Munich) would receive an oxygen mask with no oxygen? When he tried to push the mask off, 3 nurses on duty that night, jumped on him and forcefully tied him to his bed. It was a sheer luck that a man has survived that night, the first one following the operation. When disturbed relatives demanded explanations the next morning, the Chief Medical Doctor simply responded that the hospital did not have enough money to keep professional nurses in the ICU, so some students were hired. Although the case was outrageous to the point of being utterly criminal, the hospital and its insurer did everything to hush it.

At the same time we see that insurances often generously spend a lot of money on surgery procedures which are not really necessary. Thus, in Germany according to AOK Health Insurance 2014 report, which has provoked indignation among the physicians and hospitals alike, annually there were 19,000 preventable hospital deaths in the country [5]. For a comparison: car accidents took away the same year 3.290 lives.
One of the reasons for such appalling results (in particular, discussed in the German ZDF news program) was the amount of unnecessary surgeries performed in German hospitals that expose patients to the risk of infection, collateral organ damages and finally death. According to the ‘Medical Experts Online’, a company that provides patients a platform for a second medical opinion, in 66% of cases the first recommendation in favor of surgical intervention was found inappropriate [6]. The situation in the US seems to be no better. In fact, unnecessary surgeries might account for 10% to 20% of all operations in some specialties, including a wide range of cardiac procedures [7].

So why insurances are spending money on procedures that are not only unnecessary, but harmful? The more money insurances spent, the more likely they receive additional funds the next year. On the other hand, hospitals are rewarded with a premium from each operation. Considering that each surgery costs on average about 40K euro, this is quite a lucrative business. The only unhappy creature is a patient who ultimately has to pay with decreased health and with his money – an increased insurance premium.

All of the above: indifference, business-like attitude towards health increase the risk for a patient to become a victim of medical errors. Over 23% of the European Union citizens, according the “Evidence on Medical Errors”, shows that 50.0% to 70.2% of such harmful events could be prevented [8].

What patients and physicians can do about it?

•Work tightly directly together skipping bureaucratic mediators. Doctors need to listen to their patients and help them with intelligent questions to relate their story. Modern online collaborative tools such as Microsoft Personal Health Record Health Vault, US clinical decision support system UptoDate and the likes  can help a lot.
•Unfortunately, patients need to look for a second opinion. Ask why surgery is necessary, verify diagnostics with other doctors. Platforms like Medical Experts online or BUPA Medical second opinion service could be helpful.
•Patient have to educate themselves. Although an educated patient can be perceived as a challenge for a doctor, patient is in fact his best partner. There are many professional applications like NHS Choices, US Center of Disease Control and prevention (CDC) now available at Microsoft store that can help people preparing themselves for a meaningful discussion with their physician.
In the next few blogs we will take a closer look at patient/physician relationships technologies.


Preparing for a flood of personal medical Big Data

Patient generated information coming from various wearable medical devices or manually imported and collected via mobile applications such as Personal Health Records (PHR) is expected to change our approach to healthcare. Instead of vague responses to generic questions kind of: “how do you feel?” patients will provide doctors with abundance of measurable data like blood pressure, ECG, glucose level, cholesterol, diets, weight, sleep patterns – you name it. From the populational, statistically oriented health our modern health system tends to shift to personalized care based on individual data. This will make diagnostics more precise, treatments more efficient and acute conditions less probable. Patients will join forces with medical professionals to cope with preventable and chronic diseases.

This is the vision. But voices of concern are coming already from both medical professionals and consumers. The flow of personal data is going to generate a flood of Big Data people will be supplying daily, hourly, from homes, during vacations, company meetings, sleepless nights and so on. Already today 80% of patients would like to use their smartphones to interact with the doctors [1]. Ask your local GP if she/he is ready to receive numerous calls from patients worried about their ECG, blood pressure, BMI or stress level.

The answer is PHR, the tool that is expected to help patients tracking their health conditions better while preparing them to be a reliable partner to physicians in navigation through the ocean of personal medical data. “Properly designed and implemented, PHRs can help patients manage their health information and become full partners in the quest for good health” [2].

But does the existing concept of PHR really help people to turn the abundance of data into meaningful information to understand their health?

From data storage to data analysis and decision support.


The majority of consumer oriented PHR applications are based on three fundamental conceptual pillars: data collection, data protection and data storage. Preparing to this blog I went through 200+ applications that allegedly “every hospital market should know” published by the MobiHealthNews as the result of their “exhaustive search of Apple’s AppStore and the Google Play store for apps that were developed by or on behalf of hospitals and healthcare systems in the US” [3]. Despite the fact that those apps were presumably developed on behalf of hospitals and thus heavily designed to facilitate appointments or to find a physician or a facility location, about 30% of them were clearly consumer-oriented PHRs. Moreover, many applications are variations of basically the same app. The prevailing approach: patients can store their vital signs, lab results and medication in one single PHR, but often in different places with no meaningful links to each other. One is still getting a static picture that does not allow to understand how your health reacts in time if one of those values varies.

The typical example of disconnected applications are blood pressure trackers. The majority of them contain only 3 fields: “systolic”, ‘diastolic” and “pulls”. At best, one can put manual comments. You can see what is going on, but with no idea why. For example, if a person switches from angiotensin-converting enzyme (ACE) inhibitor Ramipril to Valsartan? What medication works better for the patient?

A solution would be a simple application backend analysis, e.g., for a period of three months to provide a valuable information on blood pressure and medication interaction, preparing a person for a thoughtful discussion with a physician on the impact of prescribed medications. It will also free the doctor’s time from using a “medication reminder” application if a given medication simply does not work.

Patients are of course grateful now at least to have some information about their health and ability to use some of the services which were not available before. But as soon as these first aspirations are saturated, patients will expect to move from mere data tracking to understanding how they can use data flows to improve their health.

