Key result
A convolutional neural network identified the manufacturer of cardiac rhythm devices from radiographs with 99.6% accuracy, significantly outperforming cardiologists (72.0% median accuracy; p<0.0001).
Why the study?
Medical staff frequently need to quickly and accurately identify the model of a pacemaker or defibrillator from a chest radiograph, but current methods rely on comparing radiographic appearance with a manual flowchart.
Does a convolutional neural network improve the accuracy of identifying cardiac rhythm device manufacturers from chest radiographs compared to cardiologists using a manual flowchart?
Does a convolutional neural network improve the accuracy of identifying cardiac rhythm device manufacturers from chest radiographs compared to cardiologists using a manual flowchart?
Absolute Event Rate: 99.6% vs 72%
p-value: p=<0.0001
A neural network can accurately identify the manufacturer and model group of cardiac rhythm devices from chest radiographs, significantly outperforming human cardiologists using standard flowcharts.
No takes yet. Share an insight, caveat, or question.
AI device identification from radiographs may speed emergency management; leaves open prospective validation before clinical adoption.
Howard et al. (2019) studied Cardiac rhythm devices (pacemakers or defibrillators) (n=1,676). Convolutional neural network vs. Cardiologists using a published flowchart was evaluated on Accuracy in identifying the manufacturer of a device from a radiograph (p=<0.0001). A convolutional neural network identified the manufacturer of cardiac rhythm devices from radiographs with 99.6% accuracy, significantly outperforming cardiologists (72.0% median accuracy; p<0.0001).
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: