MULTI REGRESSIVE SPLINES AND QUANTUM CORRELATED PRENATAL DIAGNOSIS IN THE IDENTIFICATION OF CHROMOSOMAL ABNORMALITIES

Prenatal diagnosis is the identification of the chromosomal abnormalities has undergone a number of changes in the last few years. Irrespective of the healthcare system, in prenatal medicine, the advantages and disadvantages of chromosomal abnormalities must be comprehended. This is because chromosomes are found in the nucleus of cells that performs as a carrier of genetic information. Several machine learning and deep learning techniques has been explored in the recent past to establish prediction models and the challenges to assess chromosomal abnormalities. Despite the advantages that machine learning and deep learning offers in screening for measuring disorders, it must be recognized that it does not adequately address many other chromosomal disorders and any of the structural fetal anomalies in an accurate and precise manner. To fill this gap in this work a method called, Multi-regressive Splines and Quantum Correlated (MS-QC) prenatal diagnosis in the identification of chromosomal abnormalities is proposed. The MS-QC method is split into two sections, namely, preprocessing the data and pertinent chromosomal feature extraction for accurate and precise chromosomal abnormalities. Initially, with the input obtained from the prenatal cytogenetic Data set a machine learning based Multivariate Logistic Regression Spline function is applied that can capture complicated and multifaceted nature. The Multivariate Logistic Regression Spline function by including multiple independent variables can account for more characteristics that influence the dependent variable and hence minimize the overall error and bias involved in analyzing the risk factors for Prenatal Diagnosis in the Identification of the Chromosomal Abnormalities. With the preprocessed data, correlation based feature selection to evaluate distinct subsets on the basis of Quantum Distribution function according to features-class correlations is measured for extracting the most pertinent chromosomal features from fetal cells. The Quantum Distribution function here assists in extracting the most probable chromosomal features from fetal cells therefore obtaining more precise information of fetal disorders indicators. Simulations will be performed to validate the proposed method in Python language in terms of precision by 18%, recall by 26% and diagnosis accuracy by 15% respectively.