Early Detection of Down Syndrome Through Ultrasound Imaging Using Deep Learning Strategies — A Review

Another name for Down syndrome is Trisomy 21, which is a complex chromosomal disorder that poses a substantial challenge in prenatal care due to its widespread prevalence. An additional copy of chromosome 21 is the defining feature of this disorder, leading to numerous physical and cognitive impairments. Typical physical traits of Down syndrome include a flattened facial structure, most notably in the nose bridge region, as well as upward-slanted almond-shaped eyes and a shortened neck. Apart from these features, individuals with Down syndrome also experience a range of intellectual disabilities, spanning from mild to severe. These challenges manifest in various areas, such as expressive communication, verbal memory, and attention. Furthermore, children with this disorder may exhibit behavioral issues that require special attention and care. Early and accurate detection of this condition is essential for informed decision-making by expectant parents. This review explores an overview of the data sets and results of recent studies that have used deep learning and machine learning methodologies to create effective models for the identification of Down syndrome.