Fighting COVID-19 with AI: The prize winners of the fifth Future Medicine Science Match

At the Future Medicine Science Match 2020, the fifth edition of the event organized by the Berlin Institute of Health (BIH) and the Tagesspiegel newspaper, nearly 1,000 participants met online this year to learn about the best projects involving artificial intelligence (AI) in translational medicine. The award for best scientific pitch went to app@work, a project that uses automated data analysis with integrated AI to predict the course of disease and help identify targeted treatment options for COVID-19 patients in intensive care units. The audience award for most entertaining pitch, which is funded by Intuitive Surgical, was won by Manouchehr Shamsrizi from gamelab.berlin – part of Humboldt-Universität zu Berlin’s Cluster of Excellence. Both winning teams received a cash prize of €500.

Anesthesiologists must document in detail everything they do before, during and after an operation. This includes all medication they administer, as well as the patient’s vital signs such as blood pressure, pulse and body temperature. Usually, these notes are handwritten on paper. The data must then be carefully stored for at least 30 years – and yet it is not made available for use by doctors or even for research purposes.

Sandman.MD automatically stores ICU data

“So, several years ago, we developed the Sandman.MD tool, which automatically transfers all data from the operating and recovery rooms to an iPad for storage,” explains Jörg Wegner from app@work GmbH in Berlin. “With a small adapter, the anesthetist can automatically transfer data from the ventilator and patient monitor to the iPad, and then only needs to add additional information such as the manually administered medication.” Sandman.MD has been in use at many hospitals for years already, including Frankfurt University Hospital.

Professor Kai Zacharowski, Director of the Clinic for Anaesthesia, Surgical Intensive Care, Emergency Medicine and Pain Therapy in Frankfurt, has already documented and stored data from well over 60,000 operations on 150 iPads. “With this data, we applied for a European Union Horizon 2020 research project to help improve the treatment of COVID-19 patients,” says Zacharowski. After all, COVID-19 patients are also treated in intensive care units. Based on their experience with Sandman.MD, the scientists and IT experts led by Zacharowski and Wegner from app@work are developing a tool together with other partners that will monitor COVID-19 patients in the ICU around the clock.

Fourteen participating European clinics

Every critically ill COVID-19 patient is linked up to an iPad that records all his or her data around the clock. This data is transferred to a central online database, where the entirety of COVID-19 data is collated and evaluated using artificial intelligence. The AI and machine-learning programs then use this information to develop predictive models that allow for patient-specific predictions to be made about how the disease will develop, and what the best personalized therapy is for the patient. The database then transfers the results back to the iPad, where the attending physician can directly incorporate the results into the treatment plan. The project is called “ENVISION,” and will start in December with the participation of 14 hospitals from eleven European countries. It is being carried out under the auspices of the European Society of Anaesthesiology and Intensive Care, where Kai Zacharowski is president. Industry partners are also on board.

Professor Sylvia Thun, Director of eHealth and Interoperability at the BIH, Scientific Director of this year’s Future Medicine Science Match, and chair of the jury, praised the pioneering spirit of the prize winner’s project: “Every single COVID-19 data record improves and enlarges the database. The artificial intelligence learns something new every time and the attending physician receives reliable support in making a decision. With the help of AI, we will be able to better control the pandemic.”

Gaining medical knowledge from video games

The audience award for most entertaining pitch went to Manouchehr Shamsrizi, co-founder of both gamelab.berlin, part of Humboldt-Universität zu Berlin’s “Image Knowledge Gestaltung” Cluster of Excellence, and social start-up RetroBrain. In his presentation, he explained the possibilities that video game technologies open up for data-driven medicine. The audience award was presented by Joachim Haes, Director of Government Affairs at Intuitive Surgical, a manufacturer of robot-assisted surgical systems and the funder of the audience award.