Abstract:
Background: Pacing leads with miniaturized accelerometers monitor myocardial function. This requires acceleration signal functional indices. A method that automatically detects aortic valve opening (AVO) and closure (AVC) will aid such extraction. We used high-fidelity pressure measurements to test whether deep learning can detect these valve events from epicardially attached accelerometers. Method: A CNN, RNN, and multi-head attention module deep neural network was trained and tested on 130 recordings from 19 canines and 159 from 27 porcines covering different interventions. Nested cross-validation assessed method accuracy due to limited data. Results: Canines had 98.9% and 97.1% correct detection rates for AVO and AVC, while porcines had 98.2% and 96.7%. AVO and AVC were incorrectly detected 0.7% and 2.3% in canines and 1.1% and 2.3% in porcines. The mean absolute error between correct detections and their ground truth in canines was 8.4 ms, and in porcines it was 8.9 ms. Conclusion: Epicardially attached accelerometers can be used with deep neural networks to accurately detect aortic valve opening and closing.
Note: Please discuss with our team before submitting this abstract to the college. This Abstract or Synopsis varies based on student project requirements.
Did you like this final year project?
To download this project Code with thesis report and project training... Click Here
