A team of Computer scientists from Duke University and Harvard University has collaborated with physicians from the Massachusetts General Hospital and the University of Wisconsin to develop a machine learning model that can predict which patients are at more risk of having destructive seizures after suffering a stroke or other brain injury.
The machine learning system developed by the team helps determine which patients should receive expensive continuous electroencephalography (cEEG) monitoring. Used nationwide, the authors claim that their model could help hospitals monitor nearly three times as many patients, saving many lives as well as USD 54 million each year. The paper detailing the methods behind the interpretable machine learning approach appeared in the Journal of Machine Learning Research.
When a brain aneurysm leads to a brain bleed, much of the damage is not done in just the first few hours; it accumulates with time as the patient experiences seizure. But because the patient’s condition does not allow them to show any outward signs of distress, the only way to tell they are having seizures is through an EEG. However, continuously monitoring a patient with this technology is expensive and requires highly trained physicians to interpret the readings.
Aaron Struck, assistant professor of neurology in the University of Wisconsin School of Medicine and Public Health, and Brandon Westover, director of the Critical Care EEG Monitoring Service at Massachusetts General Hospital, sought to optimize these limited resources. Through the help of colleagues in the Critical Care EEG Monitoring Research Consortium, they collected data on dozens of variables from nearly 5,500 patients and got to work.
While most machine learning models are a “black box” too complicated for a human to understand, interpretable machine learning models are restricted to reporting back in plain English. The team created a machine learning algorithm that produces simple models called scoring systems for other applications. The scoring systems are based on a sophisticated combination of optimization techniques called “cutting planes” and “branch and bound.”
For instance, say you were looking for the bottom point on a bowl-shaped graph. A traditional cutting plane method uses tangential lines to choose spots that quickly settle at its base like a snowboarder losing momentum in a half-pipe. But if this method is asked to find the lowest point that is also a whole number, which the unrestricted answer is not likely to be, it might continue its search between the vast amounts of nearly acceptable solutions indefinitely.
To address the issue, the team combined cutting plane optimization with another called branch and bound, which cuts out a large part of the search. The entire process then repeats until an optimal, interpretable answer is produced. Their method has already proven successful in creating screening tests for sleep apnoea, Alzheimer’s disease, and adult ADHD. They just had to refill it in EEG data.
The team explained that the machine learning tool took data from thousands of patients and produced a model called 2HELPS2B. The great thing about this model is that clinicians can memorize it just by knowing its name. It looks like something that doctors would come up with on their own, but it is a full-blown machine learning model based on data and statistics. The model has doctors give points to patients based on the patterns and spikes found in their EEGs. With a maximum tally of seven, the result provides a probability estimate of the patient having a seizure at each point interval ranging from less than five percent to more than 95 percent.
The researchers tested the model against a new set of 2,000 cases and found that it worked well. The model was then put into service at the University of Wisconsin and Massachusetts General Hospital, allowing doctors only to use cEEG where is needed the most. After a year of use, the model resulted in a 63.6 percent reduction in the duration of cEEG monitoring per patient, allowing nearly three times as many patients to be monitored while generating a combined cost savings of USD 6.1 million. The model is now being used at four more hospitals nationwide were to adopt it; the researchers calculate they could save a collective USD 54 million each year.