Pattern Classification by the Hotelling Statistic and Application to Knee Osteoarthritis Kinematic Signals

摘要

The analysis of knee kinematic data, which come in the form of a small sample of discrete curvesthat describe repeated measurements of the temporal variation of each of the knee three fundamentalangles of rotation during a subject walking cycle, can inform knee pathology classification because, ingeneral, different pathologies have different kinematic data patterns. However, high data dimensionalityand the scarcity of reference data, which characterize this type of application, challenge classification andmake it prone to error, a problem Duda and Hart refer to as the curse of dimensionality. The purpose ofthis study is to investigate a sample-based classifier which evaluates data proximity by the two-sampleHotelling T2 statistic. This classifier uses the whole sample of an individual’s measurements for a bettersupport to classification, and the Hotelling T2 hypothesis testing made applicable by dimensionalityreduction. This method was able to discriminate between femero-rotulian (FR) and femero-tibial (FT)knee osteoarthritis kinematic data with an accuracy of 88.1%, outperforming significantly currentstate-of-the-art methods which addressed similar problems. Extended to the much harder three-classproblem involving pathology categories FR and FT, as well as category FR-FT which represents theincidence of both diseases FR and FT in a same individual, the scheme was able to reach a performancethat justifies its further use and investigation in this and other similar applications.