Plenary lecture 6: digital signal processing and softcomputing methods applied to ECG and PCG to infer about subject physiological status

Electrocardiogram (ECG) and Phonocardiogram (PCG) signals embed key information about subject physiological status. Physicians use such information to diagnose the general physiological status of the patient and also to diagnose incoming pathologies. The automatic extraction of such information enables the implementation of systems that could monitor subjects and infer about some physiological and pathological condition, so that efficient prevention strategies could be enabled. ECG and PCG features extraction is a set of Digital Signal Processing-based (DSP-based) application-specific algorithms (ASA) to execute crisp measurements to be inputted to a fuzzy logic engine that infers about a physiological and/or pathological status of the subject. Fuzzy logic engine is a data driven inferential process, so most of the design efforts need to be addressed to the features extraction process and to the rule set and membership functions definition and tuning. Low level and high level signal features can be extracted, so that a very effective inference could be applied at fuzzy logic engine level. Heart rate variability (HRV) is one of the high level signal features that embeds veri significant information about physiological and pathological condition of the subject.