Artificial Intelligence Approach for Detecting Macrocephaly and Microcephaly in Avoiding Pregnancy Complications

Abstract

Ultrasound imaging is one of the vital image processing techniques that aids doctors to access and diagnose the feotal growth process by measuring head circumference (HC). This chapter gives a detailed review of cephalic disorders and the importance of diagnosing disorders in the earlier stage using ultrasound images. Additionally, it proposes an approach that uses four primary stages: pre-processing, pixel-based feature extraction, classification, and modeling. A cascaded neural network model based on ultrasound images is recommended to identify and segment the HC of the feotus during the extraction phase. According to the findings of the experiments, both the rate of head circumference measurement detection and segmentation accuracy has significantly increased. The proposed method surpasses the state-of-the-art approaches in all criteria, two assessment criteria for HC measurement, is qualitatively distinct from other prior methods, and attained an accuracy of 96.12%.