Analyses of Correlations between the data, not just data itself are especially important for the people who suffer from multiply related morbidities like diabetes, hypertension or heart failure. For example, numerous statistical data indicate that a high glucose level causes vessels blockage thus increasing the risk of hypertension and heart failure. But after a series of insulin injection or metformin tablets, can your PHR clearly show that the reduction of a glucose level brings your blood pressure down? Not for general statistics, but for you personally? Just the connection of these two parameters, vital signs and medication, not mentioning the external factors, in particular, weather conditions that can influence blood pressure, may provide patients with a much better insight into their health conditions, thus stimulating PHR adoption.

Blood pressure2

Big vendors like Microsoft with Health Vault, Apple with promised Health Kit or Samsung with coming Simband are offering citizens’ platforms to store their personal data that will be collected from more and more medical devices. That will generate more and more data. Will the developers embed the Big Data analytics into the concept of their PHR applications turning them into meaningful decision support systems? Or will these applications remain secure storage and sharing platforms, with more and more data to enter and more confusion to create?

The response to this question is crucial to predict the level of patient engagement in mHealth, the boom or the slump in its future. The consumers’ expectations for mHealth are high, and the disillusionment can be deep and irrevocable.






Mobile Health understood as the ability to receive and share information about one’s health with the help of mobile devices (such as tablets and smartphones) has created almost a universal excitement. The growth of mobile applications boosted from 40.000 in 2012 to nearly 100.000 in the first quarter of 2014 [1], that is almost 2.5 times. According to the Research2Guidance study (mHealth App Developer Economics 2014.The State of the Art of mHealth App) the application market is developing with exceptional speed – 15 times faster than the growth rate of stationary internet users [1].

But has mHealth really met the expectations of patients and caregivers? Apparently not yet. The majority of applications offered to consumers were predominately designed for healthy people aimed to stay so as long as possible. Fitness and Wellness constitute the largest chunk of commercial applications on the market, with respectively 30.9% and 15.5% [1].

But what is the offering for people who already have problems with their health conditions? Elderly or chronically ill, with ischemic heart disease, COPD or stroke? Although there are promises that applications for chronic diseases will overwhelm the market with 31% share outpacing fitness by 3%, there are still concerns about the quality and functionality of those applications. Will they help people to diagnose their illnesses earlier? Get a better treatment? Or will they remain simple trackers of multiple data assembled by patients from all sorts of medical devices? Data that doctors seldom seriously consider?

I have tried myself several heart rate apps from different app stores. Almost all of them showed drastically different results being extremely sensitive to the exact finger position or pressure inadvertently exerted on the camera. Often the results indicated that I was more dead than alive with my pulse at the level between 30 and 40 bpm.

Is it not the reason why from 40,000+ mHealth apps available in 2012 only two thirds were actually used after being downloaded? 26% were used only once, and almost two thirds were abandoned by the 10th use. The most frequent reason of abandonment: the app was not easy to navigate or lacks useful functionality [2].

The skepticism in the clinical value of the majority of commercial mHealth apps is supported by the results of the study Comparison of Mobile Apps for the Leading Causes of Death among Different Income Zones: A Review of the Literature and App Stores in June 2013. After reviewing commercial applications from the major app stores (Google play, iTunes, BlackBerry World, and Windows Phone Apps+Game) researchers from the University of Valladolid, Spain came to the conclusion that only 557 applications (out of over 2000 they have initially picked up) were helpful to tackle the leading causes of death in the World. True, only English language applications were considered, but nevertheless given that the total number of commercial health apps at that time was approximately 40.000, less than 1.4 % being related to top major killers of the world seems rather meager.


Distribution of apps by diseases. The first number before the slash indicates apps that according to criteria of researches have relevance to the given disease. The second number – total applications available in the store allegedly addressing the given disease.

Distribution of apps by diseases. The first number before the slash indicates apps that according to criteria of researches have relevance to the given disease. The second number – total applications available in the store allegedly addressing the given disease.


Applications related to medical diagnostics so far constitute only 1.4% of the total commercial health apps according to the study [3].

With the emergence of non-invasive medical sensors like ECG or glucose meters, there is a hope that the new wave of health diagnostic applications will appear on the market. At least about half of apps publishers emphasized that the availability of devices and sensors to be connected to an app is the most important requirement for a mHealth application.

InstaPulse® Heart Rate Monitors can measure your heart rate simply, instantly and accurately during aerobics, or while jogging or walking, without requiring a cumbersome chest belt or ear lobe attachments.

InstaPulse® Heart Rate Monitors can measure your heart rate simply, instantly and accurately during aerobics, or while jogging or walking, without requiring a cumbersome chest belt or ear lobe attachments.

A prototype approach to sensor driven solutions was exercised already in 1980 by Biosig Instruments, a Canadian company designing and manufacturing electrophysiological fitness systems. This company has developed a number of devices performing sensor measurements of the body health parameters.

Later the Israeli company LifeWatch, has produced 1 and 3-channel ECG designed for remote arrhythmia monitoring in any location. A German company Infratec, has already in the beginning of 2000 developed a monitor that uses principles of thermal emission spectroscopy and noninvasive measurement of tympanic membrane glucose concentration in diabetes patients [4]. The list of the examples can be prolonged.

So what does stop us from mHealth Eldorado? Four things are coming to mind.

  • Reliability of sensors and devices. Today even standard blood pressure devices may show the results differing by 20mm/Hg, this substantial inaccuracy being considered an acceptable error. So if a blood pressure is 140 or 160, it is within the same error margins i.e. basically the same value. Will physicians trust the results of portable mobile ECG or blood pressure devices obtained with such accuracy?
  • Understanding that patient is not a hub to stuff his living environment with an abundance of applications and sensors. A widely acceptable secure and open platform is required to provide people with an easy single entry point connection to multiple applications, databases and devices through open APIs.
  • Visibility and easy access to mobile health applications instead of a prolonged search through multiple sites and App Stores.
  • Sustainable business model that will allow physicians to receive reimbursements for online medical services they provide as well as the clear understanding of responsibilities. For example, if remote diagnostics had a negative impact on a patient, who is responsible? The distant consultant, physician in charge, an application or a system developer?

