Analysis of EMG During Clonus Using Wavelets

摘要: of a thesis at the University Miami. Thesis supervised by Dr. Jorge Bohorquez and Christine K. Thomas. No. pages in text. (92) Involuntary muscle contractions (spasms) are major secondary consequence spinal cord injury. These spasms disrupt mobility ability to perform daily activities. The rhythmic repetitive clonus one kind spasm. In this study an algorithm was developed automatically detect start end times EMG bursts during clonus. measures were used calculate duration bursts, frequency intensity (root mean square) each burst, parameters that characterize This relied on technique analysis (Von Tscharner 2000). Filters created non-linearly scaling Mother (Morlet) wavelet produce envelopes different bands. intermediate band (80-190 Hz) enveloped best To threshold time separation imposed eliminate multiple peaks caused baseline EMG, motor units or changes. Window regions extended between midpoints identified then resized 50 ms either side peak. 5% 95% energy contained window region, respectively. A unit constraint potentials beginning output from 31 long term (24 hr) data recorded 8 paralyzed leg muscles 7 subjects with chronic cervical injury compared generated two independent human operators. as good operator identifying (p = 0.946), (intra class correlation coefficient α 0.949), contraction (α 0.997) durations burst 0.852). On average 574 (SE 238) faster than manual people (p≤ 0.001). Analysis 24 hour dataset right medial gastrocnemius showed more prevalent stronger awake versus sleep time. can be analyze recordings accurately limited user intervention. may also prospective diagnostic tool judge effectiveness interventions such drugs like baclofen mitigate is dedicated my parents family