The application of AI medical imaging, radiation imaging and the most attention by the market, the reason for the demand side, the large amount of radiographic data, doctors need to spend considerable time and interpretation of discrepancies, from the technical aspects In the meantime, the sophistication of semiconductor computing techniques and deep learning algorithms has made AI-assisted radiologists diagnostic imaging possible.
However, DIGITIMES Research observes that regardless of whether the medical imaging diagnosis program proposed by a maker or a new entrepreneur is beyond the reach of human doctors at the present time, it is not clear why important parts of the development process such as how the database cooperates with the doctor Mode, etc., the doubts still exist in practical application.
DIGITIMES Research believes that the trend of application of AI in medical imaging is already taking shape. However, the outbreak speed will not be as fast as expected. The main reason is the data integrity and diversity in the international image database, which will result in the development of AI algorithm, And affect its commercial accuracy and clinical promotion difficult.
In addition, Taiwanese manufacturers still have a great deal of space in the AI medical imaging market. However, in the future, if they want to collaborate with or acquire new entrants, they should observe the details of the team's database establishment, such as the image labeling method and the actual participating mark Doctors professional, data marking consistency and other aspects.