Google uses AI to diagnose breast cancer

The tool is designed to help radiologists detect and assess breast cancer risk.

On November 29, 2022, Google announced that it had licensed an artificial intelligence (AI)-based breast cancer screening research model to medical technology company iCAD. Both aim to deploy the tool in real clinical settings by 2024. However, commercial implementation still requires more research and testing.

Previously, Google developed its own artificial intelligence tool to identify breast cancer. In 2020, Google researchers will publish research on this topic in the journal Nature. Specifically, they found that their AI system was able to identify signs of breast cancer better than some radiologists. Across thousands of mammograms, the model reduced false negatives by 9.4 percent and false positives by 5.7 percent.

Through the partnership, iCAD will integrate Google’s mammography AI research model into its existing tools.

The first tool, called “ProFoundAI,” analyzes images from digital mammography (DBT—an advanced imaging technique sometimes called 3D mammography). The tool scans images from DBT and then determines the density and calcification of malignant soft tissue.

The second tool that iCAD hopes to integrate with Google is a risk assessment tool. Based on recommendations, the tool estimates an individual’s risk of developing breast cancer.

The goal of the Google and iCAD collaboration is to turn AI into a tool to help radiologists detect and diagnose breast cancer.

In general, however, medical experts are wary of AI because it is not the gold standard. For example, in Google’s 2020 study, there were cases where the tool didn’t detect but the radiologist did.

Not only that, but there are currently no clear criteria to tell if you have breast cancer. However, to develop an algorithm, precise numbers and formulas are required.

So now, this AI tool will provide more than two outcomes, not just “has cancer” or “didn’t have cancer”, especially considering the “cancer area”. grey”. “.in diagnosis”, i.e. when the results can be confusing. Also, don’t rely entirely on artificial intelligence, because doctors also need to assess the patient’s overall health to make accurate conclusions, especially in the early stages of screening Stage – The stage of the cancer.

Google, on the other hand, said it was working with the UK’s National Health Service (NHS) and Imperial College London to test whether its artificial intelligence technology could work like “tittle”. or not. Second independent reading” during the two readings of the mammogram (i.e. the mammogram results will be evaluated separately by 2 independent physicians to improve diagnostic accuracy). This way, physicians can focus on priority cases and maintain screening consistency and quality.

Leave a Reply

Your email address will not be published. Required fields are marked *

GIPHY App Key not set. Please check settings

Qualcomm takes another step toward integrating AI into smartphones

Which foods help improve memory?