From Mining Tinnitus Database to Tinnitus DecisionSupport System, Initial Study.pdfVIP

From Mining Tinnitus Database to Tinnitus Decision-Support System, Initial Study Pamela Thompson, Xin Zhang, Wenxin Jiang, Zbigniew W. Ras Department of Computer Science, University of North Carolina at Charlotte pthompso@catawba.edu, {xinzhang, wjiang3, ras}@uncc.edu Abstract activity within the cochlea, and not related to external stimulation of any kind” [2]. Hyperacusis or decreased Many definitions for tinnitus exist and causes and sound tolerance frequently accompanies tinnitus and treatments are plentiful, yet not completely can include symptoms of misophonia (strong dislike of understood. A database of eleven tables, as many as sound) or phonophobia (fear of sound). Physiological 555 unique patien tuples and numerous time-stamped causes of tinnitus can be difficult or impossible to and other features was obtained for knowledge determine, and treatment approaches vary. discovery related to causes and treatments of tinnitus. Tinnitus Retraining Therapy (TRT), developed by The paper describes the knowledge discovery and Dr. Jastreboff, is one treatment model with a high rate machine learning process and introduces several new of success and is based on a neurophysical approach to temporal features to improve tinnitus evaluation, treatment. TRT “cures” tinnitus by building on its outcomes analysis, and overall understanding. association with many centers througho