Artificial intelligence (AI) is being utilized to assist doctors in examining many mammograms at Aberdeen Royal Infirmary, and their research has already shown encouraging results, as it effectively identified early-stage breast cancer.
Mammograms, low-level X-rays used in breast cancer screenings, are crucial for finding minute alterations that could be too tiny to see or feel. According to the NHS, mammograms save around 1,300 lives annually in the UK.
The number of Scottish women receiving regular breast tests has grown, while the number of radiologists reviewing the data has decreased. Between 2021 and 2022, Scotland had 1,830 screen-detected breast cancers, according to Public Health Scotland, according to the BBC.
Radiologists call 250 to 300 patients back for additional annual examinations, with 30 to 40 needing further care. Dr. Gerald Lip, clinical director of the North East Scotland Breast Screening Programme, highlighted worries about missing cancer cases due to multiple scans.
In 2021, Scotland's AI Strategy featured the Gemini initiative, a cooperation between NHS Grampian, the University of Aberdeen, and private enterprises. Mia, an AI model created by Kheiron Medical Technologies with assistance from Microsoft's cloud computing capabilities, is being used in the experiment.
Dr. Lip and other radiologists are undertaking experiments to include AI as an extra check after mammography scan assessments, despite the National Screening Council's existing restrictions prohibiting the automated use of AI in screenings.
The AI software's ability to reveal variations between mammograms of the left and right sides was demonstrated by Dr. Lip. With this technology, radiologists were able to inspect and confirm the changes the AI detected by pressing a button. Dr. Lip then called attention to a particular location that the AI software had identified as the critical problem.
A new study found that artificial intelligence is significantly better at predicting a patient’s breast cancer risk than current methods. The research analyzed mammograms of more than 18,000 women in which AI took just seconds to determine their five-year risk. pic.twitter.com/VwtFFe4skE
— CBS Evening News (@CBSEveningNews) June 6, 2023
Read Also: Scientists Discover New Treatment for Cannabis Addiction
AI's Big Help in The Fight Against Breast Cancer
In a related breast cancer research published in the journal Radiology, artificial intelligence has shown advantages in assisting with risk prediction for breast cancer. In the research, the performance of the five AI algorithms was compared to the conventional Breast Cancer Surveillance Consortium risk model.
The risk model of the Breast Cancer Surveillance Consortium (BCSC) uses family history, age, and other characteristics to predict a patient's five-year risk of breast cancer, per SciTech Daily.
The research found that AI algorithms outperformed the conventional risk model regarding accuracy. Breast density is the leading imaging biomarker used in clinical risk models. However, AI's deep learning can extract many other mammographic parameters.
The retrospective research originally included more than 324,009 participants and ran from 2016 to 2021. In the study, 13,628 patients were chosen at random, and 193 of them subsequently had incident cancer, USA Today reported.
All study participants had negative screening mammography and no evidence of malignancy. Within seconds, the AI algorithms developed risk ratings for breast cancer by analyzing these preliminary screens.
Even AI systems trained for short time horizons-as little as three months-could forecasts the risk of breast cancer for up to five years in the time ahead when no tumor was clinically diagnosed by screening mammography. The study suggested that combining the AI and BCSC risk models enhanced cancer prediction even more.
Related Article: Remote-Controlled Video Capsule Revolutionizes Endoscopy
