THE ASSESSMENT OF LH SURGE FOR PREDICTING OVULATION TIME USING CLINICAL, HORMONAL, AND ULTRASONIC INDEXES IN INFERTILE WOMEN WITH AN ENSEMBLE OF NEURAL NETWORKS

An ensemble of independently trained neural networks (NN) is proposed for the assessment of luteinizing hormone (LH) surge for predicting ov ulation time in infertile but ovulating women. The proposed ensemble i nvolves a number of parallel NN modules. Each pair of the NNs learn sp ecific data that are previously collected for monitoring timing functi on of LH levels. Training data which correspond to values of serum pro gesterone (ng ml(-1)), serum est radiol (pg ml(-1)), and follicle diam eter (mm) are used to train NN pairs to approximate the function of th e LH values. A reasonable and accurate estimation places ovulation app roximately 10-12 h after the LH peak. The double-valued (bi-phasic) re gions of training data are separated into two single-valued (bi-phasic ) regions of training data are separated into two single-valued parts (not exactly preovulatory, postovulatory division) that can be learned by each module of the NN pair. During testing, after the initial deci sion to have single-valued sides, the assessment is obtained by a line ar opinion pool (consensus rule) using the decisions of NNs on the cor responding side without waiting. The network ensemble has various desi rable properties: high assessment accuracy of a double-valued multisou rce data, minimized learning and recall times, and a parallel structur e. The ovulation time can be predicted through the assessment of LH pe ak with a better precision and fewer number of tests.