Wyniki wyszukiwania - BazTech

1

Classification of auditory brainstem response using wavelet decomposition and SVM network

Dobrowolski A., Suchocki M., Tomczykiewicz K., Majda-Zdancewicz E.

Biocybernetics and Biomedical Engineering

|

2016

|

Vol. 36, no. 2

427--436

EN

In electrophysiological hearing assessment and diagnosis of brain stem lesions are most often used auditory brainstem evoked potentials of short latency. They are characterized by successively arranged maxima as a function of time, called waves. Morphology of the course, in particular, the timing and amplitude of each wave, allow neurologist diagnosis, which is not an easy task. Neurologist requires experience, attention and very good perception. In order to support the diagnostic process, the authors have developed an algorithm implementing the automated classification of auditory evoked potentials to the group of pathological and physiological cases. The sensitivity and specificity of group numbering of 130 cases are respectively 95% and 98% and classification accuracy is equal to 97%. The procedures developed by the authors for generation of distinctive features based on wavelet decomposition with a SVM network-based classifier have been integrated into a diagnostic application directly interoperable with Nicolet Viking Select (Natus Medical Inc., USA) system data files.

2

Komputerowa analiza słuchowych potencjałów wywołanych pnia mózgu

Dobrowolski A.P., Suchocki M., Tomczykiewicz K.

Przegląd Elektrotechniczny

|

2011

|

R. 87, nr 9a

145-150

PL

W referacie przedstawiono problematykę związaną z badaniem słuchu metodą potencjałów wywołanych pnia mózgu (SPWPM). Ze względu na to, że obecne metody opierają się głównie na interpretacji przebiegów czasowych potencjałów i wymagają od badającego neurologa dużego doświadczenia, podjęto się zadania, które ma na celu eliminację subiektywnej oceny wyników oraz zautomatyzowanie badania. Autorzy korzystając z metod cyfrowego przetwarzania sygnałów wyodrębnili cechy dystynktywne SPWPM dobrze różnicujące przypadki prawidłowe od patologicznych. Wykorzystanie dekompozycji falkowej i liniowej sieci SVM zapewniło prawidłową klasyfikację 40 różnych przypadków klinicznych, w tym 20 przypadków prawidłowych i 20 patologicznych.

EN

The paper presents problems which are connected with the examinations of auditory evoked potentials. Due to the fact that current method are mainly based on timing interpretation of potentials waveform and that require a lot of experience from examining neurologist, authors have undertaken the task, which will main goal is elimination of subjective assessment results and automation of examinations. The authors use the methods of digital signal processing isolated the distinctive features of evoked potentials, which truly differentiate the correct from pathological cases. The use of wavelet decomposition and linear SVM network provide the correct classification of 40 different clinical cases, including 20 normal and 20 pathological cases.

3

Application of tone-pip stimuli of different rise-times for wave V identification in auditory brainstem responses (ABR) procedures

Hatliński G. J., Kochanek K., Piłka A., Bochenek W.

Biocybernetics and Biomedical Engineering

|

2008

|

Vol. 28, no. 4

51-57

EN

Application of tone-pip stimuli of elongated rise-times in examination of auditory brainstem responses (ABR TON) makes it possible to detect desynchronization of auditory nerve fibers in a much earlier stage of retrocochlear pathology than it would be possible when applying click-type acoustic stimuli. However, longer rise-time deteriorates synchronization of responses from individual nerve fibers. It makes the assessment of ABR morphology more difficult, in particular it hinders correct recognition of wave V. The quality of ABR recordings might be poor when examining retrocochlear pathologies, and/or in presence of excessive artifacts. The authors propose that, in the cases when accurate evaluation of wave V in ABRs recorded with the use of long-rise tone-pip stimuli (ABR TON, 4 or 8 cycles rise-time) is difficult, one should apply tone-pip stimuli of shortened rise-time (ABR TON-2, 2 cycles rise-time) to correctly identify wave V.