Mobile phone software accurately diagnoses respiratory diseases

According to a recent study published in the open-access journal Respiratory Research, scientists combine machine learning with clinical medicine to enable automated analytics embedded in a smartphone app (APP) to diagnose respiratory diseases with high accuracy, which is not easy for human doctors.the smartphone APP, presented by researchers at Curtin University and the University of Queensland in Australia, was highly accurate (81-97 per cent) in diagnosing asthma, howling, pneumonia, lower respiratory diseases and bronchitis. To develop the app, scientists used techniques similar to speech recognition to train them to recognize typical cough characteristics of five different respiratory diseases. They then used app to classify the coughing voices of 585 children aged between 29 and 12 who were treated at two hospitals in Western Australia. Accuracy is assessed by comparing the diagnosis given by app with the diagnosis of the hospital, which was obtained by a team of pediatricians after looking at the imaging results, laboratory results, hospital records, and all available clinical findings.researchers say it's not easy to tell exactly what kind of respiratory disease a child has, even for experienced doctors. This study demonstrates that new technologies, mathematical concepts, machine learning, and clinical medicine can be successfully combined to produce a new diagnostic test that combines multi-sphere knowledge.currently, the APP can provide a diagnosis without the need for a doctor's personal consultation, addressing a major limitation of existing telemedicor consultations that can be used to provide telemedic services or for different levels of health service providers. And once you don't rely on your doctor for a clinical examination, you can start targeted treatment earlier. However, the team still recommends consulting a clinician as much as possible while using the APP to maximize diagnostic accuracy. (Science and Technology Daily)