Mobile with AI helps to control respiratory diseases in children

A idea, led by scientists from the RISE-Health research unit and the Faculty of Medicine of the University of Porto (FMUP), is based on the conviction that “the wide availability of mobile phones equipped with high quality microphones offers a practical, noninvasive and scalable solution for recording respiratory sounds in clinical and domestic environments”.

This may be “the possibility of eliminating the need for additional devices and allowing data collection through a single family tool”.

To Lusa, researcher Cristina Jácome (RISE-Health/FMUP) added the objective of giving the user, in this case to parents and educators, “more power” through a tool that may even be downloading directly to clinicians’ devices, “helping them in the best health decisions for these children”.

“ At no time do we want to take away the relevance of traditional or electronic stethoscopes, but the smartphone turns out to be a familiar everyday technology. It is a technology that has enormous potential (...). The logic that clinicians have to do everything and take care of our health is part of the past. We have to be active agents and, in this case, parents and even professionals in schools, daycare centers, etc. can play a central role in monitoring children's health and even help to avoid exacerbations in periods of crisis. The logic here is the patient as power, with a tool to help him ", said Cristina Jácome.

According to the FMUP, this is the first study to apply an AI model to detect acute respiratory noise, similar to a whistling or wheezing in the chest, through respiratory sounds recorded by smartphone.

A total of 217 children and adolescents aged 0-17 years participated in the external pediatric consultation of the Local Health Unit of São João, Porto.

Emphasizing the collaboration in the study of researchers from Coimbra who “have a lot of experience in the analysis of respiratory sounds” and helped the team from RISE-Health/FMUP to create an AI algorithm dedicated to this theme, and revealing that an international database of sounds acquired with mobile phones was created, Cristina Jácome said that the next step may be in extending the study to other ULS, as well as testing the idea in more vulnerable or elderly people.

“ Thinking of the life cycle and pathologies such as dementia, there is a whole path to do also in other age groups and other pathologies. Informal caregivers may also be interested in realizing how this works”, exemplified.

Regarding the current results of a study that began in 2021, it has already experienced some difficulties with regard to funding and was published in May of this year in the scientific journal European Journal of Pediatrics, it is known that it integrated 2020 recordings of respiratory sounds via mobile phone and that the developed AI model had an 87% accuracy in the detection of respiratory noises.

Because of this, the researchers concluded that “lung auscultation through the smartphone becomes especially valuable in pediatric care, since it bypasses the barrier of the need for active cooperation of the child”, this being a “technology that enables and scales distance monitoring, acting as a central pillar for future Clinical Decision Support Systems”.

Counting that she was already contacted by professionals from Colombia and Africa, “countries with less organized health services, with less primary health care than Portugal”, described the adherence of the parents with whom she lived and stressed “the potential importance of this idea for the Portuguese National Health Service in monitoring various chronic and acute pathologies”.

“ In a clinical environment, we performed auscultation with 45 parents and adherence was very good. More than 90% said they would be available to use this application in a real environment, they would be available to allow doctors who accompany children to receive these respiratory sounds also to improve clinical decisions ", revealed, highlighting as another objective the increase in health literacy of the population.

Led by Cristina Jácome, the scientific article “Wheeze detection in real-world pediatric care: AI applied to smartphone lung auscultation” “Detection of wheezing in pediatric care in the real world: AI applied to pulmonary auscultation by smartphone”) was attended by Inês Pais-Cunha, Maria Catarina Silva, Henrique Ferreira-Cardoso, João Fonseca and Inês Azevedo, researchers from FMUP and the RISE-Health research unit, together with other national researchers, including the University of Coimbra.