HHT based lung sound crackle detection and classification
Zhenzhen Li, Minghui Du
Abstract
Zhenzhen Li, Minghui Du
Abstract
Crackles are discontinuous adventitious lung sounds, characterized by waveforms with a rapid onset and short duration. Traditional approaches to detect and classify crackles are mainly from the morphological aspect, however, the sharp patterns of crackles in frequency domain were overlooked. In this paper we employ the innovative HHT method to detect and classify crackles. By detecting peaks in time-frequency distribution derived from HHT, segments are extracted, then, crackles can be identified and classified efficiently. Relevant theories, methods and experimental results are given in detail.
OpenAlex reports 14 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Crackles are discontinuous adventitious lung sounds, characterized by waveforms with a rapid onset and short duration. Traditional approaches to detect and classify crackles are mainly from the morphological aspect, however, the sharp patterns of crackles in frequency domain were overlooked. In this paper we employ the innovative HHT method to detect and classify crackles. By detecting peaks in time-frequency distribution derived from HHT, segments are extracted, then, crackles can be identified and classified efficiently. Relevant theories, methods and experimental results are given in detail.
Key concepts: Crackles, Speech recognition, Computer science, Respiratory sounds, Waveform, Bioacoustics, Auscultation, Pattern recognition (psychology)