4/19/2024 0 Comments Resonance lung sounds![]() ![]() Furthermore, recent advances in signal processing and artificial intelligence assist clinicians in decision making when diagnosing respiratory diseases through lung sounds. It has the ability to store lung sounds as signals within a computer allowing medical doctors to investigate these signals in time-frequency analysis with a better interpretation (Shi et al. Therefore, electronic stethoscope has been gradually arising as a replacement to traditional diagnosis tools. In addition, the diagnosis of pulmonary diseases is usually affected by the quality of the tool, physician experience, and surrounding environment (Shi et al. Thus, the auscultation process loses important information, carried by lower frequency waves, about the condition of respiratory organs. Using the traditional manual stethoscope, many diseases could be misdiagnosed or go undetected due to inability of hearing its corresponding respiratory sounds. However, the human ear is sensitive to waves of 20 Hz to 20 kHz (Rosen and Howell 2011). In general, lung sounds are acoustic signals with frequencies ranging between 100 Hz and 2 kHz (Gross et al. Other respiratory sounds include coughing, snoring, and squawking. These sounds are an indication of pneumonia or heart failure (Reichert et al. Crackles are discontinuous clicking or rattling sounds of either fine (short duration) or coarse (long duration). They usually arise from fluid or mucus filling up the bronchial tubes (Sovijarvi et al. Rhonchi are low-pitch continuous waves of sounds similar to snoring with frequencies less than 200 Hz. They usually originate due to laryngeal or tracheal stenosis (Pasterkamp et al. Similarly, stridor sounds are high-pitched waves of more than 500 Hz lasting for over 250 ms. These sounds are due to an inflammation/narrowing of the bronchial tubes (Pramono et al. Wheezes are considered as high-pitch continuous waves of more than 400 Hz lasting for more than 80 ms and sounding like a breathing whistle. There are several types of abnormal (adventitious) lung sounds that superimpose normal sounds including wheezes, stridor, rhonchi, and crackles (Sarkar et al. An abnormality in the auscultated sound usually indicates an inflammation, infection, obstruction, or fluid in the lungs. Lung sounds are either normal or abnormal. In addition, chronic obstructive pulmonary disease (COPD) is expected to be the third leading cause of death by 2030 (World Health Organization 2017b). According to the world health organization (WHO) report in 2017 (World Health Organization 2017a), more than 235 million people are suffering from asthma worldwide. Through a stethoscope, the sound of air moving inside and outside the lungs during breathing can be auscultated through chest walls allowing a physiotherapist to identify any pulmonary diseases such as asthma, pneumonia, or bronchiectasis (BRON) (Andrès et al. ![]() It is considered as a safe, non-invasive, and cost-effective clinical method to monitor the overall condition of the lungs and surrounding respiratory organs (Bardou et al. Pulmonary auscultation is one of the oldest techniques used in the diagnosis of the respiratory system. This study paves the way towards implementing deep learning models in clinical settings to assist clinicians in decision making related to the recognition of pulmonary diseases. Furthermore, a total agreement of 98.26% was obtained between the predictions and original classes within the training scheme. The developed algorithm achieved the highest average accuracy of 99.62% with a precision of 98.85% in classifying patients based on the pulmonary disease types using CNN + BDLSTM. The training of the model was evaluated based on a k-fold cross-validation scheme of tenfolds using several performance evaluation metrics including Cohen’s kappa, accuracy, sensitivity, specificity, precision, and F1-score. The deep learning network architecture consisted of two stages convolutional neural networks and bidirectional long short-term memory units. ![]() These steps included wavelet smoothing, displacement artifact removal, and z-score normalization. Then, several preprocessing steps were undertaken to ensure smoother and less noisy signals. Initially, all signals were checked to have a sampling frequency of 4 kHz and segmented into 5 s segments. ![]() on Biomedical Health Informatics publicly available challenge database. In addition, 110 patients data were added to the data-set from the Int. The selected data-set included a total of 103 patients obtained from locally recorded stethoscope lung sounds acquired at King Abdullah University Hospital, Jordan University of Science and Technology, Jordan. In this paper, a study is conducted to explore the ability of deep learning in recognizing pulmonary diseases from electronically recorded lung sounds. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |