Artificial Intelligence in Colonoscopy

A Randomized Trial on the Utilization of Endocuff-assisted Colonoscopy and Computer-aided Detection in Optimizing Colonoscopies in the Elderly

N = 264 patients (50% female) aged 75 years and above undergoing colonoscopy were enrolled. Patients were randomly assigned into one of the three intervention groups: the primary intervention arm (CADe in combination with the MED), the second group with MED alone, and the control group with WLE. All detected lesions were removed and sent to histopathology for diagnosis. The primary outcome was the adenoma detection rate. Secondary outcomes were adenoma detection in the left colon in our cohort of patients.

264 patients, with an equal gender distribution (50% female), aged 75 years and above, undergoing screening and diagnostic colonoscopy at The Surgery and Endoscopy Center of Sebring, Sebring, Florida, were enrolled in this study. The eligibility criteria for a randomized controlled trial (RCT) comparing AI and mucosal exposure devices together, mucosal exposure devices alone, and white light endoscopy in patients 75 years and older could be structured as follows.

Patients were randomly allocated into one of three study groups: the primary intervention arm, where colonoscopy was performed using the CADe in combination with the MED; the second group underwent colonoscopy solely with the MED, while the control group underwent colonoscopy solely with the WLE. We used a Convolutional Neural Network-based CADe system, GI Genius, acquired for licensed use from Medtronic Inc., Minneapolis, MN. The MED employed was the EndoCuff Vision (ECV) developed by Olympus America, Center Valley, PA, which constitutes part of the standard equipment available. All detected lesions were identified and excised throughout the colonoscopy procedures, and specimens were promptly sent for histopathological analysis.

The primary outcome of interest was adenoma detection rate (ADR), defined as the percentage of patients in whom at least one histologically proven adenoma or carcinoma was identified during colonoscopy. Secondary outcomes included ADR in the left colon in our cohort of patients.

This study was conducted according to accepted ethical principles and approved by the institutional review board (IRB) of "The Surgery and Endoscopy Center of Sebring." Informed consent was obtained from all participants before enrollment, and measures were taken to ensure patient confidentiality and data protection throughout the study period.

Inclusion Criteria

1. Patients aged 75 years and older.
2. Patients who were able to provide informed consent or had a legally authorized representative who can consent on their behalf.
3. Patients who were deemed fit for colonoscopy based on a pre-procedure evaluation.

Exclusion criteria

1. Patients with acute gastrointestinal conditions such as active colitis, acute diverticulitis, or bowel obstruction.
2. Patients with a history of major colorectal surgery that might alter normal colon anatomy (e.g., colectomy).
3. Patients with severe comorbid conditions that would contraindicate colonoscopy, such as severe cardiopulmonary disease or advanced liver disease.
4. Patients with uncorrected coagulopathies or those on anticoagulation therapy that cannot be safely managed around the time of the procedure.
5. Patients who are unable to adequately prepare the bowel for colonoscopy.
6. Patients who refuse to participate in the study or have a legally authorized representative who refuses consent.