Evaluation of state-of-the-art deep learning models in the segmentation of the left and right ventricles in...

Previous studies on echocardiogram segmentation are focused on the left ventricle in parasternal long-axis views. Deep-learning models were evaluated on the segmentation of the ventricles in parasternal short-axis echocardiograms (PSAX-echo). Segmentation of the ventricles in complementary echocardiogram views will allow the computation of important metrics with the potential to aid in diagnosing cardio-pulmonary diseases and other cardiomyopathies. Evaluating state-of-the-art models with small datasets can reveal if they improve performance on limited data.

PSAX-echo was performed on 33 volunteer women. An experienced cardiologist identified end-diastole and end-systole frames from 387 scans, and expert observers manually traced the contours of the cardiac structures. Traced frames were pre-processed and used to create labels to train two domain-specific (Unet-Resnet101 and Unet-ResNet50), and four general-domain [three segment anything (SAM) variants, and the Detectron2] deep-learning models. The performance of the models was evaluated using the Dice similarity coefficient (DSC), Hausdorff distance (HD), and difference in cross-sectional area (DCSA).

The Unet-Resnet101 model provided superior performance in the segmentation of the ventricles with 0.83, 4.93 pixels, and $106{\text{pixel}}^2$ on average for DSC, HD, and DCSA, respectively. A fine-tuned MedSAM model provided a performance of 0.82, 6.66 pixels, and $1252{\text{pixel}}^2$ , whereas the Detectron2 model provided 0.78, 2.12 pixels, and $116{\text{pixel}}^2$ for the same metrics, respectively.

Deep-learning models are suitable for the segmentation of the left and right ventricles in PSAX-echo. We demonstrated that domain-specific trained models such as Unet-ResNet provide higher accuracy for echo segmentation than general-domain segmentation models when working with small and locally acquired datasets.