Classification of periodic leg movements through actigraphy signal analysis
AbstractActigraphy can assist in the detection of periodic limb movements (PLMs) and other neuromuscular disorders in sleep. Although several actigraphs have been previously reported to accurately detect PLMs, most of them sample too infrequently to accurately detect PLMs. The Philips Respironics ® Actical™ is a readily available actigraph that has the capability of sampling at relatively high frequencies. Through simultaneous recordings of EEG, EMG and actigraph signals, clinicians have been attempting to understand and diagnose neurological diseases. Although a clinician’s expertise in diagnosing diseases is unquestionable, analyzing large amount of bio-signals for hidden information would require extensive informed aid from computer-based intelligent signal processing algorithms. These algorithms not only extract hidden information from the signal but also help in classifying between normal and abnormal test subjects based on their respective signal analysis. The objective of this research study is to develop novel tools for analyzing sleep actigraphy signals, captured using the Actical™, for estimating and classifying PLMs occurring during sleep. We simultaneously recorded polysomnography and bilateral ankle actigraphy in 108 consecutive patients presenting to our sleep laboratory. After pre-processing and conditioning, the bilateral ankle actigraphy signals were then analysed for 14 simple time, frequency and morphology-based features. Using a Naive-Bayes classifier we got a correct classification rate of 78.94%, with a sensitivity of 80.26% and a specificity of 73.68%. The algorithm developed in this study has the potential of facilitating identification of PLMs across a wide spectrum of patient populations via the use of bilateral ankle actigraphy.