Community Research and Development Information Service - CORDIS

H2020

MeDX Report Summary

Project ID: 674459

Periodic Reporting for period 1 - MeDX (medx 2020 | Visual Reasoning Support for Healthcare Professionals)

Reporting period: 2015-07-01 to 2016-06-30

Summary of the context and overall objectives of the project

"Started in July 2015, Medx2020 is a phase 2 year SME instrument project funded by the EU within the H2020 framework for research and innovation.
Originally thought to develop a unique visual reasoning tool for medical professionals that offers diagnosis decision support at the point of care, Medx2020 is now in its midterm phase. Misdiagnosis is has proved a shocking reality, the US have been among the first to recognize it.
"Getting the right diagnosis is a key aspect of health care: It provides an expla¬nation of a patient's health problem and informs subsequent health care decisions.
For decades, diagnostic errors — inaccurate or delayed diagno¬ses — have represented a blind spot in the delivery of quality health care. Diagnostic errors persist throughout all settings of care and continue to harm an unacceptable number of patients" [the IOM report]
Diagnostic errors are often incredibly harmful to patients as they may lead to delays in treatment, lack of treatment, inappropriate, or unnecessary treatment. This, in turn, can have physical, psychological, and financial consequences.
Causes are varied but include inadequate communication between physicians and patients, a health care system design that does not support the diagnostic process, limited feedback to clinicians about diagnostic performance, and a health care culture that discourages transparency, so diagnostic mistakes are typically not reported (and not learned from).
Dr. Peter Pronovost, director of the Armstrong Institute for Patient Safety and Quality at Johns Hopkins, told CNN:2
"It's probably one of the, if not the, most under-recognized issues in patient safety… Much of the harm that we once labeled as inevitable we're now seeing as preventable".
But Europe is not lacking behind, and the leading healthcare organisations are making plans. Back to 2009, the Telegraph posted that one in six NHS patients are 'misdiagnosed' [telepgraph 2009], recognizing the problem in its worrying dimension.
Despite of all efforts, a lot remain to be done and the problem is far from being solved.
After 1 year of implementation Medx2020 matched its original purpose and went beyond!
“Ada”, our intelligent and caring chatbot, is now online! Find it at www.ada.com
Ada is the result of an intense year of work, hand in hand with worldwide recognized experts on the field, as well as with relevant healthcare organizations and the patient, of a team of computers scientists and medical doctors of our team.
Ada is a new approach to healthcare.
Ada is more than just an app. Ada is a virtual health companion on the go and gives you the best symptom assessment available building up a detailed picture of the patient health over time, Ada provides the patient with an accurate understanding of what is going on and helps him decide on next steps to take.
Thanks to Ada, anyone can check symptoms wherever and whenever he/she wants. Just answering simple, personalised questions about his/her health, Ada asks the patient all that is needed to really get to the bottom of what is bothering him/her. If, needed, Ada can pre-briefed doctors and book a video consultation with a doctor of the patient choice, at a time that suits him/her.
Finally Ada make available your digital health record in the patient pocket, secure and always up to date.
Thanks to the Medx2020 Ada reasoning technologies can also provide disease monitoring and prevention through integration of different data sources such as sensors, lab tests and more.
Medical doctors and specialists can find Ada extremely helpful in managing their patient in due time, and handling difficult diagnosis. Doctors are often overloaded by the number of patients they follow, Ada is a convenient way to work from the comfort of their home or any location with a stable internet connection.
The reasoning technology behind Ada is being developed as well within the MeDX2020 project. Originated by a collaboration among the co-founders, the reasoning engine is being now refined, and is capable to include complex pathophysiological connections.
MeDX2020 reasoning technology is based on Bayesian neural network approach, enriched by unique features that enable what we call “constellatory thinking”. Our brain, largely works out of analytical logics, grasping the complexity of everyday life meaningful constellations. The patient condition, is such an example. Especially when dealing with difficult diagnosis, such as for rare or chronic diseases, the brain cannot handle the different symptoms one by one and get to a sound conclusion, there are too many. The difficulty of the doctor is to grasp this complexity, giving it a meaningful outcome, a diagnosis that can help the patient improve his/her current condition.
Currently our reasoning engine can handle more than 1000 diseases and 1000 rare diseases. Recently, we have been able to introduce a new feature that enables the analysis of the side effects of a drug one sued for a given disease.
We have set up the framework for a “democratization of medicine”, as envisioned by Dr Eric Topol. Democratization of medicine means medicine for all, whenever and wherever they are.
In a globalized world, where the aging population needs more and more assistance and the poor countries struggle to set up healthcare systems, we can provide viable solutions."

