Simulating time to event prediction with spatiotemporal echocardiography deep learning
This work addresses survival prediction for patients with cardiac conditions using echocardiography, but it is incremental as it applies existing methods to simulated data.
The study tackled the problem of predicting survival times from echocardiography videos by using simulated survival data based on ejection fraction readings, and found that spatiotemporal convolutional neural networks provided accurate survival estimates.
Integrating methods for time-to-event prediction with diagnostic imaging modalities is of considerable interest, as accurate estimates of survival requires accounting for censoring of individuals within the observation period. New methods for time-to-event prediction have been developed by extending the cox-proportional hazards model with neural networks. In this paper, to explore the feasibility of these methods when applied to deep learning with echocardiography videos, we utilize the Stanford EchoNet-Dynamic dataset with over 10,000 echocardiograms, and generate simulated survival datasets based on the expert annotated ejection fraction readings. By training on just the simulated survival outcomes, we show that spatiotemporal convolutional neural networks yield accurate survival estimates.