Mar. 28, 2022

BGU MobiGuide project chosen as an EU Big Trends Research Project

 

The article​​​ about the MobiGuide project has appeared as part of the editors' choice for Big Trends in Europe

 

The prestigious scientific journal “Communications of the Association of Computing Machinery (CACM)" has selected, first as part of a dedicated international workshop, and then as part of the CACM special issue of April 2022, a series of ground breaking research projects in Europe, whose leaders were invited to submit a short article describing the selected project, when was then refereed again.

 

One of the projects featured in the Big Trends in Europe special issue was the MobiGuide project, whose focus is on the remote management at home of chronic patients, using sensors on the patients, smartphones, and a computer network, mostly without any human intervention, which was initiated by Prof. Yuval Shahar, M.D., Ph.D., head of the BGU Medical Informatics Research Center, from the Software and Information Systems Engineering Department, whose laboratory, which specializes in Artificial Intelligence in Medicine, had provided the MobiGuide project its core computational tools – medical knowledge engineering, longitudinal patient monitoring, and personally customized evidence-based decision support to the patients and to their care providers.

 

The project's coordinator was Prof. Mor Peleg from Haifa University's Information Systems Department.

The decision-support interfaces for the clinicians as well as the tools for eliciting the patients' preferences were designed and developed by Prof. Silvana Quaglini and her team at the University of Pavia, Italy.

These three collaborators are co-authors of the CACM article.

 

Overall, the MobiGuide project include 13 partners from five countries (Israel, Italy, Spain, the Netherlands, and Austria). Its four-years budget was €7M. Most of the funding was provided by the EU Community's 7th Framework initiative.

 

In the MobiGuide project, chronic patients were managed at home through sensors (blood pressure cuff, and EKG belt, a blood-glucose measurement sensor) that communicated their data via Bluetooth to the patient's smartphone device. From there, the data were transmitted, via local EU servers, to Prof. Shahar's BGU laboratory, for a context-sensitive interpretation of the multivariate time-oriented data, and for guideline-based decision making regarding a possible alert to the patient (through a dedicated application in the patient's smartphone) or a possible recommendation to the patient's care provider (through a Web-based interface used by the clinicians in their respective clinics to monitor the patients).

 

A clinical pilot in the 4th year of the project included a group of pregnant women in Barcelona, Spain, who were diagnosed with Gestational Diabetes, and concurrently, a group of cardiac arrhythmia patients in Pavia, Italy, who were diagnosed with Atrial Fibrillation.  The pilot had demonstrated that the MobiGuide system is highly feasible; the patients and the care providers were highly compliant to the system's alerts' reminders' and specific recommendations; multiple clinic visits were avoided; and on several occasions, the computer-based system had identified that the patient was misdiagnosed (e.g., as having the wrong cardiac arrhythmia type), or had recommended treatment that was different from the one originally prescribed (e.g., by suggesting the insulin be started earlier). In all cases, the clinicians had agreed with the MobiGuide system's diagnosis or therapy suggestions when the patient was referred by the system to the gestational diabetes or the cardiology clinic.

 

The MobiGuide project was originally proposed by Prof. Shahar as a cheap, high-quality, option for continuous management of the multiple chronic patients in the increasingly older world's population in general, and the EU population in particular. However, during the COVID-19 pandemic, the project turned out to have long-ranging implications for remote, home-based management of patients in general.