A new technique can reliably detect pneumonia in children using only a microphone and simple computer.
Researchers from the University of Queensland (UQ; Brisbane, Australia) recorded hundreds of coughs from 91 pediatric patients suspected of acute respiratory illness such as pneumonia, bronchiolitis, and asthma at Sardjito Hospital (Yogyakarta, Indonesia), as well as healthy controls. Working with pediatric respiratory technicians, they then extracted features such as non-Gaussianity and Mel Cepstra from cough sounds and used a logistic regression classifier to identify the coughs as either pneumonic or nonpneumonic, and trained a computer algorithm to recognize the difference.
The resulting audio-only test can separate pneumonia from other diseases at a sensitivity of 94% and a specificity of 75%, based on parameters extracted from cough sounds alone. The inclusion of other simple measurements such as the presence of fever further increased the performance. Portable and cheap, the technology could improve healthcare for children in poor, remote areas. Since it requires small processing power, it could serve as an ideal app for a smartphone, which already comes equipped with the requisite microphone and data storage. The study describing the method was published ahead of print on June 7, 2013, in Annals of Biomedical Engineering.
“Our results indicate the feasibility of taking a cough-centered approach to the diagnosis of childhood pneumonia in resource-poor regions,” said lead author Udantha Abeyratne, PhD, of the UQ school of information technology and electrical engineering. “The technology, in its simplest version, will require between five and 10 cough sounds and will automatically and immediately provide a diagnosis without requiring physical contact with patients.”
Pneumonia annually kills over 1,800,000 children annually worldwide, according to the World Health Organization (WHO; Geneva, Switzerland); the vast majority of these deaths occur in resource poor regions such as the sub-Saharan Africa and remote Asia. Prompt diagnosis and proper treatment are essential to prevent these unnecessary deaths, but is fraught with difficulties arising from the lack of field-deployable imaging and laboratory facilities as well as the scarcity of trained community healthcare workers.
University of Queensland
World Health Organization