Detection of Valve Vegetations in Native and Prosthetic Valves using Echocardiographic Radiomics and Deep Learning on Transesophageal Echocardiography Images

Background: Infective Endocarditis (IE) is a life-threatening condition that requires rapid and accurate diagnosis. Transesophageal Echocardiography (TEE) is the gold standard for detecting valve vegetations; however, its interpretation is highly operator-dependent and particularly challenging in patients with prosthetic valves. Recent advances in artificial intelligence, especially deep learning, offer opportunities to improve diagnostic accuracy and reduce observer variability.Objective: This study aimed to evaluate the performance of deep learning–based models for detecting vegetations in TEE images to support the diagnostic workflow of IE.Material and Methods: In this retrospective experimental study, a Faster Region-based Convolutional Neural Network (Faster R-CNN) was implemented to localize valve vegetations in TEE images. Four model configurations were developed using DenseNet۱۲۱ and ResNet۵۰ backbones, each trained in frozen and fine-tuned modes. All models were pretrained on RadImageNet. The dataset consisted of ۱,۰۰۰ annotated TEE frames acquired from both native and prosthetic heart valves.Results: The fine-tuned DenseNet۱۲۱ model achieved the best performance, with a mean Average Precision (mAP) of ۰.۶۵۳ and an Area Under the Curve (AUC) of ۰.۸۵۸. Its frozen version demonstrated lower performance (mAP=۰.۴۱۶, AUC=۰.۶۴۰). The fine-tuned ResNet۵۰ model reached a mAP of ۰.۵۹۳ and an AUC of ۰.۷۸۹, while the frozen ResNet۵۰ showed the lowest performance (mAP=۰.۴۰۳, AUC=۰.۶۰۱). Conclusion: Both fine-tuned DenseNet۱۲۱ and ResNet۵۰ models demonstrated effective localization of valve vegetations in TEE images, with comparable IoU performance. Although DenseNet۱۲۱ showed superior classification accuracy, the similar localization results highlight the potential of both models as physician-assistive tools for enhancing IE diagnostic workflows.

کلیدواژه ها:

Infective Endocarditis ، Transesophageal Echocardiography ، Deep Learning ، Faster R-CNN ، Valve Vegetations ، Computer-Assisted Image Interpretation ، Heart Valves