• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.11a
• /
• pp.605-607
• /
• 2023

Recent research has highlighted the effectiveness of Deep Learning (DL) techniques in automating the detection of lung sound anomalies. However, the available lung sound datasets often suffer from limitations in both size and balance, prompting DL methods to employ data preprocessing such as augmentation and transfer learning techniques. These strategies, while valuable, contribute to the increased complexity of DL models and necessitate substantial training memory. In this study, we proposed a streamlined and lightweight DL method but effectively detects lung sound anomalies from small and imbalanced dataset. The utilization of 1D dilated convolutional neural networks enhances sensitivity to lung sound anomalies by efficiently capturing deep temporal features and small variations. We conducted a comprehensive evaluation of the ICBHI dataset and achieved a notable improvement over state-of-the-art results, increasing the average score of sensitivity and specificity metrics by 2.7%.

• Proceedings of the IEEK Conference
• /
• 2003.07e
• /
• pp.2156-2159
• /
• 2003

Thoracic sound has been widely known as a good method to examine thoracic disease. But, it's difficult to diagnose with correct data according to patient's thoracic position from same patient who has thoracic disease. Therefore, it is necessary to normalize the data for lung sound objectively In this paper, we'd like to detect a useful data for medical examination by applying PCA(Principal Component Analysis) to thoracic sound data and then present a objective data about lung and heart sound for thoracic disease.

• The Journal of the Acoustical Society of Korea
• /
• v.38 no.5
• /
• pp.568-573
• /
• 2019

In the diagnosis of pulmonary diseases, auscultation technique is simpler than the other methods, and lung sounds can be used for predicting the types of pulmonary diseases as well as identifying patients with pulmonary diseases. Therefore, in this paper, we identify patients with pulmonary diseases and classify lung sounds according to their sound characteristics using various convolutional neural networks, and compare the classification performance of each neural network method. First, lung sounds over affected areas of the chest with pulmonary diseases are collected by using a single-channel lung sound recording device, and spectral features are extracted from the collected sounds in time domain and applied to each neural network. As classification methods, we use general, parallel, and residual convolutional neural network, and compare lung sound classification performance of each neural network through experiments.

• Journal of Biomedical Engineering Research
• /
• v.15 no.2
• /
• pp.159-166
• /
• 1994

A new pattern classification algorithm using cepstrum to analyze lung sounds for the classification of pattern with pulmonary and bronchial disorders is proposed. To evaluate the perfomance of the proposed method, the results are compared to the pattern classification with the AR modeling method. In the experiment lung sounds recorded for the training of physician used. As a results, the accuracy of the cepstrum classification is 92.3 % and AR modeling is the 53.8 %, therefore cepstrum modeling method has very high performance than AR and it turned out to be a very efficient algorithm.

• Journal of Information Processing Systems
• /
• v.15 no.5
• /
• pp.1068-1081
• /
• 2019

Signal complexity is one point of view to analyze the biological signal. It arises as a result of the physiological signal produced by biological systems. Signal complexity can be used as a method in extracting the feature for a biological signal to differentiate a pathological signal from a normal signal. In this research, Hjorth descriptors, one of the signal complexity measurement techniques, were measured on signal sub-band as the features for lung sounds classification. Lung sound signal was decomposed using two wavelet analyses: discrete wavelet transform (DWT) and wavelet packet decomposition (WPD). Meanwhile, multi-layer perceptron and N-fold cross-validation were used in the classification stage. Using DWT, the highest accuracy was obtained at 97.98%, while using WPD, the highest one was found at 98.99%. This result was found better than the multi-scale Hjorth descriptor as in previous studies.

• ETRI Journal
• /
• v.37 no.4
• /
• pp.824-832
• /
• 2015

Heart sounds are the main obstacle in lung sound analysis. To tackle this obstacle, we propose a diagnosis algorithm that uses singular spectrum analysis (SSA) and frequency features of heart and lung sounds. In particular, we introduce a frequency coefficient that shows the frequency difference between heart and lung sounds. The proposed algorithm is applied to a synthetic mixture of heart and lung sounds. The results show that heart sounds can be extracted successfully and localizations for the first and second heart sounds are remarkably performed. An error analysis of the localization results shows that the proposed algorithm has fewer errors compared to the SSA method, which is one of the most powerful methods in the localization of heart sounds. The presented algorithm is also applied in the cases of recorded respiratory sounds from the chest walls of five healthy subjects. The efficiency of the algorithm in extracting heart sounds from the recorded breathing sounds is verified with power spectral density evaluations and listening. Most studies have used only normal respiratory sounds, whereas we additionally use abnormal breathing sounds to validate the strength of our achievements.

