Lung cancer: the Google program that detects this disease

| |

Can Artificial Intelligence (AI) be better than medical specialists in diagnosing lung cancer?

This is suggested by a recent study by scientists at Northwestern University in Illinois, United States, who together with Google hope that this technology can increase the effectiveness of cancer detection.

Finding tumors at an earlier stage should facilitate the treatment of the disease.

The team in charge of the investigation said that IA will have a “huge” role in the future of medicine, but the current software is not yet ready for clinical use.

The study focused on lung cancer, which kills more people (1.8 million per year) than any other type of cancer.

That's why the US recommends screening tests for people at high risk due to years of excessive tobacco use.

However, screening tests may result in invasive biopsies for people who do not have cancer and may also not detect some tumors.

How was the study?

The study used artificial intelligence to decide If the analysis of the tomographies could be improved.

The first step was to train computer software with 42,290 lung CT scans of nearly 15,000 patients.

Getty Images

The Google program analyzed more than 42,000 lung CT scans.

The researchers did not tell the AI ​​what to look for, they only revealed which patients had cancer and which did not.

The results that the AI ​​showed were compared to the analysis of a team of six radiologists experts in the interpretation of computed tomography.

The program was more effective than radiologists when examining a single CT scan and was just as effective when doctors had multiple CT scans to interpret.

The results, published in the journal Nature Medicine, showed that AI could increase the detection of cancer by 5% and at the same time reduce false positives (people falsely diagnosed with cancer) by 11%.

“The next step is to use it in patients in the form of a clinical trial,” Dr. Mozziyar Etemadi, of Northwestern University, told the BBC.

According to Etemadi, sometimes the AI ​​”indicates a pulmonary nodule (a growth) that, in every sense, seems benign, but the program believes that it is not. gIt is generallyin Right

“One area of ​​scientific research is to find out why,” he added.

Getty Images

The study focused on lung cancer because it is the one that kills more people in the world.

Etemadi says that if a joint effort is made between artificial intelligence and doctors, the result would be even more effective and that AI can have a “huge” role in medicine.

For its part, Rebecca Campbell of Cancer Research in the United Kingdom believes that it is encouraging to see technological innovations that someday could help detect lung cancer at an early stage.

“Similar to how we learn from experience, these algorithms perform a task repeatedly, and each time it is adjusted a bit to improve accuracy,” he says.

“Detecting cancer early, when treatment is more likely to be successful, is one of the most powerful ways to improve survival, and developing an economic technology that is not invasive could play an important role, “adds Campbell.

“The next steps will be to test this technology even more to see if it can be applied accurately to a large number of people.”

Previous

Abemaciclib duplicates progression-free survival versus hormone therapy in metastatic breast cancer

Hector Bonilla: stealing as a lawyer was not my thing, I preferred to be an actor

Next