Research of rectal dynamic function diagnosis based on FastICA-STFT

• Published in: IET science, measurement & technology Vol. 12; no. 8; pp. 965 - 969
• Main Authors: Zan, Peng, Liu, Yankai, Chang, Meihan
• Format: Journal Article
• Language: English
• Published: The Institution of Engineering and Technology 01.11.2018
• Subjects: 15th‐order Newton; artificial anal sphincter implantation; artificial anal sphincter implants; blind source separation; clinical diagnosis; complete diagnostic experiment; complete rectal dynamic function diagnosis process; current rectal signal research; C‐C phase space reconstruction technology; different gut dynamic; different rectal signals; FastICA‐STFT; fault diagnosis; Fourier transforms; incomplete problem; independent component analysis; independent component analysis algorithm; multidimensional signal reconstruction; Newton method; nonstationary signals; nonstationary simulation signal; patient diagnosis; preprocessed rectal signal component; preprocessing method; rectal dynamic function experiment; rectal motility; Research Article; short‐time Fourier; signal reconstruction; single channel rectal pressure signal; time‐frequency analysis; wavelet transforms; clinical diagnosis; Fourier transforms; fault diagnosis; wavelet transforms; independent component analysis algorithm; artificial anal sphincter implantation; single channel rectal pressure signal; preprocessed rectal signal component; different gut dynamic; nonstationary signals; independent component analysis; different rectal signals; complete diagnostic experiment; time-frequency analysis; signal reconstruction; time–frequency analysis; preprocessing method; incomplete problem; artificial anal sphincter implants; Newton method; blind source separation; current rectal signal research; rectal dynamic function experiment; nonstationary simulation signal; FastICA-STFT; short-time Fourier; complete rectal dynamic function diagnosis process; C-C phase space reconstruction technology; multidimensional signal reconstruction; rectal motility; 15th-order Newton; patient diagnosis
• ISSN: 1751-8822; 1751-8830; 1751-8830
• DOI: 10.1049/iet-smt.2018.5179

Aiming at the lack of signal preprocessing in current rectal signal research, a new 15th-order Newton iterative fast independent component analysis (FastICA) algorithm is proposed to separate the signal. To solve the incomplete problem of the single channel rectal pressure signal, C-C phase space reconstruction technology is used for multi-dimensional signal reconstruction. In order to achieve the diagnosis of rectal dynamic after artificial anal sphincter implantation, the preprocessed rectal signal component is subjected to short-time Fourier transforms for time–frequency analysis. This constitutes a complete rectal dynamic function diagnosis process. The non-stationary simulation signal is constructed to verify the preprocessing method. A complete diagnostic experiment is performed on two rectal signals of different gut dynamic. Simulation results show that the proposed method can realise the preprocessing of non-stationary signals. Through the rectal dynamic function experiment based FastICA-STFT, two different rectal signals are distinguished, and the results are consistent with the clinical diagnosis, which has achieved the diagnosis of rectal motility. This study will provide support for the recovery of rectal dynamic function in patients with artificial anal sphincter implants.