Scientists Machine Learning Algorithm to Predict “Emergent” Inflammation Pathway in the Blood

A University of Cincinnati (UC) machine-learning research team developed an algorithm that can accurately predict the kind and severity of “pink tide” in which patients with inflammatory bowel disease (IBD) experience intermittent small blood vessel obstruction. The world-first discovery could provide a new tool that monitoring physicians using imaging changes.

“The most meaningful tool to assess treatment for IBD is blood test results,” said Mark Dye, MD, co-principal investigator of the first study and professor of surgery at the UC Health Interventional Center. “Currently, patients with IBD undergoing treatment for bypass have a five-year median survival of 3.5 years. The delay in finding treatment for patients with IBD is the factor that contributes to delayed CT blood vessel clearance and prolonged risk of developing cardiovascular disease.”

IBD is a chronic disease of the digestive system that causes pain, diarrhea and abdominal pain. Inflammation from this disease is highly prevalent in the gastrointestinal tract and can be triggered by environmental stimuli including excessive food intake, inflammation at the intestinal barrier and inflammation of the extremities and joints, collectively called “pink tide.”

Current study results showed that, in patients with pink tide, 11 supported in each of the four syndromes assessed by the researchers could be assessed with an average of 15 scans per patient. After adjusting for other factors, the researchers found that their combined analysis of both pink tide and CT coronary angiogram (laparotomy) results in an objective biomarker to predict patient response to intracolonary bypass. This milestone discovery represents a new tool for monitoring disease progression and therapeutic responses. “It is unsurprising that we found biomarkers demonstrating a significantly higher predictive power than the current gold standard of CT angiogram than laparotomy,” Dye said.

The clinical impact of the findings is anticipated because “IBD patients can live long and fulfilling lives, and knowing how a patient’s prognosis will change as a result of the color of their test results, is a crucial tool for patients treated by trained physicians.”

Dye believes that the findings are of great significance when outside researchers or doctors involved in patient care decide to test for specific biomarkers to predict patient response to treatment.

“This discovery may also help physicians on the ground at any institutions that are considering the use of IBD imaging as part of their procedures,” he said. 😀