RESUMO
The use of machine learning in biomedical research has surged in recent years thanks to advances in devices and artificial intelligence. Our aim is to expand this body of knowledge by applying machine learning to pulmonary auscultation signals. Despite improvements in digital stethoscopes and attempts to find synergy between them and artificial intelligence, solutions for their use in clinical settings remain scarce. Physicians continue to infer initial diagnoses with less sophisticated means, resulting in low accuracy, leading to suboptimal patient care. To arrive at a correct preliminary diagnosis, the auscultation diagnostics need to be of high accuracy. Due to the large number of auscultations performed, data availability opens up opportunities for more effective sound analysis. In this study, digital 6-channel auscultations of 45 patients were used in various machine learning scenarios, with the aim of distinguishing between normal and abnormal pulmonary sounds. Audio features (such as fundamental frequencies F0-4, loudness, HNR, DFA, as well as descriptive statistics of log energy, RMS and MFCC) were extracted using the Python library Surfboard. Windowing, feature aggregation, and concatenation strategies were used to prepare data for machine learning algorithms in unsupervised (fair-cut forest, outlier forest) and supervised (random forest, regularized logistic regression) settings. The evaluation was carried out using 9-fold stratified cross-validation repeated 30 times. Decision fusion by averaging the outputs for a subject was also tested and found to be helpful. Supervised models showed a consistent advantage over unsupervised ones, with random forest achieving a mean AUC ROC of 0.691 (accuracy 71.11%, Kappa 0.416, F1-score 0.675) in side-based detection and a mean AUC ROC of 0.721 (accuracy 68.89%, Kappa 0.371, F1-score 0.650) in patient-based detection.
Assuntos
Inteligência Artificial , Auscultação , Humanos , Auscultação/métodos , Algoritmos , Aprendizado de Máquina , PulmãoRESUMO
Background and objectives: As the prevalence of obesity is increasing in a population, diagnostics becomes more problematic. Our aim was to compare the 3M Littmann 3200 Electronic Stethoscope and 3M Littman Cardiology III Mechanical Stethoscope in the auscultation of obese patients. Methods. A total of 30 patients with body mass index >30 kg/m² were auscultated by a cardiologist and a resident physician: 15 patients by one cardiologist and one resident and 15 patients by another cardiologist and resident using both stethoscopes. In total, 960 auscultation data points were verified by an echocardiogram. Sensitivity and specificity data were calculated. Results. Sensitivity for regurgitation with valves combined was higher when the electronic stethoscope was used by the cardiologist (60.0% vs. 40.9%, p = 0.0002) and the resident physician (62.1% vs. 51.5%, p = 0.016); this was also the same when stenoses were added (59.4% vs. 40.6%, p = 0.0002, and 60.9% vs. 50.7%, p = 0.016, respectively). For any lesion, there were no significant differences in specificity between the electronic and acoustic stethoscopes for the cardiologist (92.4% vs. 94.2%) and the resident physician (93.6% vs. 94.7%). The detailed analysis by valve showed one significant difference in regurgitation at the mitral valve for the cardiologist (80.0% vs. 56.0%, p = 0.031). No significant difference in specificity between the stethoscopes was found when all lesions, valves and both physicians were combined (93.0% vs. 94.4%, p = 0.30), but the electronic stethoscope had higher sensitivity than the acoustic (60.1% vs. 45.7%, p < 0.0001). The analysis when severity of the abnormality was considered confirmed these results. Conclusions. There is an indication of increased sensitivity using the electronic stethoscope. Specificity was high using the electronic and acoustic stethoscope.