Text Message System for the Prediction of Colonoscopy Bowel Preparation Adequacy Before Colonoscopy: An Artificial Intelligence Image Classification Algorithm Based on Images of Stool Output

image

Inadequate bowel preparation frequently serves as a barrier in effectively screening for colorectal cancer. Experts within Penn Medicine and the CCEB formed a team to develop an artificial intelligence (AI) machine-learning algorithm to detect if colonoscopy bowel preparation was complete by analyzing photos of stool sent via text messages from patients. The machine-learning algorithm was capable of distinguishing between inadequate/fair preparation and good preparation, as well as normal colonoscopy duration and long colonoscopy duration. Read more.

Figure 1 Figure 3
Development of artificial intelligence algorithm.
Flowchart of patients who submitted photos of stool output out of total patients scheduled for colonoscopy.

 

Authors

Chethan Ramprasad, Divya Saini, Henry Del Carmen, Lev Krasnovsky, Rajat Chandra, Ryan Mcgregor, Russell T Shinohara, Eric Eaton, Meghna Gummadi, Shivan Mehta, James D Lewis