Competition Held to Develop Predictive Technology for Deterioration Rate of Coronavirus Patients
Over the last three months, a competition was held to develop the best technology for predicting the deterioration rate of patients with COVID-19, and Intel Advanced Analytics was the winner.
Sheba Medical Center led the national fight in Israel against COVID-19. Against this backdrop, the CORONA HUB was established at ARC (Accelerate, Redesign, Collaborate) as an innovation center focused on finding technological models to help predict deterioration from the coronavirus. ARC, in collaboration with TAKEDA Israel, initiated the recent contest.
“Competition is another pillar of ARC’s efforts to promote the discovery of innovative solutions to help manage the novel coronavirus and save lives,” said Dr. Eyal Zimlichman, VP of Medicine and Innovation.
“We do not see this type of competition as a competition for its own sake. Rather, it is an opportunity to share information and connect communities in the world of medicine; it promotes innovation for research and advancement to find solutions that will empower the medical world,” he added.
The Intel Advanced Analytics Team designed a system for assessing the risk of each COVID-19 patient, starting from the first admission. The analytic models that the team developed work to predict whether the patient will experience respiratory deterioration and require ventilation and whether intensive care or extended hospitalization will be needed. Additionally, the results are visually accessible to doctors, displaying “heat maps” and decision trees that clearly show the complex considerations taken by the models.
“COVID-19 is a new disease, and we don’t know a lot yet about how it behaves. Three months ago, we knew even less – but then we realized the key to understanding the course of the disease better is artificial intelligence and data compiled from many COVID-19 patients,” explained Zimlichman. “The idea for the competition was borne from this understanding, which aims to encourage relevant researchers and scientists to find innovative solutions to help medical teams better manage the virus and save lives.”
“One of the major difficulties faced by the medical teams is the inability to predict how coronavirus will behave. It’s still a mystery. The more patients we meet, the more we learn about the behavior of the virus, but we still don’t know the exact mechanism of action,” said Prof. Galia Rahav, head of the Infectious Disease Unit at Sheba. “This type of contest, initiated by the ARC, helps to identify models that can be a therapeutic tool for medical teams, helping to predict the deterioration rate of COVID-19 patients and leading to life-saving healthcare.”
More than 30 large organizations and over 100 researchers participated in the competition. Throughout the months that it was ongoing, anonymous data was provided from COVID-19 patients at the medical center, which spanned the entire hospitalization period and longer. Using this data, researchers built models to predict the course of the disease. In particular, the predictions indicated which patients were most likely to deteriorate and who will need ventilation.
The competing companies were assisted by mentors from the ARC and medical specialists from Sheba, including Prof. Robert Klempner, Prof. Galia Rahav, Dr. Gadi Segal, Dr. Yael Habib, Dr. Itzik Levy, Dr. Sharon Amit, Dr. Itai Pesach and other top experts.
There were two rounds of judging, carried out by more than ten doctors and data specialists from Israel and overseas, including from Mt. Sinai Hospital in New York, Massachusetts General in Boston, Ottawa, Canada and more.
The winning project was implemented at Sheba and validated at Mt. Sinai.
“Sheba Medical Center leads the national effort to fight the virus and initiating competition is an integral part of that effort. It is essential for us to use the capabilities and data we have to empower our research and advance technological solutions that will revolutionize dealing with COVID-19 in particular – and change the face of medicine in general,” said Zimlichman.