Clinical trial begins for AI software to detect lung cancer in unanalyzed X-rays

• This tool, named Optimal XR, aims to prevent potentially malignant lung nodules from continuing to go unnoticed by diagnosing them in a timely manner
• The innovation is part of the Optimal Lung project, a clinical decision support system based on artificial intelligence for the detection of lung nodules, driven by the Parc Taulí University Hospital and the Vall d'Hebron University Hospital, together with the Eurecat technology center.

Parc Taulí started a multicenter clinical trial this July to validate AI software that detects lung cancer in unanalyzed chest X-rays. This tool, named Optimal XR, aims to prevent potentially malignant lung nodules from going unnoticed. Currently, these nodules are not diagnosed in time because the X-rays have not been examined by a radiologist.

The study, led by pulmonologist and emeritus researcher at the Parc Taulí Institute of Research and Innovation (IXNUMXPT) Eduard Monsó, aims to confirm the algorithm's efficacy and safety, as well as validate its use in a real hospital environment and generate the evidence needed to apply for approval for commercialization.

The clinical trial is expected to last sixteen months—ten for recruitment and six for clinical follow-up—with an additional passive follow-up for three years after its completion. During this period, around XNUMX chest X-rays from the Parc Taulí Digital Medical Imaging Center (CIMD) database will be analyzed in real-time. These X-rays are not necessarily related to respiratory diseases and may have been taken for other medical reasons. “The study will consist of using the algorithm to identify lung nodules in these X-rays, but not all of them will contain nodules. The nodules detected by the system will be reviewed by a radiologist and, if confirmed, will be referred to the Pulmonology Service for further evaluation and follow-up,” explains Monsó.

Optimal XR and lung cancer

 Lung cancer is the leading cause of cancer death worldwide. In Catalonia alone, in 2023, nearly 5,000 new cases and more than 3,400 deaths from this disease were recorded. The late appearance of its symptoms means that 80% of lung cancers go unnoticed and are diagnosed in very advanced stages, when the five-year survival rate is already below 20%. Early detection not only significantly increases the survival rate—up to 60%—but also improves the quality of life of those affected.

Currently, early detection of lung cancer is a major challenge for Primary Care, where thousands of chest radiographs are performed annually on a significant proportion of patients as part of standard evaluations or diagnostic procedures that are not initially related to suspicions of this disease. However, due to the growing shortage of specialized radiologists, it is not possible to examine these X-rays as efficiently as desired, with the possible risk of failing to identify cancerous lung nodules in early stages, which are later detected when the disease is already in advanced stages.

In this context, Parc Taulí University Hospital and Vall d'Hebron University Hospital, together with the Eurecat technology center, developed Optimal Lung, a clinical decision support system based on artificial intelligence that integrates two algorithms for the detection of lung nodules: Optimal XR, which focuses on X-rays from Primary Care and emergency services to avoid missing cancerous lung nodules; and Optimal CT, which focuses on computed tomography to cover all levels of lung cancer diagnosis.

Optimal XR is AI software that uses deep learning technology to analyze X-rays. The algorithm processes all X-rays performed at the hospital and identifies those with a high probability of containing lung nodules, based on over a thousand real X-rays it has previously trained with. The detected X-rays are sent directly to a radiologist for review and confirmation of the presence of nodules. This way, the work of radiologists is facilitated, and early and timely detection of possible lung problems that might have gone unnoticed is improved.

The development of this solution involved departments from various clinical specialties such as radiology, pulmonology, and oncology, as well as engineers, information systems experts, and innovation personnel. The ultimate goal is to implement Optimal XR in other hospitals, with a special focus on Primary Care settings where it is now most needed. In the future, according to Monsó, “the team intends to incorporate it in other countries where chest X-ray care or reading is not as developed as it is here.”