Post written by Carlos Robles-Medranda, MD, FASGE, AGAF, from the Endoscopy Division, Instituto Ecuatoriano de Enfermedades Digestivas (IECED), Guayaquil, Ecuador.

We describe a case of a patient cared for in a medical facility abroad (Colombia) who presented with jaundice and proximal common bile duct stenosis detected with MRCP. EUS identified a homogeneous hypoechoic lesion localized at the emergence of the cystic duct. EUS fine-needle biopsy revealed normal histopathology. ERCP with brush cytology depicted a benign lesion.

The team in Colombia performed a digital single-operator cholangioscopy (DSOC) and classified it as a nonneoplastic inflammatory lesion using the Carlos Robles-Medranda DSOC criteria. Tissue sampling results were inconclusive.

Because of uncertainty among the visual impression and pathology results, the Instituto Ecuatoriano de Enfermedades Digestivas was remotely consulted for a second opinion for visual evaluation of the recorded DSOC video by a DSOC expert and for assessment by a cholangioscopy-based artificial intelligence (AI) system (AIWorks-Cholangioscopy; mdconsgroup, Guayaquil, Ecuador). Visual interpretation per the expert showed a type I/IV neoplastic lesion. The AI system highlighted areas of neoplasia, and the complete detection video was shared with the crew in Colombia.

A second DSOC was performed in Colombia using the detection video as a guide. New biopsies were taken near the highlighted areas, revealing cholangiocarcinoma, thus changing management of the patient.

We feel that physicians should make the most of technological improvements and intelligence systems that may aid during complex procedures. DSOC is an advanced technique that allows direct visualization of bile duct lesions. Several classifications help operators identify neoplasia within the biliary tract.

However, despite the availability of different classifications, interobserver mismatches occur. The application of an accurate AI model that can be used during real-time live cases and previously recorded videos, such as in this case, can provide a truthful second opinion and change management of the patient accordingly. AI tools may assist in targeted tissue sampling and with diminishing the visual impression gap between experienced and less-experienced endoscopists.

The present and future of GI endoscopy is aimed toward intelligent endoscopic procedures. The AIWorks-Cholangioscopy system is a valuable tool that offers consultations and/or a second opinion on lesions with dubious results. Its implementation allows for prompt and adequate management of patients with biliary lesions and may reduce the number of interventions required and overall health-related costs.

Histopathology of the biopsy sample obtained from the area suggestive of neoplasia detected with the AIWorks-Cholangioscopy software (mdconsgroup, Guayaquil, Ecuador). A, H&E, orig. mag. ×4; B and C, H&E, orig. mag. ×200.

Read the full article online.

The information presented in Endoscopedia reflects the opinions of the authors and does not represent the position of the American Society for Gastrointestinal Endoscopy (ASGE). ASGE expressly disclaims any warranties or guarantees, expressed or implied, and is not liable for damages of any kind in connection with the material, information, or procedures set forth.