From X-rays, AI can identify race?

arts/

AI medical imaging refers to the specific application of artificial intelligence technology in the diagnosis of medical imaging. In fact, AI medical imaging is also considered by industry insiders to be the most likely to take the lead in the commercialization of artificial intelligence medical field.

According to the data report of Global Market Insight, from the perspective of application division, the artificial intelligence medical imaging market is the second largest market segment in the field of artificial intelligence medical applications (in the global artificial intelligence medical market, the largest market segment is drug research and development). , the largest share, accounting for 35%), will develop at a growth rate of more than 40%, reaching a scale of 2.5 billion US dollars in 2024, accounting for 25%.

Both patients and physicians will benefit from the application of artificial intelligence in the field of medical imaging. For patients, Al medical imaging will help them complete health examinations more quickly, including X-ray, B-ultrasound, MRI, etc., and can obtain more reliable diagnosis results.

For radiologists, the application of artificial intelligence technology will reduce their reading time, greatly improve work efficiency, and reduce the possibility of misdiagnosis. The combination of artificial intelligence and medical imaging can provide assistance and reference for doctors to read and delineate images, greatly save doctors' time, and improve the accuracy of diagnosis, radiotherapy and surgery.

In the current AI medical imaging application scenarios, there are mainly three requirements : First, the requirements for lesion identification and labeling, that is, AI medical imaging products are required to perform image segmentation, feature extraction, quantitative analysis, and comparative analysis for medical images. The second is the need for automatic target delineation and adaptive radiotherapy, that is, the need for Al medical imaging products to process the images of the tumor radiotherapy link. The third is the demand for 3D reconstruction of images, that is, the need for Al medical imaging products to perform 3D reconstruction on the basis of artificial intelligence recognition for the operation.

Of course, scientists are also trying to make AI medical imaging have more applications, so that AI medical impact has more comprehensive functions. Often, doctors cannot tell if someone is black, Asian, or white just by looking at their X-rays. But computers can do it, according to a new paper recently published by an international team of scientists including researchers from MIT and Harvard Medical School.

The study found that an artificial intelligence (AI) program trained to read X-rays and CT scans could predict a person's ethnicity with 90 percent accuracy.

The study also gives scientists new hints — at a time when AI software is increasingly being used to help doctors make diagnostic decisions, is it possible that AI-based diagnostic systems could inadvertently produce racially biased results? For example, AI (where X-rays are available) could automatically recommend a specific treatment for all black patients, regardless of whether it's best for a particular person. At the same time, the patient's human doctor would not know that the AI's diagnosis was based on racial data.

The paper was published Wednesday in the medical journal The Lancet Digital Health.