Sadish Kumar Thangavelu1, Nehru Kasthuri2, Vivekanandan Sundaram3, Navneet Aravind3, Nishant Bilakanti3. 1. 1 Department of Electronics & Communication Engineering, Nandha Engineering College , Erode, Tamilnadu, India . 2. 2 Department of Electronics & Communication Engineering, Kongu Engineering College , Erode, Tamilnadu, India . 3. 3 School of Electrical Engineering, VIT University , Vellore, Tamilnadu, India .
Abstract
BACKGROUND: In low-income countries with limited infrastructure, remote diagnoses are a sustainable solution to healthcare issues. Current electroencephalography (EEG) monitoring devices are very expensive and may not be feasible in rural areas. Because a seizure is nonstationary, composed of multiple frequencies, and difficult to acquire, visual and conventional frequency-based methods have limited application. MATERIALS AND METHODS: One hundred healthy individuals at their best possible conscious state and 25 patients with an epileptic seizure a few hours prior to the time of recording were chosen. The recordings were carried at Shri Preethi Hospital and Nanda Engineering College, Tamilnadu, India, during the period of June 2014-December 2014. The primary objective was to differentiate the diseased from normal individuals, as well as quantifying the EEG signal and diagnosing remotely. RESULTS: Using wavelet coefficients, the separation between seizure and nonseizure states was measured and compared with that of healthy individuals of different age groups. The systems performance was measured by parameters such as sensitivity, specificity, region of convergence, and precision rate. CONCLUSIONS: The novelty of this article lies in the design, development, and clinical analysis of a low-cost EEG device for remote diagnosis.
BACKGROUND: In low-income countries with limited infrastructure, remote diagnoses are a sustainable solution to healthcare issues. Current electroencephalography (EEG) monitoring devices are very expensive and may not be feasible in rural areas. Because a seizure is nonstationary, composed of multiple frequencies, and difficult to acquire, visual and conventional frequency-based methods have limited application. MATERIALS AND METHODS: One hundred healthy individuals at their best possible conscious state and 25 patients with an epilepticseizure a few hours prior to the time of recording were chosen. The recordings were carried at Shri Preethi Hospital and Nanda Engineering College, Tamilnadu, India, during the period of June 2014-December 2014. The primary objective was to differentiate the diseased from normal individuals, as well as quantifying the EEG signal and diagnosing remotely. RESULTS: Using wavelet coefficients, the separation between seizure and nonseizure states was measured and compared with that of healthy individuals of different age groups. The systems performance was measured by parameters such as sensitivity, specificity, region of convergence, and precision rate. CONCLUSIONS: The novelty of this article lies in the design, development, and clinical analysis of a low-cost EEG device for remote diagnosis.