Detection and Prediction of Cardiac Amyloidosis disease using deep learning algorithms- Comparative Study

Abstract

Extracellular deposition of amyloid fibrils within the heart greatly expands the extracellular volume in Cardiac Amyloidosis (CA). Patients who are affected with CA disease have significant heart failure and have a poor prognosis too. To resolve these issues, this paper mainly focused on deep learning based classification algorithm such as VGG-16, VGG-19, Xception model and DenseNet201 approach for predicting heart disease especially CA using ECG images. Moreover, metrics such as precision, recall, F1-score and accuracy are estimated to find the performance of deep learning models in CA disease diagnosis. Our experimental outcomes show that DenseNet201 model attains higher accuracy of 92.86% in disease identification and classification too.