AI-aided home stethescope promising and straightforward in diagnosing asthma exacerbations

According to a study published in the Annals of Family Medicine, researchers have concluded that AI-aided home stethoscope offers reliable insights into asthma flare-ups, particularly for young children. Its implementation in healthcare may significantly improve the detection of asthma exacerbations and simplify remote patient monitoring.

New medical devices enable asthma patients to monitor their condition at home, offering a more comprehensive understanding of their disease than intermittent clinic visits. The question arises here: "Which devices and parameters excel in detecting exacerbations?"

In a 6-month observational study, 149 asthma patients (90 children, 59 adults) used three devices daily for the first two weeks, then weekly for 5.5 months, with increased frequency during exacerbations. They used an AI-aided home stethoscope, a peripheral capillary oxygen saturation meter, and a peak expiratory flow meter and filled out a health state survey. The resulting 6,029 examinations were evaluated by physicians for exacerbations. Machine learning models were trained for each registered parameter, and the area under the receiver operating characteristic curve (AUC) was calculated to assess their utility in detecting exacerbations.

The best single-parameter discriminators of exacerbations were wheeze intensity for young children (AUC 84%, rhonchi intensity for older children (AUC 81%), and survey answers for adults (AUC 92%). The greatest efficacy was observed for a combination of various parameters.

The current study demonstrates that an AI-aided home stethoscope can detect exacerbations by measuring wheezes, rhonchi, coarse, HR, RR, and I/E. PEF measurements are not necessary, and the stethoscope may simplify asthma diagnosis and monitoring for younger children.

Reference:

Andrzej Emeryk et al. Home Monitoring of Asthma Exacerbations in Children and Adults with Use of an AI-Aided Stethoscope. The Annals of Family Medicine Nov 2023, 21 (6) 517-525; DOI: 10.1370/afm.3039