A deep learning (DL) model based on mammograms has been created by researchers from the Jameel Machine Learning Clinic and MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL). The model could help detect precancerous lesions in high-risk women. The artificial intelligence (AI) system, named Mirai, can outperform existing risk assessment algorithms when estimating the likelihood of breast cancer from mammograms. In a post on X, entrepreneur Anand Mahindra cited an article discussing an AI that helped detect breast cancer five years earlier.
“If this is right, AI will deliver much more value than we ever imagined, much sooner than we ever imagined,” the Mahindra president said in a repost.
With over 200,000 views, social media users seem to very much agree.
“This is very accurate,” one person said. “We are also working on AI to detect arrhythmias from heartbeats and COVID-19 from patients' coughs. AI will change the way we treat patients.”
Another user said: “In any case, that's the best use of technology – helping humanity become better and reduce risks.”
“Systems based on AI, machine learning and big data analytics will dramatically improve living standards for all segments of society,” another comment read.
Video Carousel
One user further suggested, “It's time to incorporate AI into healthcare as a topic of medical research. Doctors who leverage technology can achieve amazing things in preventing, diagnosing, and treating diseases.”
Video Carousel
Meanwhile, Mirai was tested on pending patients and other datasets at Massachusetts General Hospital (MGH), said the study published in Science Translational Medicine. Mirai was trained using MGH's large-scale screening mammography dataset.
Mirai has four modules. To create a mammogram diagram, the image aggregator module first collects and processes all traditional mammography images. From there, it aggregates the image data from all views. Next, the risk factor prediction module uses the mammograms to predict the patient's risk factors, if applicable. In the last stage, the additive hazard layer uses the patient's risk variables and the mammography analysis to predict the patient's risk annually for the next five years.
One user further suggested, “It's time to incorporate AI into healthcare as a topic of medical research. Doctors who leverage technology can achieve amazing things in preventing, diagnosing, and treating diseases.”
Top Videos
Show all
Manu Bhakar wins India's first medal at Paris 2024 Olympics; Deepika, Kangana, Alia and Kareena react
Fantastic Four: The First Step cast announces filming will begin after years of delays
Pedro Pascal and Vanessa Kirby talk about their excitement for filming 'Fantastic Four: The First Step'
Phir Aayi Haseen Dilruba Limited Cast: Taapsee Pannu, Vikrant Massey, Sunny Kaushal on Love
Lisa Ray opens Rahul Mishra's show at India Couture Week, walking for 'Afreen Afreen'
Meanwhile, Mirai was tested on pending patients and other datasets at Massachusetts General Hospital (MGH), said the study published in Science Translational Medicine. Mirai was trained using MGH's large-scale screening mammography dataset.
Mirai has four modules. To create a mammogram diagram, the image aggregator module first collects and processes all traditional mammography images. From there, it aggregates the image data from all views. Next, the risk factor prediction module uses the mammograms to predict the patient's risk factors, if applicable. In the last stage, the additive hazard layer uses the patient's risk variables and the mammography analysis to predict the patient's risk annually for the next five years.
First revealed: 29 July 2024 12:44 IST