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

Started in July 2015, Medx2020 is a phase 2 year SME instrument project funded by the EU within the H2020 framework for research and innovation.
Originally thought to develop a unique visual reasoning tool for medical professionals that offers diagnosis decision support at the point of care, Medx2020 is now in its midterm phase. After 1 year of implementation Medx2020 matched its original purpose and went beyond!
“Ada”, our intelligent and caring chatbot, is now online! Find it at www.ada.com
Ada is the result of an intense year of work, hand in hand with worldwide recognized experts on the field, as well as with relevant healthcare organizations and the patient, of a team of computers scientists and medical doctors of our team.
Ada is a new approach to healthcare.
Ada is more than just an app. Ada is a virtual health companion on the go and gives you the best symptom assessment available building up a detailed picture of the patient health over time, Ada provides the patient with an accurate understanding of what is going on and helps him decide on next steps to take.
Thanks to Ada, anyone can check symptoms wherever and whenever he/she wants. Just answering simple, personalised questions about his/her health, Ada asks the patient all that is needed to really get to the bottom of what is bothering him/her. If, needed, Ada can pre-briefed doctors and book a video consultation with a doctor of the patient choice, at a time that suits him/her.
Finally Ada make available your digital health record in the patient pocket, secure and always up to date.
Thanks to the Medx2020 Ada reasoning technologies can also provide disease monitoring and prevention through integration of different data sources such as sensors, lab tests and more.
Medical doctors and specialists can find Ada extremely helpful in managing their patient in due time, and handling difficult diagnosis. Doctors are often overloaded by the number of patients they follow, Ada is a convenient way to work from the comfort of their home or any location with a stable internet connection.
The reasoning technology behind Ada is being developed as well within the MeDX2020 project. Originated by a collaboration among the co-founders, the reasoning engine is being now refined, and is capable to include complex pathophysiological connections.
MeDX2020 reasoning technology is based on Bayesian neural network approach, enriched by unique features that enable what we call “constellatory thinking”. Our brain, largely works out of analytical logics, grasping the complexity of everyday life meaningful constellations. The patient condition, is such an example. Especially when dealing with difficult diagnosis, such as for rare or chronic diseases, the brain cannot handle the different symptoms one by one and get to a sound conclusion, there are too many. The difficulty of the doctor is to grasp this complexity, giving it a meaningful outcome, a diagnosis that can help the patient improve his/her current condition.
Currently our reasoning engine can handle more than 1000 diseases and 1000 rare diseases. Recently, we have been able to introduce a new feature that enables the analysis of the side effects of a drug one sued for a given disease.
We have set up the framework for a “democratization of medicine”, as envisioned by Dr Eric Topol. Democratization of medicine means medicine for all, whenever and wherever they are.
In a globalized world, where the aging population needs more and more assistance and the poor countries struggle to set up healthcare systems, we can provide viable solutions.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

