Determination of disease severity in COVID-19 patients using deep learning in chest X-ray images
Problem
During the acute COVID-19 outbreak, clinicians needed to determine disease severity in patients using chest X-rays and to test whether deep learning could support this.
Why it matters
Rapid severity assessment from widely available chest X-rays could guide triage and management of COVID-19 patients.
What the AI did
A deep learning model was applied to chest X-rays from 48 RT-PCR-positive patients to assess feasibility of automated severity determination, alongside analysis of clinical and radiographic feature associations.
Results
not stated in source (abstract truncated before deep learning results)
Impact
Only feasibility of the approach on a small retrospective sample is claimed; performance results are not stated in the evidence.
Record
| Venue | Diagnostic and Interventional Radiology |
|---|---|
| Published | 2020 |
| Work type | article |
| Authors | Maxime Blain, Michael T. Kassin, Nicole Varble, Xiaosong Wang, Ziyue Xu, Daguang Xu, Gianpaolo Carrafiello, Valentina Vespro et al. |
| Citations | 61 (OpenAlex, at import time) |
| References | 50 |
| Model or method | not stated in source |
| Organization | tertiary care hospital in Milan, Italy (data source) |
| Primary source | https://doi.org/10.5152/dir.2020.20205 |
| Open access | bronze |
| Keywords | deep learning, chest X-ray, COVID-19, disease severity, radiology |
Importance rating
significant — thousands to millions affected
COVID-19 severity assessment from widely available chest X-rays could guide triage and management during an acute outbreak; only a small 48-patient feasibility study with no performance results stated.