An Automated Deep Learning Model for the Cerebellum Segmentation from Fetal Brain Images

Abstract

Cerebellum measures taken from routinely obtained ultrasound (US) images have been frequently employed to determine gestational age and identify developing central nervous system’s anatomical abnormalities. Standardized cerebellar assessments from large-scale clinical datasets are required to investigate correlations between the growing cerebellum and postnatal neurodevelopmental results. These studies could uncover structural abnormalities that could be employed as indicators to forecast neurodevelopmental and growth consequences. To achieve this, higher-throughput, precise, and impartial measures must be used to replace the existing human, semiautomatic, and advanced algorithms, which seem to be time-consuming and inaccurate. In this article, we presented an innovative deep learning (DL) technique for automatic fetal cerebellum segmentation from 2-dimensional (2D) US brain images. We present ReU-Net, a semantic segmentation network tailored to the anatomy of the fetal cerebellum. Moreover, we use U-Net as a foundation models with the incorporation of residual blocks and Wiener filter over the last 2 layers to segregate the cerebellum (c) from the noisy US data. 590 images for training and 150 images for testing were taken; also, we employed a 5-fold cross-assessment method. Our ReU-Net scored 91%, 92%, 25.42, 98%, 92%, and 94% for Dice Score Coefficient (DSC), F1-score, Hausdorff Distance (HD), accuracy, recall, and precision, correspondingly. The suggested method outperforms the other U-Net predicated techniques by a quantitatively significant margin ( p 0.001 ). Our presented approach can be used to allow high bandwidth imaging techniques in medical study fetal US images as well as biometric evaluation on a broader scale in fetal US images.

Enhancing Fetal Anomaly Detection in Ultrasonography Images: A Review of Machine Learning-Based Approaches

“…These challenges can be mitigated through the incorporation of techniques such as regularization parameter tuning and the strategic use of skip connections, as exemplified by ResNets. The inclusion of skip connections not only alleviates the vanishing gradient problem but also [58,[95][96][97][98][99][100][101][102][103]107,109,[115][116][117]120,[124][125][126][127][128][129][130][135][136][137]141,142,145,150,153,157,160,161,169]. Each study's color code reflects its relevance to a certain organ, including the heart, brain, lung, or analyzing for chromosomal abnormalities.…”

“…Each entry provides information about the employed methods, total number of images, key performance metrics, and application domain. [58,[95][96][97][98][99][100][101][102][103]107,109,[115][116][117]120,[124][125][126][127][128][129][130][135][136][137]141,142,145,150,153,157,160,161,169]. Each study's color code reflects its relevance to a certain organ, including the heart, brain, lung, or analyzing for chromosomal abnormalities.…”

“…During the 18-to 21-week pregnancy period, US imaging is used to measure the cerebrum, midbrain, cerebellum, brainstem, and other regions of the brain as part of the screening for fetal abnormalities [113,114]. To detect fetal brain abnormalities, Sreelakshmy et al developed a model (ReU-Net) based on U-Net and ResNet for the segmentation of fetuses' cerebellum using 740 fetal brain US images [115].…”

Advanced neuroimaging in the fetal brain: An update on technical advances and clinical findings

“…Therefore, quantitative methods usually rely on automatic segmentation algorithms to achieve accurate and reproducible measurements. Deep convolutional neural networks have performed well in fetal medical image analysis, especially fetal brain segmentation [17].…”

Evolving the Era of 5D Ultrasound? A Systematic Literature Review on the Applications for Artificial Intelligence Ultrasound Imaging in Obstetrics and Gynecology

“…Researchers Xie et al [6,93] and Sahli et al [94] reported a method for classifying US images into a binary system of 'normal' and 'abnormal' cases, in which Xie et al additionally localized the structural lesions, which lead the algorithm to declare it 'abnormal' and thus recommend the clinician to recheck the labeled area. Lastly, the studies of Burgos-Artizzu et al and Sreelakshmy et al portrayed AI methods for the estimation of GA through an analysis of transthalamic axial planes or cerebellum measurements [95,96].…”