But the new wave of much more sophisticated sensors related to diagnostics is irrevocably approaching. The scientists from the Weizmann Institute in Israel have recently made a breakthrough with a microscopic device that operates autonomously inside bacterial cells. The device“scans” the cell to see if all genes in it are expressed as they should. The detected malfunctioning molecule will cause a disruption in gene expression thus allowing one to diagnose the danger on a very early stage [5].


The recent Samsung Simband cloud platformthat will enable to “gather vital diagnostic information – from your heart rate to your skin’s electrical conductivity, 24 hours a day, seven days a week” is one of the newest promises of the sensor driven healthcare.

Apple is about to announce in October 2014 its mobile HealthKit that will connect a vast array of healthcare devices and applications, from monitoring your activity level, heart rate and weight to your chronic medical conditions such as high blood pressure and diabetes.

microsoft-sufrace-smartwatch-1Microsoft is coming with a Surface Smart Watch that will include multiple sensors such as heart rate monitor, accelerometer, gyroscope, GPS and a galvanic skin response detector (to measure the changes in the electrical resistance in the skin using sweat). The received integral information can be connected to Microsoft Health Vault, an open platform for storing and sharing health related data like images, medications and vital signs.

If all promises are fulfilled, a different paradigm of healthcare will evolve as compared to today’s mainstream medicine. The development of non-intrusive sensor-based techniques of picking up parallel information from multiple body areas in real time, thus measuring the distributed state of human health can provide better understanding of the organism stability margins, the emergence of diseases as well as ensure more patient safety under drug and physiotherapy prescription [6].

Hopefully, with big vendors like Apple, Microsoft or Samsung stepping into the game, the barriers will be surpassed and some of us may enjoy the transformation of Mobile Health into the trustworthy Measurable Health.


1.mHealth App Developer Economics 2014.The State of the Art of mHealth App Publishing

2.MobileSmith. Patient Engagement and Mobile Aps

3. mHealth App Developer Economics 2014.The State of the Art of mHealth App Publishing



6.Mobile Health: A Conceptual view. Horizon Research


As a family caregiver I need to see the results of many medical tests such as computer tomography (CT), ultrasound or magnetic resonance images performed by different medical specialists often from a variety of medical institutions. It seems quite natural and feasible,  in order to get insight into patients health conditions, considering all modern technologies that rapidly appear on the market.

Reality, though, is different. I vividly remember myself roaming between cardiology, neurology and archive departments of one of the most prominent hospitals in Munich, Germany – Bogenhausen Krankenhaus – seeking for the results of radiography performed on a patient after a complicated cardio operation. Despite the appalling fact that at least five X-ray tests were conducted within a week – each day an X-ray investigation without notifying the relatives, all obtained images mysteriously disappeared.

The next unforgettable experience was in the German Cardiocenter (Deutsches Herzzentrum) in Munich where we arrived carrying a CD in our hands, with the cardiovascular image produced by the gamma-camera at local nuclear medicine facilities about 20 km from Munich. To my big surprise, a cardiology “expert” bluntly returned the CD with a barely disguised embarrassment requesting “written comments” from the physician who had created it. That presumed to drive back to the small city of Unterschleißheim asking for the recorded CD image to be explained by words instead of looking through the authentic images that carry much more information.

I do not want to question the competence of experts from the German Cardiocenter. I suspect the issue was more technical than professional, requiring the familiarity with certain viewers. Even if our doctor in this respectable institution had an appropriate software on his PC that supports DICOM (Digital Imaging and Communication in Medicine), the standard itself is quite complex and needs a prolong training. Even major vendors have their own modifications of MRI and CT scanners which export DICOM data in proprietary ways so that one can hardly blame physicians [1,2].

Available medical imaging modalities.

As patients we  are  wondering: which medical examination is better and more accurate for us? Should we go through  CT or MRI? Though the final decision is with our doctor, it is important for patients  to understand the pro and contras  of the imaging techniques.

Modern imaging studies of physiological objects are using different physical principles depending on the specific location of the body organ to be investigated and the quality of image required. Quality in this case means the contrast and the resolution of a picture reflecting different levels of physiological information.

For example, the system described in plain radiography reflects the absorption of X-rays by different tissues giving a 2D image of structures within a body. While passing through more dense and structured tissues, e.g., through bones, X-rays are scattered, diffract and dissipate thus creating contrasts between the body objects (dark vs. light shades). That is why X-rays tests are more frequently used to locate fractures or bone pathologies accompanied by density variations, lung pathologies or abdominal obstructions.


A more advanced CT system computed tomography is also based on the contrast between the unperturbed and attenuated X-rays, but here X-rays are emitted in different angles thus creating cross-sectional view of body slicers. The resulting image is produced by a set of attenuation coefficients integrated along the irradiation lines. Given a large series of 1d or 2d projections obtained by scanning a patient from various directions, a 3d image of an investigated object (organ or a body structure) can be restored using some well-known mathematical (Radon) transform. Often this method is used in combination with special contrast agents. Due to the cross-sectional (slice) nature CT images are more informative, and applicability to medical fields can be ensured: muscle or bone disorders, looking for tumors or a blood clot, injuries to internal organs (kidneys, liver, spleen, etc.). CT is also used today to look into the heart (cardiac CT). The CT method has some obvious disadvantages as well: due to the multiple exposure to radiation, the doses received by a patient are much higher than in other radiographic techniques.


Contrary to CT, magnetic resonance imaging (MRI) is using strong magnetic fields combined with a relatively weak radio frequency wave signal to produce real time 3d images of the inside of the body. Magnetic field applied to the body aligns vectors of magnetic moments (proportional to nuclear spins) along the direction of the magnetic field. The image contrast is determined by the signal difference from various parts of the examined organs, this signal difference mainly depending, in its turn, on the relative density of hydrogen nuclei (protons) in the tissue: magnetic moments of protons in such tissues align themselves easier, mostly due to simplicity of the hydrogen atoms. Thus, tissues containing hydrogen or water (and this is approximately 70% of the human body) produce more contrast images. That is why compared to CT, MRI gives a better insight into soft tissues such as muscles, joints, breasts, heart and blood vessels or most internal organs, e.g., the liver, womb or prostate gland. Nowadays special chemical agents exist that are designed to enhance the contrast.

