VNG technique for a convenient vestibular neuritis rating

ABSTRACT

Vestibular Neuritis (VN) joins to the most significant public health concern. A lot of videonystagmographic (VNG) datasets are admitted to clinical assessment methods, which impose a serious problem in term of complexity. The aim of this work is to develop a simple, fast and intelligent method to identify subjects with a high risk of VN disease. This paper proposes a real-time digital signal controller (dsPIC) based system with a digital output display indicating the VN existence. The clinical feature inputs are extracted from the current VNG analysis. The proposed method has been experimented on a database including 73 patients affected by vestibular neuritis (VN) proceeded with saccadic, kinetic and caloric tests for basic measures. Moreover, the VNG characteristics are divided into two groups: VN and HL cases. The obtained classification results have achieved the best precision when applying the supervised multilayer neural network (MNN). As stated in the performance assessment, we recorded more than 0.9576 for accuracy of detected vestibular neuritis supplied by ENT pathologists which reveals the highest Positive Predictive Values with the specialist’s result (PPV = 0.9528, Negative Likehood ratio <0.2). This framework shows ENT application of vestibular dysfunction as a successful tool for automatic VN evaluation without expert intervention.

KEYWORDS:

• Vestibular neuritis (VN)
• dsPIC implementation
• videonystagmographic test (VNG)
• multilayer neural network (MNN)
• classification

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Hanene Sahli

Hanene Sahli, PhD, received a doctorate degree in science and technology-electrical engineering, from the University of Tunis, ENSIT. She is a member of the research group in Signal Image and Energy Mastery laboratory (SIME) at the same university. Her research interests include machine learning methods, pattern recognition,biomedical video and image analysis and biomedical signal processing.

Amine Ben Slama

Amine Ben Slamaa is a PhD Doctor in Biophysics and medical imaging and member of research group in Laboratory of Biophysics and medical technologies at ISTMT - University of Tunis El Manar. His research interests include image and signal processing, classification and intelligent data processing for vestibular disorder diagnostic.

Sami Bouzaiane

Sami Bouzaiane is a Professor in electronics at the Naval Academy in the Ministry of Defense of Tunisia. He is actually a member of the research laboratory of the Naval academy and a research collaborator of the research laboratory Signal Image and Energy Mastery (SIME). His research interests are focused on analog and digital electronics, microcontrollers, embedded systems and automatic control.

Jihene Marrakchi

Jihene Marrakchi is a doctor-hopsital university assistant of otorhinolaryngology in La Rabta hospital, Tunisia. Her research interests include oral, thyroid, vocal cord surgery and vertigo exploration.

Seif Boukriba

Seif Boukriba is a doctor-hopsital university assistant of radiology in La Rabta hospital, Tunisia. His research interests include oncology, neuroradiology and medical technology.

Mounir Sayadi

Mounir Sayadi is a Professor at ENSIT- University of Tunis and head of the research laboratory Signal Image and Energy Mastery (SIME) - University of Tunis. His research interests are focused on adaptive signal processing and filtering, medical image and texture classification and segmentation.