Classification of Paediatric Inflammatory Bowel Disease using Machine Learning
Problem
Accurate diagnosis of paediatric inflammatory bowel disease subtypes (Crohn's, ulcerative colitis, IBDU) is needed for prompt, effective treatment.
Why it matters
Misdiagnosis of PIBD subtype can delay appropriate treatment in children with a complex, rising-incidence condition.
What the AI did
Machine learning models were developed, trained, tested and validated on endoscopic and histological data from 287 children to classify disease subtype; hierarchical clustering also identified four novel subgroups.
Results
Classification accuracy of 71.0% (endoscopic only), 76.9% (histological only) and 82.7% (combined); four novel subgroups identified.
Impact
ML classified PIBD subtypes with 82.7% accuracy using combined data, aiding diagnosis.
Record
| Venue | Scientific Reports |
|---|---|
| Published | 2017 |
| Work type | article |
| Authors | Enrico Mossotto, James John Ashton, Tracy A F Coelho, R Mark Beattie, Ben D. MacArthur, Sarah A. Ennis |
| Citations | 157 (OpenAlex, at import time) |
| References | 38 |
| Model or method | not stated in source |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1038/s41598-017-02606-2 |
| Open access | gold |
| Keywords | machine learning, paediatric IBD, classification, endoscopy, histology |
Importance rating
real but narrower advance
Abstract states PIBD is complex with increasing incidence and that accurate diagnosis is needed for prompt effective treatment; a real but narrower diagnostic advance.