The disadvantage of MRI compared to CT is that it takes longer time for examination (about 30 min. compared to 5 min. for CT), thus MRI is difficult to use in emergency cases, e.g., stroke or potential brain bleedings which obviously requires faster reaction [3]. It also contradicts with metal body parts or implants, e.g., cardio pacemakers. On top of that, MRI examination is quite sensitive to patient movements, so a patient should remain still during the whole investigation in order not to deteriorate the image quality.


Advances of Nuclear Medicine open new opportunities for a deeper insight into human body. A sensitive crystal embedded in a gamma camera detects the distribution of tracers injected into a body. The results are converted into a digital format to produce 2d and 3d images that reveal the cellular level metabolic changes occurring in an organ or tissue. That is why nuclear imaging methods such as Positron Emission Tomography (PET) can often detect changes earlier than CT or MRI and thus works better for an earlier diagnostics [4]. One of course has to be cautious of the potential allergy on isotopes that a patient is to swallow.

The widely used ultrasound scan or sonogram utilizes high frequency sound waves that bump into tissues creating an echo captured by the computer in the real time. Though relatively safe, this method has some disadvantages as well. The correct ultrasound image strongly depends on human operator, since plenty of artifacts may appear on the picture as the incidence angle changes slightly. Ultrasound tests rely on acoustical wave reflection from inhomogeneities so that test results may identify a potential area of concern, but often one is not capable of telling, based solely on ultrasound examination, if some lesion is malignant or just a physiological peculiarity of a given organism. The false-positive results might require more tests and even further complicated procedures, up to biopsy or surgery.



To give credit to this method it went through various stages of improvements since its discovery in 1953. By combining the traditional ultrasound method with Doppler blood flow investigations (and some specific improvements such as, e.g., transesophageal viewing), engineers and physicians have achieved ultrasound images with much better resolution compared to the initially obscure shapes. Thus, the mentioned transesophageal echography involves a flexible tube (probe) with a transducer at its tip. The probe is guided through the throat and into the esophagus (the passage leading from a mouth to a stomach). Because esophagus is located right behind the heart, the physician can get a much better view of the heart and blood vessels [5]. The approach is primarily used in cardiology under the name of transesophageal echocardiography. Though more informative compared to traditional cardio ultrasound tests, this method loses the biggest advantage of ultrasound: noninvasive approach to examinations.

From many diverse images to  a single comprehensive one.

All of the above mentioned medical imaging modalities have their advantages and disadvantages, but none of them alone could provide a physician with the sufficient insight to arrive to a final diagnose. Patients, in their turn, have to go through multiple tests often associated with stress, side effects and potential health damages. Basically, what all of us need is a single, consolidated, all-in-one image of a suspicious zone or internal body organs that would aggregate multiple imaging data generated by different modalities such as CT, MRI, PET, ultrasound, etc. into one picture, without losing or replicating data.

Luckily for clinicians, there are certain steps done in this direction based on the 21st century medical imaging fusion technologies. For instance, big companies such as General Electric Medical Systems and Varian Medical Systems set up a joint program called See and Treat Cancer care that combines metabolic (PET) and anatomical imaging (CT, MRI) examinations with intensely modulated radiation therapy (IMRT) to better treat cancer [6].

Others start producing software that allows to view a variety of scans no matter where they are located from one entry point. For example, the Germany-based Brainlab company, Brainlab Buzz has created a single hub designed for operating theaters to view various medical images like CT, MRI, and Ultrasound on a single screen from a dedicated workstation of just a PC.

Another step forward is the tool designed by the Canadian company Calgary Scientific ResolutionMD that retrieves data from multiple PACS (Picture Archiving and Communication Systems) archives no matter where images are located. The application is Web enabled so that images can be accessed from any PC, tablet or mobile device (with no right to download data to any of device in order not to compromise privacy). Despite the fact that all those technologies are making the life of a physician much easier, they are still used for clinical purposes only being unavailable to individual patients.

Three criteria to meet patient’s expectations.

Visualized information is much easier to perceive than multiple text documents. Once to see is better than one hundred times to hear. Therefor, for patients and caregivers computer applications with visualized representation of their own inner organs that maybe damaged  are crucial to understand and manage diseases. Unfortunately, such applications are absent as yet.

 It is, however, possible that personal medical imaging applications based on image fusion will appear as extensions to routine clinical software to be used by patients in cooperation with their physicians.  Nevertheless, these new applications have to meet three main criteria for a patient to be able to consume them.

  1. All-in-one multimodality image of your inner self.

So far applications that allow patients to see the results of their personal MRI, ultrasound or CT examinations are a rarity. Sure, there are solutions like, e.g., designed by Microsoft Health Vault for uploading DICOM images to a personal electronic health record stored in the Cloud infrastructure and to be shared with family caregivers, home doctor or nearest emergency center. But those are still sequences of separate X-ray images that do not produce the holistic picture of the inside structure of an individual human.

  1. Clear insight into visualized organs.

Let us not forget that with all available algorithms that are reconstructing medical data to create sophisticated images, the final algorithm is our own brain nourished with our life-long thesaurus. Just as snakes see objects differently from humans, physicians view medical images differently from patients. For a patient to get a deep understanding of her or his health conditions medical images have to be clearly interpreted and ideally supported by a physician’s written or voice comments. An educated patient can be a true partner for a doctor.

For example, a patient normally receives, prior to an operation, a declaration explaining what will happen to her/him. Instead of multiple drawings to discuss the pending surgery a physician can produce a single image combining visual results of medical examinations performed on this person.

