The invention belongs to the technical field of physiological information measurement, and provides a measurement
system and method based on multi-
modal physiological information and a wearable device, and the
system collects various different types of physiological signals including optical signals through a
signal collection module; physiological information such as heart driving force,
peripheral blood flow state and
tissue volume change is reflected respectively or synergistically; multi-
modal physiological features are extracted through the
feature extraction module, individual
physiological condition parameter vectors containing multiple information are constructed through the individual modeling module, and finally physiological parameter measurement results are obtained through the parameter calculation module by means of a neural
network model trained by introducing a physiological consistency constraint mechanism. According to the scheme, the defects that a single
signal is prone to being interfered and information is limited can be overcome through multi-
modal physiological
information fusion, the stability and accuracy of measurement are improved, personalized modeling and physiological consistency constraints are matched, the neural
network model conforms to the physiological mechanism while keeping the data driving expression capacity, and the method has wide application scenes.