"Misdiagnosis is has proved a shocking reality, the US have been among the first to recognize it.
"Getting the right diagnosis is a key aspect of health care: It provides an expla¬nation of a patient's health problem and informs subsequent health care decisions.
For decades, diagnostic errors — inaccurate or delayed diagno¬ses — have represented a blind spot in the delivery of quality health care. Diagnostic errors persist throughout all settings of care and continue to harm an unacceptable number of patients" [the IOM report]
Diagnostic errors are often incredibly harmful to patients as they may lead to delays in treatment, lack of treatment, inappropriate, or unnecessary treatment. This, in turn, can have physical, psychological, and financial consequences.
Causes are varied but include inadequate communication between physicians and patients, a health care system design that does not support the diagnostic process, limited feedback to clinicians about diagnostic performance, and a health care culture that discourages transparency, so diagnostic mistakes are typically not reported (and not learned from).
Dr. Peter Pronovost, director of the Armstrong Institute for Patient Safety and Quality at Johns Hopkins, told CNN:2
"It's probably one of the, if not the, most under-recognized issues in patient safety… Much of the harm that we once labeled as inevitable we're now seeing as preventable".
But Europe is not lacking behind, and the leading healthcare organisations are making plans. Back to 2009, the Telegraph posted that one in six NHS patients are 'misdiagnosed' [telepgraph 2009], recognizing the problem in its worrying dimension.
During the years the problems didn’t soften, on the contrary, the more the awareness has been raised the more wider the problem revealed itself.
More recently, in 2014, a new study published in the latest issue of the UEG Journal revealed 1 in 10 Inflammatory Bowel Disease patients are misdiagnosed with Irritable Bowel Syndrome and in 3% of cases the misdiagnosis can persist for five or more years. Leading IBD researcher and United European Gastroenterology (UEG) spokesperson Dr Michael Scharl says, “IBS has been estimated to affect at least 10% of the population in Europe and it causes distressing symptoms that disrupt normal life.” He explains, “We have known for some time that there are similarities between symptoms of IBS and IBD, but when it comes to diagnosis and treatment, this differs greatly. [1]” The initial misdiagnosis of IBD has serious consequences for patients, in particular those with Crohn's disease, since delays in diagnosis is correlated with an increased risk of bowel stenosis and CD-related intestinal surgery.
Already in 2009, EU estimated 8 - 12% of patients admitted to hospital in the EU suffer from adverse events whilst receiving healthcare, for example:
• healthcare-associated infections (Accounting for approximately 25% of adverse events)
• medication-related errors
• surgical errors
• medical device failures
• errors in diagnosis
• failure to act on the results of tests.
The EU launched an initiative, spreading around all the member states, called Patient safety package http://ec.europa.eu/health/patient_safety/policy/index_en.htm ]. Patient safety package highlights how the Commission and EU countries are addressing the challenge of patient safety.
Despite of all efforts, a lot remain to be done and the problem is far from being solved.
In fact, it not only a matter of the difficulty to dealing with the complexity of many diseases by the doctor (namely chronic and rare diseases) , but also a matter of the relationship between doctors and patients. Doctors 'more likely to misdiagnose patients' if they are 'difficult' says the Independent [2]. A new study finds patients who obstruct clinicians' efforts by being aggressive increase chance of misdiagnosis by 42%
To effectively tackle this problem, all the different facets must be taken into consideration.
First of all, our brain does not work in an analytical way when addressing complexity. Dating back to the Gestalt psychologists (Duncker, 1945; Katona, 1940; Koffka, 1935; Metzger, 1986; Wertheimer, 1959) becoming famous for their work on the perceptual domain and founding the Gestalt principles, they realized that perceptual processes do more than processing pieces of information, but integrating the information in a meaningful way. Max Wertheimer (1959) claimed that this holds not only true for the domain of perception but also for the faculty of thinking. He showed that for difficult problems it was often helpful to reorganize the given information in way that new constellations of information resulted, that often provide an elegant solution to a problem.
In several studies, it was shown (see Öllinger & Knoblich, 2009 for an overview) that the relaxation of given self-imposed constraints often results in new constellations of information that provide new meanings and facilitate the solution of difficult and complex problems. Dual process models are particularly interested in the interplay of the constellatory mode of thinking that integrates the given information in a meaningful way and analytical processes that use details of the given information and consciously processing the information according to law-like principles (Evans, 2003; Gigerenzer & Todd, 2001; Kahneman, 2012).
Based on these recent findings, A.v. Mueller and Martin Hirsch (MeDX Project Manager) created now a conceptual and technological approach to implement the interplay between analytical and Gestalt-like reasoning processes on the technological level, and, at the same time, foster and support both thinking modes in the physician with a new form of user-interface.
This is the how MeDX was born. Today, after one year of project implementation, our cahtbot “Ada” can talk with the patient and be an interface to our reasoning engine, providing both doctor and patient with reliable insights, and at the end diagnosis.
After one year of intense work at MeDX gmbh, a team of 70 medical experts and computers scientists enabled the intelligent and smart chatbot capable of complex reasoning.
Ada is able to provide a response to the different facets of the misdiagnosis problem:
1 – support the doctor when dealing with complexity
2 – support the patient in his/her relationship with the doctor
Providing a realible decision support tool for medical experts, companies, insurers, etc.
Whether a BtB or BTC, “Ada” bear the potential to revolutionize the healthcare system making the democratization of medicine a real fact.
You can meet Ada at www.ada.com
You can learn more about Medx2020 project at: www.medx2020.eu

[1] Timothy R Card, Jesse Siffledeen and Kate M Fleming. Are IBD patients more likely to have a prior diagnosis of irritable bowel syndrome? Report of a case-control study in the General Practice Research Database. 2: 505-512, UEG Journal December 2014. Available at: http://ueg.sagepub.com/content/2/6/505.full
[2] http://www.independent.co.uk/life-style/health-and-families/health-news/doctors-more-likely-to-misdiagnose-patients-if-they-are-difficult-a6930781.html";

Related information

Record Number: 192891 / Last updated on: 2016-12-15
Follow us on: RSS Facebook Twitter YouTube Managed by the EU Publications Office Top