Machine Learning-Based Risk Stratification for Fistula Formation After Perianal Abscess Drainage

A Prospective Cohort Study for Machine Learning-Based Prediction of Anal Fistula Formation After Perianal Abscess Drainage Based on Drainage Setting, Provider Experience, and MRI Interpretation (PRISM)

Status

Not yet recruiting

Phases

Unknown

Study type

Observational

Source

ClinicalTrials.gov

Registry ID

NCT07019532

Acronym

PRISM

Enrollment

450

Registered

2025-06-13

Start date

2025-07-01

Completion date

2026-06-01

Last updated

2025-06-13

For informational purposes only — not medical advice. Sourced from public registries and may not reflect the latest updates. Terms

Conditions

Anal Fistula, Perianal Abscess

Keywords

Artificial Intelligence, MR Imaging, Colorectal Surgery, Surgical Outcomes, Drainage, Fistula Prediction

Brief summary

This prospective cohort study investigates the influence of provider experience and drainage location on fistula formation within 6 months following perianal abscess drainage. Additionally, the study explores the role of artificial intelligence (AI)-based interpretation of magnetic resonance (MR) images in early identification of fistula development.

Detailed description

Perianal abscess drainage is a common surgical procedure. However, subsequent fistula formation remains a significant complication. This study aims to determine whether the procedure setting (operating room, emergency department, or outpatient clinic) and the experience level of the performing clinician affect fistula development rates. Furthermore, the study evaluates the use of AI-assisted analysis of selected MR images to identify early signs of fistula formation. Selected image slices will be labeled based on radiological reports, and a machine learning model will be trained to predict fistula risk. The study will also compare AI-generated interpretations with expert radiologist assessments to validate performance. The ultimate goal is to create a risk stratification tool to support clinical decision-making in surgical management of perianal abscesses.

Interventions

None listed

Study design

Observational model

COHORT

Time perspective

PROSPECTIVE

Eligibility

Sex/Gender

ALL

Age

18 Years to No maximum

Healthy volunteers

Inclusion criteria

* Age ≥ 18 * First-time perianal abscess * Surgical drainage performed

Exclusion criteria

* Existing anal fistula history * Crohn's disease * Immunosuppressive treatment * Incomplete 6-month follow-up

Design outcomes

Primary

Measure	Time frame	Description
Fistula formation within 6 months	6 months	Confirmed by clinical exam, surgical findings, or MR imaging

Secondary

Measure	Time frame
Correlation between drainage location and fistula rate	6 months
Correlation between provider experience and fistula complexity	6 months
Diagnostic accuracy of AI-based MR analysis vs radiologist	6 months

Contacts

Primary ContactKayahan Eyüboğlu, MD

kayahaneyuboglu@gmail.com+905546813327

Outcome results