Artificial Intelligence can improve lung disease diagnosis

A new study suggests that artificial intelligence (AI) can improve the diagnosis of long-term lung diseases.

Results from the study, presented at the recent European Respiratory Society’s International Congress, are the first to investigate the use of AI in lung disease diagnosis.

The study included data from 968 participants who were undergoing complete lung function testing for the first time – all of which had already received a clinical diagnosis using traditional testing methods, including lung function testing, CT scans and electrocardiograms.

Researchers subsequently investigated whether machine learning could help detect patterns in patient data to spot disease more accurately.

The algorithm made likely diagnosis suggestions based on routine lung function parameters and clinical variables such as smoking history, body mass index and age.

"We have demonstrated that artificial intelligence can provide us with a more accurate diagnosis in this new study,” said Wim Janssens, the senior author of the study from the University of Leuven in Belgium. “The beauty of our development is that the algorithm can simulate the complex reasoning that a clinician uses to give their diagnosis, but in a more standardised and objective way so it removes any bias."

The current testing for lung function is based on a complex set of data taken from spirometry tests (which measure the amount of air and airflow speed during breathing), plethysmography tests (which measure static lung volume and airway resistance), and diffusion tests (which measure the amount of oxygen and other gases crossing the lungs' air sacs).

The study leaders concluded AI resulted in a more accurate and automated interpretation of pulmonary function tests, and thus better disease detection.

Marko Topalovic, first author of the study from the University of Leuven in Belgium. “Not only can this help non-experienced clinicians, but it also has many benefits for healthcare overall. as it is time saving in achieving a final diagnosis as it could decrease the number of redundant additional tests clinicians are taking to confirm the diagnosis."

AI is already being used for healthcare data analysis is areas such as the automated analysis of electrocardiograms.

Various companies are attempting to integrate AI into care decisions around the world. Perhaps the biggest company in the field IBM is lending its Watson Health AI to a number of different projects, the latest being in China to help personalise cancer care.

Similarly, many companies are now investigating the use of AI in the analysis of visual data from medical scans. Imaging Advantage launched a radiology-focused machine learning initiative earlier this year whilst Google’s DeepMind Health is deploying its AI in a collaboration with the UK Moorsfield Eye Hospital NHS Foundation Trust to help analyse optical scans.

The study researchers will now test their algorithm in different populations in order to ‘train’ their AI platform and increase its the decision capabilities.