News briefs

Automated cough analysis technology in a smartphone app could help diagnose childhood respiratory disorders, according to an Australian study published in Respiratory Research. Researchers at Curtin University and the University of Queensland found that their smartphone app had high accuracy (81–97%) for diagnosing asthma, croup, pneumonia, lower respiratory tract disease, and bronchiolitis. The authors used similar technology to that used for speech recognition software, which they trained by a machine learning approach to recognise features of coughs characteristic for five different respiratory diseases. The researchers then used the app to categorise the coughs of 585 children aged between 29 days and 12 years at two hospitals in Western Australia. The accuracy of the automated cough analyser was determined by comparing its diagnosis with that of a panel of paediatricians after they had reviewed the results of imaging, laboratory findings, and hospital charts, and undertaken all appropriate clinical investigations. Both positive agreement (both the app and the panel diagnosed the disorder) and the negative agreement values (neither the app nor the panel diagnosed the disorder) exceeded 80% for all orders, and were particularly good for asthma (97% positive agreement, 91% negative agreement). The authors noted that the technology can provide a diagnosis without the need for clinical examination by a doctor, addressing a major limiting feature of existing telehealth consultations for providing clinical services remotely. Removing the need for a clinical examination may allow targeted treatments to begin sooner. Dr Porter said: “As the tool does not rely on clinical investigations, it can be used by health care providers of all levels of training and expertise. However, we would advise that where possible the tool should be used in conjunction with a clinician to maximise the clinical accuracy.”