Rhonchi

Abstract: To characterize lung sounds objectively, we examined, by means of time-amplitude plots, selected tape recordings of auscultatory phenomena considered by six observers to be typical of those in a standard classification. Normal lung sounds could not consistently be visually distinguished from adventitious sounds at conventional chart recorder speeds of 100 mm per second or less, but the differentiation was easily achieved when the time scale of the plots was raised to 800 mm per second. When discontinuous sounds (rales, crackles or crepitations) were heard clinically, the time-expanded wave forms showed intermittent "discontinuous" deflections usually less than 10 msec in duration. When continuous sounds (rhonchi or wheezes) were heard, the deflections were usually more than 250 msec. Time-expanded wave-form analysis provides reproducible visual displays that allow documentation of the differentiating features of lung sounds and enhances the diagnostic utility of the sounds. (N Engl J Med 296:968–...

Abstract: Adults presenting to an emergency department with acute respiratory illness were studied prospectively in an effort to identify sensitive clinical criteria for the diagnosis of pneumonia. Of 308 patients studied, 118 (38%) had definite or equivocal infiltrates and were considered to have pneumonia. No single symptom or sign was reliably predictive of pneumonia. Cough was the most common symptom in patients with pneumonia (86%), but was equally common in those with other respiratory illness. Fever was absent in 36 patients with pneumonia (31%). Abnormal findings on lung examination, that is, rales, rhonchi, decreased breath sounds, wheezes, altered fremitus, egophony, and percussion dullness, were each found in fewer than half of the patients with pneumonia. Twenty-six patients (22%) with a completely normal chest examination had pneumonia. Abnormal vital signs (temperature greater than 37.8 degrees C (100 degrees F), pulse greater than 100/min, or respirations greater than 20/min) were 97% sensitive for the detection of pneumonia. These criteria retained their sensitivity when films were subjected to a second, blinded interpretation by a senior radiologist. We conclude that restricting chest roentgenograms to patients with at least one abnormal vital sign will detect almost all radiographically demonstrable pneumonia in adult emergency department patients.

Abstract: Auscultation has been essential part of the physical examination; this is non-invasive, real-time, and very informative. Detection of abnormal respiratory sounds with a stethoscope is important in diagnosing respiratory diseases and providing first aid. However, accurate interpretation of respiratory sounds requires clinician's considerable expertise, so trainees such as interns and residents sometimes misidentify respiratory sounds. To overcome such limitations, we tried to develop an automated classification of breath sounds. We utilized deep learning convolutional neural network (CNN) to categorize 1918 respiratory sounds (normal, crackles, wheezes, rhonchi) recorded in the clinical setting. We developed the predictive model for respiratory sound classification combining pretrained image feature extractor of series, respiratory sound, and CNN classifier. It detected abnormal sounds with an accuracy of 86.5% and the area under the ROC curve (AUC) of 0.93. It further classified abnormal lung sounds into crackles, wheezes, or rhonchi with an overall accuracy of 85.7% and a mean AUC of 0.92. On the other hand, as a result of respiratory sound classification by different groups showed varying degree in terms of accuracy; the overall accuracies were 60.3% for medical students, 53.4% for interns, 68.8% for residents, and 80.1% for fellows. Our deep learning-based classification would be able to complement the inaccuracies of clinicians' auscultation, and it may aid in the rapid diagnosis and appropriate treatment of respiratory diseases.

Abstract: BACKGROUND: To determine whether objectively detected lung sounds were significantly different in patients with pneumonia than those in asymptomatic subjects, and to quantify the pneumonia findings for teaching purposes. METHODS: At a community teaching hospital we used a multi-channel lung sound analyzer to examine a learning sample of 50 patients diagnosed with pneumonia and 50 control subjects. Automated quantification and characterization of the lung sounds commonly recognized to be associated with pneumonia were used to generate an “acoustic pneumonia score.“ These were examined in the learning sample and then prospectively tested in 50 patients and 50 controls. RESULTS: The acoustic pneumonia score averaged 13 in the learning sample and 11 in the test sample of pneumonia patients. The scores were 2 and 3 in the controls. The positive predictive value of a score higher than 6 was 0.94 in the learning sample and 0.87 in the test sample. The sensitivities in the 2 groups were 0.90 and 0.78, and the specificities were 0.94 and 0.88, respectively. Adventitious sounds were more common in pneumonia patients (inspiratory crackles 81% vs 28%, expiratory crackles 65% vs 9%, rhonchi 19% vs 0%). CONCLUSIONS: Our lung sound analyzer found significant differences between lung sounds in patients with pneumonia and in asymptomatic controls. Computerized lung sound analysis can provide objective evidence supporting the diagnosis of pneumonia. We believe that the lung-sound data produced by our device will help to teach physical diagnosis.