Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
6950634 | Biomedical Signal Processing and Control | 2018 | 10 Pages |
Abstract
This paper presents a new hardware architecture for real-time wheezing detection in pulmonary sounds. The proposed architecture is based on Mel-frequency cepstral coefficients (MFCC) and artificial neural network (ANN). It has been implemented on field programmable gate array (FPGA) chip using a rapid prototyping design tool, Xilinx system generator (XSG), that provides a high-level of abstraction. The proposed architecture has been optimized, in terms of the required resources, to fit in a low cost FPGA chip. It has been also pipelined to ensure real-time performance. Resources utilization, maximum operating frequency and total power consumption obtained for an Artix-7 FPGA chip are presented and discussed. The wheezing detection/classifications rates of the proposed architecture have been evaluated using normal and wheezing respiratory sounds. Sensitivity, specificity, performance and accuracy obtained by the proposed hardware architecture are almost the same as those obtained by MATLAB software.
Related Topics
Physical Sciences and Engineering
Computer Science
Signal Processing
Authors
Mohammed Bahoura,