Data Mining to Support the Discrimination of Amyotrophic Lateral Sclerosis Diseases Based on Gait Analysis

Abstract

In many medical researches, dealing with huge dataset is crucial. However, it is difficult to use standard methodologies to analysis huge dataset. In such cases, data mining tools can support doctors for better diagnosis. Data mining tools have been utilized in clinical data analysis such as biological signals, clinical images to analysis and detect diseases. The utilization of these techniques can increase diagnostic sensitivity and specificity. It can help reduce the misdiagnosis, in addition to early prediction. This paper discusses the use of machine learning algorithms for the detection and classification of amyotrophic lateral sclerosis disease using gait data. Our analysis indicated that a number of machine learning algorithms such as the linear discriminant classifier and quadratic discriminant classifier can discriminate between normal and abnormal cases of amyotrophic lateral sclerosis disease.