Objective: To develop a probabilistic computer model to predict the preovulatory days of the menstrual cycle by given hormonal parameters such as follicle-stimulating hormone (FSH), luteinizing hormone (LH), estradiol (E2), and progesterone (P).
Design: A computerized analysis program is specifically designed for the normal human menstrual cycle. The algorithm of this model incorporates the statistical aspects and the daily variations of FSH, LH, E2, and P.
Patients: This study includes 16 healthy fertile women with ovulatory cycles.
Interventions: Daily venous blood samples were collected from each subject from the first to the last day of the menstrual cycle. Radioimmunoassays were used to measure the four hormones in each blood sample.
Results: The computer implementation of this menu driven code that is flexible to accommodate differences in the hormone measurement procedures was accomplished. The distribution of measured hormone concentrations was validated to be normal. The best estimation of the ovulation day was highly dependent on the use of the band width (acceptable range) of the daily SD, which was +/- 1.45 in our laboratory.
Conclusion: This computer model, heretofore named the CESME (Computer Enhanced Systems in Medicine and Engineering) developed by two of the authors, Ger and Karamete, with its reconfigurability enables the user to adopt the program to the normal values in their laboratory and use its clinically for predicting the probable periovulatory day(s) of the menstrual cycle.