Fetal Health Analysis Using Machine Learning

Accurate detection of fetal birth weight plays a crucial role in assessing prenatal health and guiding appropriate care. In this study, we propose a method for fetal high birth weight, normal birth weight, and low birth weight detection using image processing techniques in machine learning (ML). A dataset of ultrasound images, along with corresponding birth weight outcomes and relevant demographic information, was collected. The images were pre-processed to enhance quality and normalize brightness and contrast. Key features, including biometric measurements such as abdominal circumference, head circumference, and femur length, were extracted from the ultrasound images. Texture analysis techniques were applied to capture textural features indicative of fetal size and growth patterns. The proposed methodology achieved promising results in accurately classifying fetal birth weight categories. The findings suggest that image processing in ML holds great potential for non-invasive fetal weight estimation, enabling early identification of high birth weight and low birth weight cases for targeted prenatal interventions. Further research and validation on larger datasets are warranted to establish the clinical utility and generalizability of the proposed approach.