The potential of artificial intelligence in lung cancer

Gradually, artificial intelligence tools are being applied in clinical practice to enhance knowledge and approach to lung cancer. Knowing how patients respond to treatments such as immunotherapy, avoiding gender bias in studies that harm women, or improving survival are some of its capabilities.

Presentation of artificial intelligence-assisted radiation therapy at the Department of Radiation Oncology at the Charité Hospital in Berlin (Germany). EFE/EPA/CLEMENS BILAN

“There are already artificial intelligence tools that we use every day, especially for the early diagnosis of cancer,” radiologist Raquel Pérez-López explains to EFEsalud, clarifying that clinical trials are still needed to confirm the capabilities it can offer. a technology based on the analysis and mass aggregation of data.

As part of World Lung Cancer Day, November 17, a researcher from the Vall d’Hebron Institute of Oncology (VHIO) in Barcelona analyzes one of the research avenues that artificial intelligence (AI) could have in this area. tumor, one of the most common: more than 32,000 cases are diagnosed annually.

Artificial intelligence recognizes response to immunotherapy

This is a line of research being carried out VHIO

with financing from outside Cancer Crisis Foundationand which involves using artificial intelligence to predict whether patients will respond to treatment with immunotherapy, a new generation of drugs that have contributed to improved survival in this type of tumor.

“We want to select the best candidates who will benefit from immunotherapy,” the doctor says of the clinical trial already underway.

And it relies on patient information provided by three sources: CT image, MRI, and tumor biopsy to see what biomarkers it expresses. The data is analyzed by AI to find out each patient’s likelihood of reacting to these medications.

“AI allows us to do this kind of analysis on medical images. The information contained in CT images has not been used until now. AI allows us to integrate this type of data based on disease volume, location, and tumor pattern to better understand which patients are most likely to benefit from immunotherapy,” points out Dr. Raquel Perez. Lopez.

“We are trying to capture the characteristics of cancer cells and their microenvironment, such as lymphocytic cells, vascularization… factors that are involved in the likelihood of response and that can be captured through imaging and interpreted through artificial intelligence models,” he notes.

AI against false positives and negatives

Among the disadvantages of lung cancer late diagnosis (its symptoms can be confused with other pathologies) and false positives or negatives that imaging tests may reveal and subsequent need for confirmation with biopsy

It is in diagnostics that the greatest progress has been achieved with the help of artificial intelligence, which makes it possible to identify and analyze the characteristics of nodules in the lungs, excluding false-positive and false-negative results.

But AI also allows us to take steps to find biomarkers, or combinations of genetic changes, that are associated with a greater likelihood of response and prognosis.

The doctor mentions important research work aimed at using AI to predict the risk and therefore prevent lung cancer, taking into account not only already controlled variables such as smoking, but also an analysis of the patient’s clinical condition. information, lifestyle or family history.

End gender bias

Annual report “Cancer figures in Spain in 2024” belonging Spanish Society of Medical Oncology (SEOM)in collaboration with Spanish Network of Cancer Registries (REDECAN)reflects that lung cancer in women is the third most common cancer (after breast and colorectal cancer) with an estimated 10,285 cases for that year.

The numbers, which are rising in women, are expected to increase by 12.3% by 2024 compared to the previous year, and one of the reasons is the effects of tobacco smoking.

This makes it necessary to work from a gender perspective, with women’s own data, in the prevention, approach and research of lung cancer, as stated by Association for Research into Women’s Cancer Lungs (ICAPEM).

“There are gender differences in cancer in general, as well as in lung cancer, differences in their susceptibility or risk of contracting it, in their prognosis, in their tolerance to treatment, which are secondary to various biological as well as sociocultural aspects that must be thoroughly investigated “, – note in this organization.

In this sense, artificial intelligence can also play a role, being able to find specific aspects of women compared to men, but as VHIO researcher Raquel Perez-Lopez says, AI models are trained to take this diversity into account. also necessary among representatives of different nationalities.

The present and future of artificial intelligence tools in lung cancer, which over time can help reduce mortality from this tumor, which is currently the leading cancer mortality rate and which is associated with tobacco habit in more than 80% of cases.