GMV designs a groundbreaking artificial intelligence-based simulator for diagnosing interstitial lung diseases

Respiratory diseases are among the leading causes of mortality and disability worldwide. Diffuse interstitial lung diseases (DILD), such as idiopathic pulmonary fibrosis (IPF) or sarcoidosis, which cause progressive scarring of the lung tissue, impairing respiratory capacity and the body's ability to supply adequate oxygen, are of particular concern.

Diagnosing these types of pathologies is not easy as medical imaging often shows characteristics that overlap between different diseases, making it difficult to interpret. Furthermore, once the diagnosis has been made, there are currently no factors capable of predicting the course of the disease or its response to treatment. This is important in this type of disease, which has a high morbidity and mortality rate, where correct treatment has been shown to slow the disease, but delayed diagnosis and suboptimal treatment are associated with a worse prognosis.

Detecting this type of disease is not easy, as medical images often show similar signs between different pathologies, which makes analysis challenging. Furthermore, once the diagnosis is confirmed, there are still no tools available to predict how the disease will progress or how it will respond to treatment. This is particularly important because these are serious diseases with a high likelihood of complications or even death. It has been shown that proper treatment can slow their progress, but if diagnosis is delayed or treatment is inadequate, the prognosis worsens.

With this in mind, the multinational tech company GMV, in collaboration with the Hospital Universitario La Paz and the Universidad Complutense de Madrid, has taken up the challenge proposed by the Center for Technological Development and Innovation (CDTI): developing a simulator based on artificial intelligence (AI) that makes it possible to understand and predict the course of interstitial lung diseases.

The company just presented its design proposal as part of the first phase of the pre-commercial public procurement launched by the CDTI. This project is financed with the center's own funds and through the Recovery and Resilience Mechanism (RRM), within the Recovery, Transformation and Resilience Plan (RTRP), financed by the European Union - NextGenerationEU.

AI and Deep Learning, critical parts of medical diagnostics of the future

According to the World Health Organization (WHO), 80% of medical decisions are based on X-ray evidence, making medical imaging a critical tool in diagnosing, monitoring, and even treating patients. However, it is particularly challenging to interpret DILD images 

Given this context, technologies such as artificial intelligence (AI) and, specifically, deep learning (supervised deep machine learning, inspired by the functioning of the human brain) are emerging as fundamental aids for specialists in diagnostic imaging.

Carlos Illana, Product Manager of Secure e-Solutions at GMV, explained: "In designing the planner, we used a combination of a complete and refined dataset and deep learning techniques to obtain a high level of accuracy, explainability and bias control. However, it is only when it is validated in a real clinical environment—in this case, Bellvitge Hospital—that we will be able to confirm whether there are any deviations from the results obtained in test environments."

Illana added that “we are well on the way to ensuring that this tool will help to avoid delayed diagnosis of DILDs, thus enabling us to offer personalized and more effective treatments. For example, in the case of IPF, where the average survival rate is 3 to 5 years from diagnosis.”

He also highlighted how unique the project is: “For us, this development represents a real challenge as there is currently no commercial solution on the market that covers all the clinical capabilities we are proposing. This makes it a milestone in the field of medical imaging and the application of artificial intelligence in healthcare.”

Disruptive advances in medical imaging

The GMV-developed simulator will use advanced analysis of computed tomography (CT) images, enabling the artificial intelligence to pinpoint all visible patterns in X-rays related to DILDs. It will also be able to determine the main pattern and indicate if it is a fibrotic disease.

Additionally, one of the project's most innovative features is the AI's ability to predict the course of the disease by integrating medical imaging with respiratory function tests. This predictive capacity is expected to help specialists predict how the disease will progress and adapt treatments to each patient's specific needs, significantly improving their quality of life.

Beyond lung diseases, this project lays the foundations for a broader integration of artificial intelligence in other medical areas, thanks to a multidisciplinary approach that brings together experts in biomedical engineering, radiology, pulmonology, and AI.

More info:

Marketing and Comunicación
GMV Secure e-Solutions
[email protected]