Grant: Project Grant

Acute appendicitis is a very common surgical condition in children. While treatment for simple appendicitis is straightforward, once the appendix bursts there are several ways to manage the problem, and the hospital stay can be long and complicated. The burden on the child, family, and the hospital depends on how severe the condition is found during the surgery. Being able to identify the grade of perforation before the surgery - right in the emergency room - would be very useful, helping families in the counseling process, the surgical team in choosing the best way to manage the child, and the hospital in allocating resources for the patient. In order to achieve this, we plan to use artificial intelligence (AI) to identify the grading of the appendicitis based on child's symptoms, blood tests in the emergency room, and ultrasound examination before surgery. We will collect data from children who have been operated for appendicitis at the Montreal Children's Hospital over the past decade. We will then use this data to create several AI algorithms, by first training them on part of the data, then testing them on the remaining data. We will then compare the different algorithms, to find the one which best predicts the perforation severity in children with appendicitis. Finally, we will use the winning algorithm on a new group of children presenting with acute appendicitis, and evaluate the accuracy and usefulness of the algorithm in their care. In keeping with the interdisciplinary nature of this study, our team includes pediatric surgeons, emergency pediatricians, AI engineers, and patient collaborators. The study fits in and is supported through McGill's CommiSur Lab, which explores the connections between clinical communication and technology. The resulting diagnostic tool will be used to guide the future management of perforated appendicitis in children and the discussions between patients, families, and the surgical team.