An algorithm for detecting features of the hormone profiles of the human menstrual cycle

An algorithm for detecting features of the cycles of the gonadotropic and o varian hormones in women is described. The algorithm can detect hormone pea ks and normal cycles defined in terms of the peaks in sequences of measurem ents that have an arbitrary starting point in the menstrual cycle and are o f arbitrary length. The algorithm makes use of fuzzy set theory and is opti mized using signal detection theory. Published by Elsevier Science Ltd.