In the field of medicine, with the introduction of computer systems that can collect and analyze massive amounts of data, many non-invasive diagnostic methods are being developed for a variety of conditions. In this study, our aim is to develop a non-invasive method of classifying respiratory sounds that are recorded by an electronic stethoscope and the audio recording software that uses various machine learning algorithms. In order to store respiratory sounds on a computer, we developed a cost-effective and easy-to-use electronic stethoscope that can be used with any device. Using this device, we recorded 17,930 lung sounds from 1630 subjects. We employed two types of machine learning algorithms; mel frequency cepstral coefficient (MFCC) features in a support vector machine (SVM) and spectrogram images in the convolutional neural network (CNN). Since using MFCC features with a SVM algorithm is a generally accepted classification method for audio, we utilized its results to benchmark the CNN algorithm. We prepared four data sets for each CNN and SVM algorithm to classify respiratory audio: (1) healthy versus pathological classification; (2) rale, rhonchus, and normal sound classification; (3) singular respiratory sound type classification; and (4) audio type classification with all sound types. Accuracy results of the experiments were; (1) CNN 86%, SVM 86%, (2) CNN 76%, SVM 75%, (3) CNN 80%, SVM 80%, and (4) CNN 62%, SVM 62%, respectively. As a result, we found out that spectrogram image classification with CNN algorithm works as well as the SVM algorithm, and given the large amount of data, CNN and SVM machine learning algorithms can accurately classify and pre-diagnose respiratory audio.

Objective: To investigate the value of C-reactive protein (CRP) as a marker of chronic obstructive pulmonary disease (COPD) exacerbations or specifically bacterial exacerbations and to evaluate a correlation between raised CRP levels and other markers of inflammation in patients with an acute exacerbation (AECOPD). Subjects and Methods: The medical records of patients with AECOPD were retrospectively analyzed. They were categorized according to the nature of sputum as mucoid or purulent and to the findings on chest radiographs as with pneumonia (PCOPD) or without pneumonia. Stable COPD (SCOPD) patients and a group of asymptomatic nonsmokers were also included in the study. Results: All COPD patients (SCOPD: 30; AECOPD: 51; PCOPD: 32) and control subjects (30) were male. The mean CRP levels and WBC counts of the groups were PCOPD: 108.1 ± 61.8 mg/l and 13.7 ± 6.8 × 10⁹/l; AECOPD: 36.8 ± 43.9 mg/l and 11.4 ± 4.8 × 10⁹/l; SCOPD: 3.9 ± 1.4 mg/l and 7.9 ± 1.9 × 10⁹/l; control: 2.1 ± 0.9 mg/l and 7.7 ± 1.1 × 10⁹/l. The mean CRP level of AECOPD was statistically different from those of PCOPD and SCOPD (p = 0.0001, p = 0.002, respectively). The sensitivity and specificity of CRP to determine an acute exacerbation were 72.5 and 100%, respectively. Among the patients with AECOPD, 25 had purulent sputum and a mean CRP level of 46.4 ± 48.6 mg/l, which is significantly higher than the CRP level (28.0 ± 44.5 mg/l) of the 18 patients with mucoid expectoration (p = 0.015). Among the mucoid-expectorating subgroup, the patients with leukocytosis had significantly higher CRP levels than the patients without leukocytosis (p = 0.034). Conclusion: A high serum CRP value may indicate an infectious exacerbation in COPD patients and it correlates with sputum purulence and increased serum WBC counts.

No abstract

Robotic-assisted prostatectomy in the steep Trendelenburg position led to an increase in upper airway resistance directly after surgery that normalized within 24 h. The development of chemosis can be indicative of increased upper airway resistance. In patients without COPD, VC and FEV1 were reduced after surgery and recovered within 5 days, while in patients with COPD, the alteration lasted beyond 5 days.

No abstract