Celiac Disease, Non-celiac Enteropathies, Small Bowel Mucosal Atrophy or Lesions
Conditions
Keywords
Artificial Intelligence, Endoscopy, Deep Learning, Computer-Aided Diagnosis, Small Bowel
Brief summary
Coeliac disease (CD) is an immune-mediated enteropathy leading to small intestinal mucosal atrophy. Diagnosis relies on serology and duodenal biopsies, but it can be complicated by patchy lesions and differential diagnosis with Non-Celiac Enteropathies (NCEs). This multicenter observational study aims to develop and validate an Artificial Intelligence (AI) system to detect and characterize small bowel mucosal atrophy and other pathological findings using endoscopic imaging. The study involves a retrospective phase for training the AI model and a prospective phase to validate its diagnostic accuracy compared to standard human assessment.
Detailed description
The study is a multicenter observational non-profit study with a total expected duration of 36 months. It aims to address the challenges in diagnosing CD and NCEs, specifically the subjective nature of endoscopic evaluation and inter-observer variability. The study proceeds in two phases: 1. Model Development (Retrospective): Training of Deep Learning algorithms on anonymized endoscopic images/videos to identify mucosal atrophy and other lesions (e.g., angiodysplasia, ulcers, polyps). 2. Validation (Prospective): Enrolling patients undergoing small bowel endoscopy to validate the AI system's performance. The system aims to provide analysis to assist endoscopists, reducing missed lesions and improving diagnostic accuracy. Validation of the AI system will be performed offline on recorded anonymized endoscopy videos collected prospectively during the validation phase.
Interventions
None listed
Sponsors
Istituti Clinici Scientifici Maugeri SpA
Fondazione IRCCS Policlinico San Matteo, Pavia, Italy
Ospedale Valduce, Como
Azienda Ospedale Università Padova, Padova
IRCCS Azienda Ospedaliero-Universitaria di Bologna
ASST Cremona, Cremona
Azienda Ospedaliero-Universitaria Città della Salute e della Scienza di Torino
Fondazione Poliambulanza Istituto Ospedaliero, Brescia
ASST degli Spedali Civili di Brescia , Brescia
IRCCS Policlinico San Donato, Milano
Azienda Socio-Sanitaria Territoriale (ASST) di Pavia
Azienda Unità Sanitaria Locale della Romagna, Ravenna
ASST Lecco, Lecco
Fondazione IRCCS Ca' Granda Ospedale Maggiore Policinico, Milano
Academic Unit of Gastroenterology, Sheffield Teaching Hospitals, Sheffield, UK
Study design
Observational model
COHORT
Time perspective
OTHER
Eligibility
Sex/Gender
ALL
Age
18 Years to No maximum
Healthy volunteers
No
Inclusion criteria
* Adult patients (age ≥18 years). * Patients undergoing endoscopic investigation of the small bowel (endoscopy or capsule endoscopy)
Exclusion criteria
* Inability to provide informed consent.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Diagnostic Performance of the AI System | Through study completion (36 months) | Evaluation of Accuracy, Sensitivity, and Specificity of the AI system in detecting intestinal mucosal atrophy and other major pathological findings compared to the gold standard (histology/expert consensus). |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Comparison of Diagnostic Performance (No AI-assistance vs AI-assisted) | Through study completion (36 months) | Comparison of accuracy, sensitivity, and specificity of gastroenterologists performing assessment with and without the assistance of the AI system. |
| Inter-observer Agreement | Through study completion (36 months) | Measurement of agreement between the AI system and expert gastroenterologists in evaluating the extent of mucosal atrophy. |
Countries
Italy
Contacts
CONTACTFederico Biagi, MD
Outcome results
None listed