Diagnosing silicosis: could a new AI-powered breath test lead to mass screening?

A recent study has highlighted the ongoing development of a non-invasive rapid breath test which can potentially diagnose silicosis with the aid of artificial intelligence (AI) in less than 2 minutes.

Summary

The study, Rapid, non-invasive breath analysis for enhancing detection of silicosis using mass spectrometry and interpretable machine learning, published in the Journal of Breath Research, reports that the new test is over 90% accurate in differentiating silicosis patients from healthy individuals, which is reported to be better than traditional lung function tests.

Whilst the study acknowledges that further research needs to be undertaken before the test could be utilised in real world scenarios, the implementation of a diagnostic tool such as this could significantly increase the number of diagnosed silica cases in the UK and in turn result in an increase in claims for personal injury.

Conversely, the early diagnosis of silicosis cases could prevent further exposure to at risk workers, thereby reducing the severity of injury and in turn the claims potential.

Background

In Merryn J Baker et al 2025, it is noted that traditional methods of screening for silicosis, includes the use of questionnaires, spirometry and chest radiographs. However, whilst chest x-rays are the traditional mainstay of diagnosis, they are insensitive for detecting disease in early stages as there is a delay between histopathological onset and radiographically viable lesions. Similarly, spirometry has shown limited sensitivity when detecting abnormalities in lung function in early-stage disease and even in severe disease results may fall within reference ranges.

The study notes:

“Current silicosis surveillance methods rely on progression of lung disease to the stage that this is visible in imaging techniques or has significantly restrictive or obstructive lung function below the lower limit of normal. This reduces the likelihood of OLDs being detected in the very early stages where disease progression can more effectively be slowed, highlighting the need for new sensitive detection methods.”

Human breath contains hundreds of volatile organic compounds (VOCs). The composition of these molecules changes in response to physiological conditions like disease. However, VOCs are often present in extremely low concentrations and therefore the authors of the study indicate that they have developed tools, which include AI machine learning, which analyse breath samples for these low concentrations in order to more accurately identify and diagnose cases of silicosis:

“The presence of disease changes metabolic processes which alter the chemical profile of breath [20]. When crystalline silica is deposited in the lungs, macrophages ingesting the dust particles cause an inflammatory response which releases potential biomarkers, altering the abundances and identities of aldehydes, alkanes, and other VOCs in breath [29–32]. Similarly, pulmonary fibrosis has been shown to increase the concentrations of a range of chemicals (carbon monoxide, nitric oxide, protein, 3-nitrotyrosine and 8-isoprostane) in exhaled breath condensate [33] and in exhaled breath [34]. Such 'breathprints' may enable the detection of early disease in individual workers; however, the concentrations of such chemicals are low, which highlights the need for ultra-sensitive methods in chemical analysis.”

The study concludes:

“This study provides early evidence for the potential of exhaled breath analysis using APCI-MS as a rapid, non-invasive diagnostic tool for silicosis. The VOC measurement takes less than two minutes per sample and does not require a preconcentration sample processing step, making it a promising candidate for large population screening, offering a significant advantage over methods that are invasive and time-consuming particularly in at-risk occupational groups." [emphasis added]

It is noteworthy that the study acknowledges its limitations which includes a relatively small sample size and the need for further validation of the findings using higher-resolution mass spectrometry and ion fragmentation data, supported by authentic standards.

What this means for Defendants and Insurers

As we have previously reported, there has been increasing concern and media coverage, fuelled in part by a spike of silicosis cases in Australia associated with engineered stone, that a resurgence of silicosis cases will occur in the UK.

However, other than a relatively modest (although arguably significant) spike in reported silicosis cases connected with acute exposure to engineered stone, we have not seen any significant increase in claims numbers.

The potential introduction of mass screening, which can more readily identify early-stage disease progression and potentially asymptomatic silicosis cases, could in turn lead to an increase in claims numbers.

It is noteworthy that in response to an “increase in reported silica cases” Irwin Mitchell has launched a “National Register for Stoneworkers” to record incidents of contact with RCS. It seems entirely feasible that claimant firms and claims management companies could offer mass screenings of workers for silicosis, should inexpensive and rapid breath testing become readily available.

Clyde & Co are specialists in dealing with disease claims, including silicosis, and we closely monitor developments around this topic. For more on this subject, you can read all of our previous articles here, and if you have any questions about this topic you can contact Edward Sainsbury or any of our Occupational Disease and Legacy Claims team.