  1. Ease of access with widely used formats.

Applications that require even minor efforts to access the data are doomed to die. Complexity and variety of DICOM and other imaging standards have to be drastically reduced; the standards themselves must probably be converted into widely accepted standards like jpg, png, gif, tiff, JPEG, etc. for the patients to be able to open and work with them easily. Moreover, it is important that the consolidated multimodality images obtained through MRI, CT, PET, echographic, etc. tests can be uploaded to a personal Cloud, e.g., Microsoft OneDrive or DropBox as parts of a personal record. If a patient wants to discuss the results with the professional she or he should be able to grant the latter an easy and secure access to the data.


One of examples of the multimodality imaging fusion software is developed by Manzoma Technology Solutions, Infinitt Xelis Fusion. The product enables to consolidate multimodality studies (e.g. CT-PET, MR-PET) and to visualize the combined 3D images with the aim to perform quantitative analysis based on the images obtained by different examinations, regardless of whether they are from the same modality or from different modalities.


Another product Integra ImageFusion Software is primarily used in neurosurgery. It helps clinicians to use complementary features of different scan types. For example, the offered system combines the advantage of CT spatial accuracy with the superior soft tissue MRI scan.Integra

With the advancement of 3D printers, one can even create an inner self atlas based on the results of recent medical examinations. Sounds like a dream? Not at all. I am sure that despite all technical and administrative problems somebody is working on such modeling applications at this very moment.










You should always do what your doctor is telling you, it’s a must. That is what we are persistently told. Medication compliance (sometimes referred to as “medication adherence”) normally is defined as “the degree to which a patient correctly follows medical advice” [1]. It is considered to be one of the fundamental requirements for the effective treatment especially for patients with a long term chronic disease. Medical community is more or less unanimous in labelling non-compliance as a major obstacle to the effective delivery of healthcare. Nevertheless, according to many respectable sources like WHO and Mayo Clinic although medications are effective in combating disease, their full benefits are often not realized because approximately 50% of patients do not take their medications as prescribed [2]. The statement presupposes that medications are effective no matter who and why is to take them.

The team of researchers from the University Hospital of Leicester came to a quantitative conclusion through an experimental test of compliance. A total of 40 most commonly prescribed antihypertensive medications (or their metabolites) were screened for in spot urine samples to test patients’ compliance. After conducting a research among 208 hypertensive patients (125 new referrals, 66 follow-up patients with inadequate blood pressure control and 17 renal denervation referrals) who underwent assessment of antihypertensive drug intake using high-performance liquid chromatography-tandem mass spectrometry (HP LC-MS/MS) urine analysis at the time of clinical appointment the team came to the sad conclusion: “Overall, 25% of patients were totally or partially non-adherent to antihypertensive treatment (total non-adherence 10.1%, partial non-adherence 14.9%)” [3].

non adherence

The damage of non-compliance is measured not only in medical terms, but also in terms of material losses that the healthcare industry has to bear. Thus the WHO shares the results of the studies which point that “the poor adherence to medication leads to increased morbidity and death and is estimated to incur costs of approximately $100 billion per year” [5]. The National Health Service (NHS) UK estimates the costs of medication treatment for patients with coronary heart disease (CHD) in excess of £2 billion on medicines, 50% of which is wasted through poor understanding and poor adherence [6].

But why 50% of patients even though suffering from acute chronic diseases such as CHD or, e.g., cerebrovascular accident (CVA) are consciously ignoring their doctors’ prescriptions?

Many highly regarded medical publications would name multiple reasons including a “poor communication between physicians and patients”, “lack of patients involvement in the decision process”, “patient illiteracy and lack of understanding of medication benefits and side effects”…

To avoid scholarly discussions, but as a care giver to patients with both CHD and CVA, I came to the firm conclusion: the main reason that patients do not take their medications is because those medications simply do not work for them, at least the patients fail to notice tangible improvements and have to rely on the doctors’ assertions.

For several years, I have been observing a patient with high blood pressure, cardio-vascular problems and diabetes mellitus on top of its all. The traditional medication treatment was a combination of ACE inhibitor or angiotensin-converting enzyme inhibitor (in this case Ramipril), calcium channel blocker (e.g., Lercanidipin or similar preparates) and beta-blockers (e.g., Bisoprolol).

Having taken Ramipril for 2 week the patient started complaining of unpleasant chest pains (presumably heart vessels reacted) and sleep problems. The physician’s response was unconditional and definite: “it is not possible. These side effects are atypical for Ramipril. There may be only slight cough.” Only due to the patient’s persistency, Ramipril was substituted by Valzartan, an angiotensin II receptor antagonist, more commonly called an ARB. The effect of Valsartan and Lercanidipin combination unfortunately had no effect on the reduction of the blood pressure and apparently did not help with CHD. Recently, I came across an article published in Jama Networks (thanks to the Internet and social networks!) “Effect of Angiotensin-Converting Enzyme Inhibitors and Angiotensin II Receptor Blockers on All-Cause Mortality, Cardiovascular Deaths, and Cardiovascular Events in Patients With Diabetes Mellitus: ”[7].The authors, Jun Cheng, MD, of the Medical School of Zhejiang University, China and his colleagues compared the effect of two types of medication on patients with cardio-vascular diseases and diabetes mellitus: ACE inhibitors and ARB. The conclusion was that “Angiotensin-converting enzyme inhibitors reduced all-cause mortality, CV mortality, and major CV events in patients with DM, whereas ARBs had no benefits on these outcomes. Thus, ACE-Is should be considered as first-line therapy to limit excess mortality and morbidity in this population.” For my patient, it meant that the swallowed Valsartan had probably no effect on his CHD.

Bisoprolol reduces the activity of the heart by blocking tiny areas (called beta-adrenergic receptors) where the messages are received by the heart muscle. This could provoke a bradicardia – one of the many possible side effects of this medication. We simply had to drop bisaprorol to recover the normal pulse rhythmus.

Almost all medications aimed at treating people with long term chronic conditions have severe side effects that are revealed gradually and tend to accumulate in the body.Cold sweats, fainting, fast or irregular heartbeat, nausea, shortness of breath to mention just a few are numbered as the “typical” side effects of antihypertensive medications.


