CheXaid: deep learning assistance for physician diagnosis of tuberculosis using chest x-rays in patients with HIV
Problem
TB in HIV-positive patients often goes undiagnosed, and chest x-ray diagnosis is difficult due to atypical presentation and radiologist shortages.
Why it matters
TB is the leading cause of preventable death in HIV-positive patients.
What the AI did
A deep learning algorithm trained on chest x-rays and clinical data from 677 HIV-positive patients was deployed as a web-based diagnostic assistant for clinicians.
Results
Clinician mean accuracy improved from 0.60 (95% CI 0.57-0.63) to 0.65 (95% CI 0.60-0.70), p=0.002.
Impact
AI assistance produced a modest but statistically significant improvement in clinician TB diagnosis accuracy.
Record
| Venue | npj Digital Medicine |
|---|---|
| Published | 2020 |
| Work type | article |
| Authors | Pranav Rajpurkar, Chloe P. O’Connell, Amit Schechter, Nishit Asnani, Jason J. Li, Amirhossein Kiani, Robyn L. Ball, Marc Mendelson et al. |
| Citations | 136 (OpenAlex, at import time) |
| References | 40 |
| Model or method | CheXaid deep learning algorithm |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1038/s41746-020-00322-2 |
| Open access | gold |
| Keywords | tuberculosis, HIV, chest x-ray, deep learning, clinical decision support |
Importance rating
major — millions of lives, an industry or a field
Abstract states TB is the leading cause of preventable death in HIV-positive patients, a major global health burden; AI assistance improved clinician diagnosis accuracy in high co-infection settings.