The recognition of severe thoracic aortic dissection using conventional chest radiography in conjunction with a whale optimized bilateral residual convolutional neural network

Abstract

A life-threatening condition called severe thoracic aortic dissection (STAD) is brought on by blood leakage from the aorta's injured inner layer, which separates the intimal and adventitial layers. It is difficult to make a diagnosis for this disease. Chest x-rays are frequently used for initial screening or diagnosis, however their diagnostic accuracy is not great. Deep learning (DL) has recently been effectively used for a variety of medical image processing tasks. By using DL techniques, we try to enhance the accuracy of the diagnosis of STAD made on the basis of chest x-rays in this work. The significant thoracic aortic dissection was detected using the Whale optimized bilateral residual convolutional neural network (WO-BRCNN). The WO-BRCNN accuracy was found to be 99.21%, with precision at 94.93%, recall at 97.89%, F1-score at 95.71% and specificity at 93.42%. To increase diagnostic accuracy using aorta segmentation, further study is required.

Imprint

Naveen Meena, Amit Kumar Bishnoi, Malathi.H The recognition of severe thoracic aortic dissection using conventional chest radiography in conjunction with a whale optimized bilateral residual convolutional neural network. Cardiometry; No.26 February 2023; p.-; DOI: .; Available from: https://www.cardiometry.net/issues/no26-february-2023/the-recognition-of-severe-thoracic-aortic-dissection