But what about the ones we are buying in our nearby drug stores? Even less sophisticated widely available drugs such as ibuprofen or even aspirin could cause unpredictable side effects in certain individuals.

Jason Ryan, 28, from Washington, near Sunderland, suffered a severe allergic reaction which is believed to have been sparked by taking the over-the-counter drug ibuprofen and turned into Stevens Johnson Syndrome (SJS), which causes the skin cells to die before shedding [8].

If 50 percent of patients are not taking their prescribed medications, there is definitely something fundamentally wrong with the way our healthcare and pharmaceutical systems are working. For centuries, the traditional healthcare system was addressing a “typical” patient with a “standard” reaction on medications. This approach appears to be methodologically wrong since it is based on an implicit assumption of the universal validity of the Gauss distribution underlying medical statistics. Yet the statistics of emergent diseases, similarly to heavy accident statistics, seem to obey other distribution laws such as, e.g., power law distributions rather than sharply dropping exponential ones that imply negligible stray areas.

Unfortunately, the statistical results, e.g. produced in clinical trials, in fact do not answer the crucial question: what is the best strategy for a given patient. The variations of human organisms are much more complex and diverse to bring them under the unified umbrella of a statistically average person. This is pretty much the same as to measure “an average temperature” of all patients in a hospital – even accurately computed standard deviations would leave many “stray” patients outside of the main massive. But it is these latter patients that have to be dealt with sometimes taking much more doctors’ time than the average ones. Someone is crouching in fever, while somebody had already peacefully passed away. The average temperature though is normal.

The so-called “personalized devices” (both hardware and software) such as smart pill bottles or health feedbacks systems that are now coming out on the market are predominantly focused on reminding a patient to take prescribed medications vs. contemplating what medication a given individual requires at the current instant. Similarly, personal portals that allow patients getting a direct access to their physicians do not completely solve personal health issues: doctors will be still restricted by numerous health protocols as well as by health insurance plans. The latter will not allow to cover the variety of medications outside of their agreements with certain hospitals and/or pharmaceutical companies.


Certain hope lies on personalized approach to medications based on preemptive genotyping (PG). Thus, the Mayo Clinic initiated a study “Right Drug, Right Dose, Right Time” using genomic data of Mayo Clinic biobank participants, with a recruitment goal of 1000 patients connected to their electronic medical records (EMR). The multivariate prediction model was applied to identify patients with a high risk of hyperlipidemia requiring treatment with statins within three years. The model included 6 chronic diseases categorized by the Clinical Classifications Software for International Classification of Diseases, the Ninth Revision codes (dyslipidemia, diabetes, peripheral atherosclerosis, disease of the blood-forming organs, coronary atherosclerosis and other heart diseases, and hypertension). The Mayo Clinic researchers hope that clinical implementation of PG at the bedside could make it possible to avoid adverse drug reactions, maximize drug efficacy, and select medications to optimize effect for specific indications on the basis of the genetic profile of individual patients [9].

Of course such endeavors would require time and money. But can’t it be more efficient to invest $100 billion per year in the therapy that brings the result helping people to survive than totally wasting the money?


  4. Osterberg L, Blaschke T. Adherence to medication. N Engl J Med. 2005;353(5):487-497. [PubMed]


People who once struggled through stroke are persistently chased by a burning question: shall it ever happen again? How can I avoid it?

Obviously there are some worrying signals of potentially revocable brain damages such as increased blood pressure, sudden nausea or dizziness one needs to watch, but the accuracy of stroke predictions based on occasional body symptoms i.e. external to the brain is rather dubious. All the above mentioned signs can be attributed to disorders of a completely different nature.

Unfortunately, today’s medicine is not able to precisely predict the possibility of a repeated cerebrovascular accident. At best, neurologists can capture its first manifestations based on cognitive or motor disabilities with the help of Computer Tomography (CT), Electroencephalography (EEG) or Magnetoencephalography (MEG). All those methods have their limitations. For instance, the CT X-ray radiation – as many other invasive interventions – can ultimately be damaging for delicate cell structures inside the brain. Ionizing parcels of hard electromagnetic radiation can bump with the cell DNA, causing damage that may lead to cancer. EEG registers mainly the neuron activity on the surface of the cerebral cortex, giving little information about subcortical neuron activity so that clinicians have to extrapolate the picture. MEG is capable of registering signals produced by the currents excited in subcortical areas, thus providing 3d information, but this method is so far quite cumbersome requiring a lot of space and extremely expensive, so not every hospital can afford the MEG equipment.

Could the future technologies offer patients and physicians a more affordable and accurate way to predict the threat of a repeatable stroke early enough to avoid it? I think now we can observe the first signs of this prevention.



The group of BCI researchers from Colombia, Jhon Edison Muñoz Cardona, have designed a novel Brain Computer Interface (BCI) or rather a Brain Kinect Interface (BKI) system which combines biomechanical signals coming from Microsoft Kinect sensors, brainwave signals acquired from Emotiv EPOC EEG [1] and the so called Steady State Visually Evoked Potentials (SSVEP) signals that are our natural responses to visual stimulation at specific frequencies [2].

While a patient is immersed in a rehabilitation game with a predetermined visual stimulus, e.g., trying to manipulate objects with hands motions or eyesight, Brain Kinect Interface is tracking the dependencies (correlations) between the visual, cognitive and motor signals generated by a patient.

Every stimulus is associated with a command that controls a specific action inside the video game. By registering and analyzing data reflecting motions together with visual reaction in combination with EEG signals, the therapist can get a better understanding of what areas of the brain are exactly responsible for a certain motional or visual stimulus and how they are affected by the game. For example, if a patient has certain difficulties in raising an arm due to the acquired paresis, which part of the brain has been damaged by the stroke and is now responsible for the disorder, and is there any progress in the course of rehabilitation therapy?

