Correlation of BAI psychological test scores with heart rate variability using wavelet packet transform and artificial neural networks in fibromyalgia syndrome

Abstract

Fibromyalgia syndrome which is appeared in the form of common pain in women is a musculoskeletal disorder. Heart rate variability (HRV) is a signal as measured time between each successive QRS time obtained from ECG signal. HRV parameters are associated with autonomic nervous system in literature. FMS affects patient's psychology. Consequently some psychological tests are applied to patients for evaluation of psychological effects. Beck Anxiety Inventory (BAI) Test being applied as writing and speaking is a test consisting of 21 questions. In the study, HRV signals obtained from FMS patients and control group are decomposed into wavelet packets using Wavelet Packet Transform (WPT) and frequency bands which related to autonomic nervous system are included to study. Wavelet Packets within these frequency bands are applied to inputs of multilayer perceptron artificial neural networks (MLPNN). BAI psychological test scores are included as target values for MLPNN and each input is trained and tested as relating to target values. According to obtained accuracy values, the packets within Very Low Frequency (VLF) band has minimum accuracy values, however the packets within Low Frequency + High Frequency (LF+HF) has the best accuracy values. This study is targeted on presenting a solution for clinical studies which are evaluated using psychological test scores by physiological support.