Doctors of today are often inundated with signals from charts, test results, and other metrics to keep track of. It can be hard to integrate and monitor all of these data for multiple patients while making real-time treatment decisions, especially when data is documented inconsistently across hospitals.
In a new bunch of papers, “researchers from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) explore ways for computers to help doctors make better medical decisions.
One team built a machine-learning approach called ‘ICU Intervene” which takes large amounts of intensive-care-unit data, from vitals and labs to notes and demographics, to determine what form of treatments are needed for different symptoms. The system utilises “deep learning” to make real-time predictions, learning from past ICU cases to make suggestions for critical care, while also explaining the reasoning behind these decisions.
“The system could potentially be an aid for doctors in the ICU, which is a high-stress, high-demand environment,” says PhD student Harini Suresh, lead author on the paper about ICU Intervene. “The goal is to leverage data from medical records to improve health care and predict actionable interventions.”
Another development made by another team is a new approach called “EHR Model Transfer” which can assist the application of predictive models on an electronic health record system (EHR) system, despite being trained on data for a different EHR system. Concentrating on this approach, the team showed that the predictive models for mortality and prolonged length of stay can be trained on one EHR system and used to make predictions on a different system.
Both models were trained utilising data from the critical care database MIMIC, which includes de-identified data from roughly 40,000 critical care patients and was developed by the MIT Lab for Computational Physiology.
Integrated ICU data is vital which helps in automating the process of predicting patients’ health outcomes.
“Much of the previous work in clinical decision-making has focused on outcomes such as mortality (likelihood of death), while this work predicts actionable treatments,” Suresh says. “In addition, the system is able to use a single model to predict many outcomes.”
The ICU data plays an important role and it depends on how the data is leveraged with respect to its storage and what happens when that storage method gets changed. And this is where the EHR Model Transfer comes in. This approach can work across different versions of EHR platforms, using natural language processing to identify clinical concepts which are encoded differently across systems and then mapping them to a common set of clinical concepts (such as “blood pressure” and “heart rate”).