Brain Interface


Who knows, maybe due to more and more precise mapping between bodily functions and brain topology in the not so distant future one can even address an inverse problem: by tracking external motional manifestations to reconstruct the activities deeply hidden in the brain. Changes in motion patterns would exactly indicate the brain damage within certain areas.

Of course, the ambitious attempt to diagnose brain failures with non-invasive methods using standard, almost consumer-level technologies will take time and require the development of a new generation of highly sensitive, accurate and miniature sensors vs. expensive and bulky contemporary MEG systems.

But there is a noticeable progress in this direction as well.

One of the examples is a miniature atom-based magnetic sensor developed by the National Institute of Standards and Technology (NIST) that was successfully tested already in 2012 as an instrument to measure human brain activity. Experiments verify the sensor’s potential for biomedical applications such as studying mental processes and advancing the understanding of neurological diseases [3].NIST and German scientists used the NIST sensor to measure alpha waves (the deep relaxation wave (7.5-14Hz))in the brain that arise, e.g., when a person is opening and closing her/his eyes.


Signals resulting from stimulation of the patient’s hand were also explored. The measurements were verified by comparing them with the signals recorded by SQUID systems (superconducting quantum interference device) SQUID, the world’s most sensitive commercially available magnetometers that are considered the “gold standard” for such experiments. The NIST mini-sensor is slightly less sensitive than SQUID as yet, but has the potential for a comparable performance while promising advantages in size, portability and cost. Many other similar experiments are on the way.

Would it be possible to use exergames with personalized exercises to diagnose, treat and ultimately cure patients with the conditions caused by the stroke (e.g., hemiparesis), brain trauma, Parkinson’s disease, sclerosis and other neuropathies?

15 million people worldwide who suffer a stroke each year are looking today towards the upcoming new technologies and medical studies with hope and expectations mixed with anxiety [5].



  1. BKI: Brain Kinect Interface, a new hybrid BCI for rehabilitation .J. Muñoz, O. Henao, J. F. López, J. F. Villada. Games for Health Proceedings of the 3rd European conference on gaming and playful interaction in healthcare.
  2. Steady state visually evoked potentials Wikipedia
  3. Human Computer Interaction Group
  4. NIST Mini-sensor Measures Magnetic Activity in Human Brain
  5. World Stroke Organization











In my recent blog we spoke about people hospitalized with an acute stroke, but deprived of a timely medical assistance. As a result they were discharged with long-term disabilities. The usual path for those people is the rehabilitation clinic (the German abbreviation: REHA). One of such REHA clinics I have visited is located in Bad Aibling, in the heart of the picturesque Alpine meadows of Bavaria. The offered rehabilitation programs are extensive, but expensive. Health Insurances often allocate limited budgets to cover a full recovery process, and patients are soon released to the homecare of their families.  Only in extreme cases, certain patients are eligible to receive some support from their local community services.

What can families do to continue with professional and affordable rehabilitation care? How to control that exercises assigned by a physiotherapist are performed correctly and the physical stress matches the capacity of an individual patient? Alternatively, muscles will contract resulting in spasticity, the side effect extremely difficult to get rid of.

I had a chance to discuss the issue with Pablo Gagliardo, from the Spanish company Fivan. To support patients and their families both in hospital and at home Fivan has designed Neuro@Home – a telerehabilitation platform, aimed to treat individuals with neurological condition. The system includes more than 100 rehabilitation tasks destined to treat specific motor or cognitive functions by using the virtual reality and computer technologies with natural interface.

Fotogramas Neuroathome Video_09

The therapist in the hospital assigns a rehabilitation program to a patient based on her/his physical conditions. The program consists of a series of computer-based exercises controlled by body gestures. Every time a patient moves, his avatar reproduces  gestures on the TV or computer screen. By watching the progress, medical professionals can adjust the program to the physical conditions of a person at a given instant. Once this is done, Neuro@Home stores a detailed session report in the patient management application. While patients carry out scheduled tasks, the therapeutic applications record their progress and transmit data back to the clinician.

Each rehabilitation task performed by a patient is quantified: Neuro@Home Pro measures the number of successfully completed tasks, response time and total number of completed tasks. The rehabilitation team can evaluate the progress not only by observations, but by “measurable” facts. New rehabilitation tasks could be added, existing ones modified or the rehabilitation task can be entirely stopped if not considered useful anymore. In a word: the therapy can be personalized and adjusted to the individual needs of a patient.

Fivan is now offering this program as online services for a very affordable monthly fee to families whose members were affected with long-term cognitive or motor disabilities. By subscribing to Neuro@Home a neurological patient can get a daily access to personalized exergames therapy with the assistance of a professional physician.


Neuro@Home is based on Microsoft Kinect for Windows sensors that can capture the slightest body motions.  Recently, Microsoft has launched a new version of Kinect sensors that can significantly enhance neurorehabilitation programs. Physicians will be able to track facial contractions to measure the emotional impact of the exercise (e.g., stress or fatigue), review rotation angles of the joints and  detect changes in the skin tint (such as, e.g., redness or paleness) to judge the variations in the heart beat.

The first clinical studies using Neuro@Home were conducted at a post-acute and long stay hospital in the Valencian Health Agency and have been presented at the International Brain Injury Association’s World Congress in San Francisco.

The studies covered cognitive rehabilitation of 12 patients (4 women and 8 men) who had suffered a stroke with a mean age of 56 years old. After receiving 40 one-hour sessions during two months (5 days a week), significant improvements were observed in the participants attention, working memory and executive functions.

Similarly, 33 patients (10 women and 23 men, mean age = 58 y.o.) who were accepted into an inpatient rehabilitation programme after suffering a stroke participated in a physical rehabilitation study with Neuro@Home. In this case, significant improvements were observed in patient’s balance, coordination and gait. As a result, both studies reported that patients also obtained significant improvements in their functionality.

Less stress, more enjoyment, is it not what helps people to recover faster?

