Women from the age of 40 are usually suggested annual mammography when it comes to breast cancer. Several studies have revealed that screening mammography helps lower breast cancer mortality by reducing the chances of advanced cancer. However, conventional methods available today for breast cancer risk assessment use clinical factors for analysis and are not considered very effective.
Recently research has been published that could provide a better way for accomplishing breast cancer risk assessment. The findings in the study suggest that AI has great potential in being the second reader for radiologists. The technology can facilitate a reduction in unnecessary imaging and related costs. The study might contribute significantly to the Breast Cancer Diagnostics Market as it would empower doctors to look at the woman’s individual risk and then decide the frequency at which she needs to be monitored.
Deep Learning, a sophisticated type of Artificial Intelligence (AI), might better distinguish between mammograms of women that might turn into breast cancer in the future and which will not. This is in comparison to currently used clinical risk factors, which is not very reliable at this task.
To reach this conclusion, the team took a data set of more than 25,000 digital screening mammograms from 6,369 different women who were participants in screening mammography. From these participants, it was noted that more than 1,600 women developed screening-detected breast cancer, and around 351 developed intervals invasive breast cancer.
After that, researchers instructed the deep learning model to identify signals or details within the mammogram that might be related to increased cancer risk. The tests demonstrated that the deep-learning-based model underperformed when determining the risk factors for interval cancer risk. However, it outmatched clinical risk factors and breast density in predicting screening-detected cancer risk.
Further, the results also displayed that extra signals received due to AI enables an improved risk estimate for screening-detected cancer. Thus, the team was able to accomplish its goal of differentiating between women based on high or low risk of screening-detected breast cancer.
The findings proposed in the research may impact the clinical practices significantly, wherein breast density still accounts for most management decisions. However, instead of being recommended next year for screening, women with negative mammograms can now be sorted on risk into three pathways, i.e., low risk of breast cancer, high interval invasive cancer, or elevated screening-detected risk for the next three years.
Not only this, but the deep learning model also holds great promise towards supporting decisions on addition imaging in MRIs and other modalities.
Other Related Reports:
Global Breast Cancer Treatment Drugs Market 2020 by Manufacturers, Regions, Type and Application, Forecast to 2025
Global Breast Cancer Monoclonal Antibodies Market Growth (Status and Outlook) 2020-2025
Global Breast Cancer Screening Market Research Report - Industry Analysis, Size, Share, Growth, Trends and Forecast Till 2026
