A fuzzy constraint satisfaction approach to identify and characterize apnea episodes

This paper presents an algorithm that permits the identification of apneas - cessations in the sleeping patient's respiratory flow - in the respiratory airflow signal and relates them to the drops in blood oxyhemoglobin saturation that they produce. The structural nature of the algorithm allows us to perform a detailed characterization of the identified events and to easily modify the morphological detection criteria. This proposal is based on the fuzzy set theory for the representation and manipulation of the vagueness of the medical knowledge on which it is based, and on the constraint satisfaction problem formalism to provide a computable support to medical knowledge.

keywords: Sleep Apnea Syndrome, Biosignal Processing, Structural Pattern Recognition, Fuzzy Constraints