For more information please contact Pablo Gagliardo

Stroke is one of the leading causes of deaths taking away 650.000 lives in Europe annually [1].   After the first warning signs (e.g., sudden headache, weakness of an arm or leg, trouble of walking, seeing, loss of balance or dizziness) are detected there is a time window of approximately 3 hours when the damaged tissues could still be recovered. If the appropriate measures are not taken, the brain tissues can simply suffocate without the oxygen supply. Within those 3 hours a physician in the hospital has to make a computer tomography (CT) of the brain to determine the nature of the stroke. In case of ischemic (which is over 80% of all strokes), some of the blood vessels are blocked by a thrombus (fat or blood clot) that disrupts the blood flow from certain areas of the brain.  The thrombolytic drugs can reestablish the blood flow to the brain by dissolving the clots, which are blocking the flow. But the same treatment could be killing for patients with hemorrhage stroke when a weakened blood vessel ruptures and blood is spilling out into the brain. Stroke Unfortunately not all hospitals nowadays  have stoke units at their disposal or specialists who can provide a CT image and carefully read it. As our friend called us on Sunday evening suspecting stroke we knew we need to react immediately and have connected to the closest call center in the area. That was the first mistake. The person in the call center probably had some doubts in regard to the diagnosis and advised to call a home doctor. The home doctor arrived in 20 minutes. At that time the systolic blood pressure raised above 200, but nevertheless our friend had no problem with speech, movements or walking and could assess the situation very well. Since the blood pressure was still high the ambulance was summoned: the local hospital in Freising close to Munich, Germany, which was the regional center of the area.

That was the second mistake. Sunday evening is the worst time getting to the hospital: most of the specialists are spending time with their families and friends. A lonely and tired assistant doctor on duty was the only one left for the emergency cases. The doors of the admission department have opened swallowing our friend, with the family left behind. We were waiting outside watching some of the physicians and nurses silently wandering through the lobbies and looking through us. No information was available, but we hoped the appropriate measures were taken: after all, the doctors should know that the stroke has to be treated fast.  After 6 hours the wife was allowed to enter the intensive therapy.  Her husband was sitting on a coach pleading to the staff  “to do something”. The lady doctor on duty was a bit irritated since she was busy with writing, obviously preparing reports. She tried to convince the patient’s wife that there was no need to worry: after all “the rehabilitation system in Germany is so good, and moreover, three hours have already passed”. The CT equipment, although available in this hospital, but there was nobody to use it or, equally plausible, no one wanted to take a risk of a decision. The sleepy neurological team arrived the next morning after a well-deserved weekend. They made a CT and diagnosed the ischemic stroke. At that time there was no doubt: the patient had a full left side paresis.

The experience we went through was a classic example of a medical error due to the late provision of services. The conclusions we drew later: 1) Never call the local emergency number: go for any of the numbers in your country that connects to the central emergency services (US 999, UK 911, Germany 112, etc.). At least they will try to get you to the hospital with the available stroke unit. 2) Don’t be shy just sitting and waiting. Demand and insist on information about the actions taken – up to the decision of calling another emergency service. Although the doctors and the medical staff intensely dislike demanding patients, it is serious: you will not get another chance. 3) Always keep a bottle of a good Cognac at home: in case of a high blood pressure or of a peculiar sensation (e.g., numbing or an awkward sticky feeling in your head), 20 ml will quickly widen the vessels and the clot might get through. This does not exclude the call for a professional assistance, but may increase the chances for the recovery or survival.

Our experience with the German hospital in Freising was frustrating. Should the anti-stroke measures have been taken properly and on time, a patient could have gone out of the hospital in a couple of days. That would cost the health insurance about € 400-500 whereas getting into the full spiral of rehabilitation process could be roughly estimated within the range of € 50.000 – 60.000. If the governments want to address the cerebrovascular accidents (CVA) more seriously, they actually have all the means and technologies to do it. For example, in England, where stroke is costing the National Health Services (NHS) over £2.8 billion [2], the relevant strategy was launched in 2007 to set a clear direction for the development of stroke services over the next 10 years. The strategy was developed in partnership with representatives from stroke charities, stroke professionals in the NHS, social care professionals and those affected by stroke.

Using a telestroke solution designed by MultiSense Communications based on Polycom telemedicine platform, the NHS Cumbria and Lancashire Cardiac and Stroke Network (CSNLC) established the telestoke network. It enables a specialist to view remotely the latest patient MRI scans, vital signs and medical history and to confirm if they should be thrombolysed or not. As long as there is a wireless or mobile connection, experts and clinicians can be reached at their homes at any time of the day or night to remotely recommend treatment for a stroke victim. Once the CT scan results are ready, the remote expert receives a video call on his mobile device (Windows Phone, iPhone, iPad, MS Surface, etc.) where the nurse shares the scan image. The specialist can check the impact of the thrombolytic medication, for example, by viewing via his mobile device the dilation of the pupils, which indicates the treatment progress. Polycom 15 round the clock consultants in NHS trusts of Lancashire & Cumbria are on the call from their homes to support remotely approximately 2.2 million people across the region. Installed at the end of July 2011, the service has already positively impacted the treatment of 26 patients by providing out of hours thrombolysis treatment. In addition to huge patient benefits, the service is anticipated to save Lancashire & Cumbria NHS Trusts over £8 million a year. If proven examples already exist, why not to install similar remote consulting services for emergency case in every clinic as the standard part of their medical equipment pool? For sure, it would be less expensive than complicated rehabilitation treatments, without mentioning the patients who will have much better chances to walk out of the hospital with no long term disabilities? Since some local structures (hospitals, insurances, and rehabilitation centers) can oppose such modernization, this should be a governmental program – as it would save a lot of budget costs (taxpayers’ money).

It is here that the governmental regulatory functions can be useful and not stifling. For years we hear long discussions that health insurances are not sure how to compensate the telemedicine services. Well, they have to think faster. If it is finally us, the citizens and eventually patients who are paying to health insurance funds, directly or through our taxes. Is not our right to demand the better quality of services we anticipate?


  2. Statistics referenced from NHS, National Stroke Strategy