Evaluation of 'WAYMED Endo CL CS' for Automated Detection and Diagnosis of Colorectal Adenoma and Non-Adenoma Lesions

The Effectiveness of the Computer-aided Detection/Diagnosis Software, 'WAYMED Endo CL CS', Which Assists Medical Specialists by Automatically Analyzing Colorectal Endoscopic Images With Confirmed Adenoma or Non-Adenoma Lesions, Detecting the Lesion, and Providing Probabilities as to Whether the Lesion Belongs to the Adenoma or Non-Adenoma Group: A Single-center, Single Arm, Retrospective, Superiority, Single-blind, and Pivotal Clinical Trial

The purpose of this clinical trial is to evaluate the effectiveness of the computer-aided diagnosis (CADx) software, 'WAYMED endo CL CS', which assists medical specialists by automatically analyzing colorectal endoscopic images, identifying lesions, and providing probability values to classify them as Adenoma or Non-Adenoma. This pivotal trial is designed to confirm that 'WAYMED endo CL CS' can support clinicians in diagnostic decision-making by improving the classification of colorectal lesions.

This clinical trial aims to assess the clinical sensitivity and specificity of 'WAYMED endo CL CS' in detecting colorectal lesions and classifying them into Adenoma (colorectal cancer, adenoma) or Non-Adenoma(Hyperplastics, etc) groups. The trial is designed as a retrospective, single-center, single-arm, single-blind, superiority, and pivotal study.

Colorectal endoscopic images and corresponding histopathology results are retrospectively collected from adult patients. Images meeting all inclusion/exclusion criteria are enrolled. Each lesion is annotated and verified by a Reference Standard Establishment Committee, which consists of experienced endoscopy specialists. The investigational software is applied to the enrolled images to automatically detect and classify the lesions, while the reference standard serves as the comparator.

The primary endpoints include clinical sensitivity (%) and specificity (%) of 'WAYMED endo CL CS' in distinguishing Adenoma from Non-Adenoma lesions. The secondary endpoint is the overall diagnostic accuracy (%).

This study is intended to provide confirmatory clinical evidence required for regulatory approval of 'WAYMED endo CL CS' as a computer-aided diagnosis (CADx) software for colorectal endoscopy.

Inclusion Criteria:

1. Adult patients (≥19 years) with colorectal endoscopic images showing one lesion.
2. Histopathology confirming the lesion as Adenoma (colorectal cancer, adenoma) or Non-Adenoma (sessile serrated lesion, hyperplastic polyp).

Exclusion Criteria:

1. Images previously used for training or internal validation of the investigational software.
2. History of colectomy.
3. Diagnosed with inflammatory bowel disease (e.g., ulcerative colitis, Crohn's disease) or neuroendocrine tumor.
4. Poor image quality (blurred, incomplete lesion capture).
5. Images with multiple lesions or fewer than two images per lesion.
6. Determined by the investigator to be inappropriate for inclusion.