MS Tez Sunumu: “Analysis of Spine Sounds for Spinal Health Assessment,” M. Arda Ahi (EE), EE-314, 14:00 29 Mayıs (EN)

SEMINAR: “Analysis of Spine Sounds for Spinal Health Assessment,” by M. ARDA AHI
M.S. in Electrical and Electronics Engineering
Prof. Dr. A. Enis Çetin

The seminar will be on Monday, MAY 29, 2017 at 14:00 @ EE-314

ABSTRACT
This thesis, in the light of accomplished joint health assessment techniques, proposes a spinal health assessment system based on acoustic bio-signals. The aim of this study is to offer an alternative to the conventional spinal health assessment techniques such as MR, CT or x-ray scans. As conventional methods are time-consuming, expensive and harmful (radiation risk caused by medical scanning techniques), a cheap, fast and harmless method is offered. In accordance with this purpose, a diagnosis algorithm was developed by using joint sounds collected from the vertebrae of subjects and automatic speech recognition (ASR) algorithms were used in order to accomplish feature extraction and classification. That being said, by using one of the most popular feature extraction method for speech recognition called Mel Frequency Cepstrum Coefficients (MFCC), a classification algorithm has been offered with popular classification method, Artificial Neural Networks (ANN). Furthermore, the scattering transform cepstral coefficients (STCC) algorithm is implemented as an alternative of mel filterbank in MFCC to increase the classification performance. The correlation between the medical history of the subjects and the click sound in the collected sound data is the basis of the classification algorithm. In the light of collected data, it is observed that the ‘click’ sound is detected in the individuals who have suffered low back pain (slipped disk) but not in healthy individuals. The identification of the ‘click’ sound is accomplished by using MFCC/STCC and ANN. The system has 83 percent success rate of detecting ‘click’ sounds in the given sound signal for MFCC algorithm, and with the alternative STCC algorithm, this success rate is 80 percent. Therefore, proposed method may detect individuals with spinal health problem without conventional medical scanning methods.