• Journal of Biomedical Engineering Research
• /
• v.10 no.2
• /
• pp.151-156
• /
• 1989

Lung sound analyer which can provide an objective diagnosis of patients with pulmonary and bronchial disorders is designed. For the purpose of power spectrum analysis, adaptive digital filtering technique and TM - S320C25 DSP chip is used. As a results, adaptive lattice Wiener filter could eliminate heart sounds with a few of 10th order and on the distribution of power spectrum each patterns has shown in normal vescicular breathy from 100 Hz to 200 Hz, in crackle sound from 100 Hz to 400 Hz, in wheeze sound from 150 Hz to 600 Hz.

• The Journal of Korean Medicine
• /
• v.32 no.6
• /
• pp.10-17
• /
• 2011

Background: Vibration response imaging (VRI) is a new technology that records energy generated by airflow during the respiration cycle. Analysis of lung sound using VRI may overcome the limitations of auscultation. Objectives: To set a VRI standard for healthy Koreans, we conducted a clinical assessment to evaluate breath sound images and quantification in healthy subjects and compared the findings with reported breath sound characteristics. Methods: Recordings were performed using the VRIxp. Eighty subjects took a deep breath four times during a 12-second interval while sitting upright. The quantitative aspect was analyzed using the VRI quantitative lung data (QLD) for total left lung, total right lung and for six lung regions: left upper lung (LUL), left middle lung (LML), left lower lung (LLL), right upper lung (RUL), right middle lung (RML), right lower lung (RLL). The qualitative aspect was provided through image assessments by three reviewers. Results: In all regions the left lung had significantly higher QLD than the right lung (P<0.005, paired t-test). The inter-rater agreement was 0.78. 84% of the images were found normal by the final assessment. Among the 16% (n=13) of images with abnormal final assessment, the most common flawed features were dynamic image (77%, n=10) and maximum energy frame (MEF) shape (77%, n=10). No significant differences were found between males and females for QLD but there were significant differences in qualitative aspects including dynamic images, MEF shape, and missing LLL. Conclusion: The characteristics of healthy Koreans are similar to those of Western subjects reported previously. VRI is easy to use and objective, and so is helpful to diagnose patients with respiratory diseases and to monitor the progress of diseases after medical treatments.

• Journal of Chest Surgery
• /
• v.11 no.1
• /
• pp.92-96
• /
• 1978

Congenital diaphragmatic eventration is a rare disease and generally accepted as an abnormally high position of part or all of the diaphragm, usually associated with a marked decrease in muscle fibers and a membranous appearance of the abnormal area. There were 4 cases of the congenital diaphragmatic eventration at the Dept. of Thoracic Surgery, Seoul National University Hospital, from 1957 to 1977. They were two boys and two girls and ranging from 1 day to 3 years of age. They were all repaired by surgical operation and one was expired postoperatively, another one was dead one year later due to complication. The ratio between right and left was 1:3 and their symptoms were cyanosis, dyspnea and frequent respiratory disease. In physical examination there was noted decreased breathing sound on the affected lung field and bowel sound was audible in some cases. Diagnosis was done by Chest X-ray and plication of the affected diaphragm was usually done in operation. There were noted atelectasis and cystic change of the affected side lung. And the liver, colon, spleen and small intestine were found in the dome of the eventrated diaphragm.

• Journal of Physiology & Pathology in Korean Medicine
• /
• v.20 no.2
• /
• pp.320-327
• /
• 2006

This study was written in order to help understanding of listening diagnosis to vocal sound and speech. The purpose of listening diagnosis is that we know states of essence(精), Qi(氣) and spirit(神). Vocal sound and speech are made by Qi and spirit. Vocal sound originates from the center of the abdominal region(丹田) and comes out through vocal organs, for example lung, larynx, nose, tongue, tooth, lip and so on. Speech is expressed by vocal sound and spirit. They are controled by the Five Vital organs(五臟). Various changes of vocal sound and speech observe the rules of yinyang. For example, if we consider patient likes to say or not, we can diagnose heat and coldness of illness. If we consider he speaks loudly or quietly, we can diagnose weak and severe of illness. If we consider he speaks clearly or thick, we can diagnose inside and outside of illness. If we consider he speaks damp or dry, we can diagnose yin and yang of illness. If we consider change of voice, we can diagnose new and old illness. Symptoms of changes of five voices, five sounds, dumbness and huskiness are due to abnormal vocal sound, and symptoms of changes of mad talk, mumble, sleep talking and so on are due to abnormal speech.