Anaerobic threshold measurement using dynamic neural network models

This paper deals with the subject of anaerobic threshold measurements for a thletes involved in aerobic or aerobic/anaerobic sports. Traditionally, ana erobic threshold has been determined using invasive tests or using a non-in vasive technique using steady-state heart-rate/work rate data. Non-invasive tests have the advantage of not requiring specialised equipment, but the a cquisition of steady-state information can be problematic. This paper demon strates how dynamical data can be used to accurately determine the steady-s tate heart-rate/work-rate curve (SSHW curve) using neural network dynamic m odels. (C) 1999 Elsevier Science Ltd. All rights reserved.