Biosensors and Machine Learning Algorithms to Detect Complications in Post-Pancreatectomy Patients
The Opportunity
Surgical complications can be devastating to patients, as well as a significant financial burden to healthcare systems. Pancreatectomy is an especially complex operation with post-operative complications occurring in 40-60% of cases. The goal of this project is to leverage remote telemonitoring devices in conjunction with machine learning to address poor surgical outcomes in patients undergoing pancreatectomy.
Our Approach
Dr. Hammill and his team used wearable activity monitors and gamification methods to gather data and correlate surgical outcomes with patient preoperative activity levels, improved patient progressive mobility during the inpatient postoperative period, and evaluated machine learning models by incorporating activity monitor data to predict early onset of surgical complications.
The Solution
There are 103 patients who have successfully been enrolled into the clinical trial for this project. This trial is now being expanded to pancreatic cancer patients receiving chemotherapy, hernia repair patients, and hepatectomy patients.
