| 摘要 |
The present disclosure relates to a device, comprising a memory and a processing unit, for classifying heart sounds presented in a phonocardiogram. The processing unit performs a first step of receiving a phonocardiogram comprising at least one cardiac cycle (X). A first point (P1 ) in relation to a first heart sound (S1 ), and a second point (P2) in relation to a second heart sound (S2), in said cardiac cycle (X), are determined. A first analysis on said heart sound is performed over a first time window (w1) having e.g. a starting point (SP1 ) essentially equal to said first point (P1 ), and a duration of a first predetermined time (T1), which is shorter than a time period (P1 P2). At least one further analysis is performed on said heart sound over at least one further time window (wn), having a duration of a predetermined time (Tn), which is the sum of the duration of the previous window (w1, wn) and a predetermined time period (&Dgr;t), but shorter than said time period (P1 P2). Said analysis steps comprise extracting spectral energies from said phonocardiogram, and from said analysis, said heart sound is classified as non-pathological or pathological. |
