The invention discloses a sleep
hypopnea type automatic discrimination method and
system based on a multi-
modal signal, and aims to solve the technical bottlenecks of strong subjectivity, poor consistency and limited single
signal feature discrimination capability of the existing visual scoring method. The method comprises the following steps: collecting multi-
modal signals in a polysleep pattern (PSG), wherein the multi-
modal signals comprise nasal ventilation
airflow, thoracico-abdominal
breathing movement, blood
oxygen saturation and other signals; extracting multi-dimensional features related to
hypopnea typing, wherein the multi-dimensional features comprise an inspiration flow limited index, thoracico-abdominal movement coordination, an event termination mode, an awakening sequential relationship and a sleep stage associated feature; the features are fused and optimized based on a
machine learning
algorithm, a high-precision classification model is constructed, and automatic discrimination of obstructive and central hypoventilation is realized. According to the method, the key problems that a traditional method depends on manpower in
hypopnea typing, feature utilization is not comprehensive, and judgment precision is insufficient are solved, and a reliable basis is provided for accurate diagnosis,
typing and
treatment strategy formulation of sleep
breathing disorder.