A review of approaches investigated for right ventricular segmentation using short‐axis cardiac MRI

Abstract

The right ventricular assessment is crucial to heart disease diagnosis. Unfortunately, its segmentation is quite challenging due to its intricate shape, ill-defined thin edges, large variability among patients, and pathologies. Besides, it is a very laborious and time-consuming task to be done manually. Therefore, automated segmentation techniques are very suitable to reduce the strain on the expert. Here, it is attempted to review the taxonomy of the current RV segmentation approaches adopted to handle the afore-mentioned issues. Enhanced by our expert's interpretation, the results of over forty research papers were evaluated based on several metrics such as the dice metric and the Hausdorff distance. Synthetic tables and charts were also used to discuss the reviewed approaches. The following study shows that none of the existing methods has proved accurate enough to meet all the RV challenging issues. Many misestimated results were reported for several cases. Eventually, global guidance is outlined, which supports combining different methods to enhance the expected results during the MRI short-axis slice processing.

Analysis of Right Ventricle Segmentation in the End Diastolic and End Systolic Cardiac Phases Using UNet-Based Models

“…To tackle the challenges of Right Ventricle segmentation, various works were proposed employing different segmentation techniques [8]. As reviewed in [6] and [9], the most recently proposed methods are more oriented to use deep learning techniques. In fact, Good progress in the medical imaging field has been reached thanks to the introduction of Artificial Intelligence technologies that became a popular approach for detection and segmentation problems due to their powerful feature representation [10].…”

“…To analyze the RV function, radiologists have to delineate its boundaries over the entire slices which is a time-consuming task. For this reason, automatic segmentation of this cardiac cavity has been studied using multiple approaches [6]. Despite the inspiring results obtained in the End Diastolic (ED) phase, lower results were detected in the End Systolic (ES) phase for many proposed approaches [7].…”

A fusion‐attention swin transformer for cardiac MRI image segmentation

“…Due to the complexity of the physiological geometric structure of the heart, artifacts in the imaging process caused by blood flow, the uneven image grey distribution, and the blurred target boundary caused by the interference of papillary muscles, it is especially difficult to image segments of heart organs [1]. Moreover, the non‐uniformity and anatomical variability of imaging, as well as the inherent geometric and dynamic complexity of the heart, bring great challenges to accurate segmentation of cardiac organs and tissues based on MRI images [2].…”

Ventricular segmentation and modeling using topological watershed transformation and harmonic state model

“…This progress requires acquisition based on various slice planes. The main slice planes used for the right ventricle (RV) [9], which is the object of interest in this paper, are: the long axis 4 cavities slice and the short axis slice [10], [11].…”

“…The main slice planes used for the RV, 10 which is the object of interest in this paper, are: the long axis 4 cavities slice and the short axis slice. 5,11 Figure 1 presents an example.…”