- Researchers at University of Michigan developed AI model Prima for brain MRI diagnosis
- Prima achieved up to 97.5% accuracy identifying over 50 neurological conditions quickly
- Model trained on 220,000 MRI studies and patient histories over one year
Researchers at the University of Michigan have created an artificial intelligence (AI) model that can interpret and diagnose brain Magnetic Resonance Imaging (MRI) scans in just seconds, unlike humans. The model, named Prima, achieved accuracy as high as 97.5 per cent and outperformed other advanced AI tools, according to the findings published in the journal Nature Biomedical Engineering.
The model was able to identify more than 50 neurological conditions, including strokes, brain tumors and haemorrhages. Trained on hundreds of thousands of real-world scans along with patient histories over a one-year period, researchers claim this model has the potential to reshape how brain imaging is handled across health systems.
"Trained on over 220,000 MRI studies, Prima uses a hierarchical vision architecture that provides general and transferable MRI features. Prima was tested in a 1-year health system-wide study that included 29,431 MRI studies," the study highlighted.
Todd Hollon, the lead author of the study and a neurosurgeon at U-M Health, described Prima as "ChatGPT for medical imaging", stating that it could ease workflows, especially in rural areas where there might not be a neurology specialist immediately available to review imaging.
“As the global demand for MRI rises and places significant strain on our physicians and health systems, our AI model has potential to reduce burden by improving diagnosis and treatment with fast, accurate information,” said Hollon.
"We believe that Prima exemplifies the transformative potential of integrating health systems and AI-driven models to improve healthcare through innovation."
Study co-first author Samir Harake, an MD candidate and data scientist in Michigan's Machine Learning in Neurosurgery Lab, explained that Prima works like a radiologist by combining a patient's medical history and imaging data to paint a conclusive picture.
AI Tools For Medical Diagnosis
AI tools are increasingly being used by the public to better understand their medical conditions. Last month, a video of Elon Musk stating that AI can help provide efficient medical diagnoses went viral on social media.
"I think AI will be very helpful with the medical stuff. Right now, you can upload your X-rays or MRI images to Grok, and it will give you a medical diagnosis. I have seen cases where it's actually better than what doctors tell you," said Musk.
Musk highlighted the case of a 49-year-old Norwegian man who credited Grok with saving his life following a misdiagnosis. The man experienced severe abdominal pain for over 24 hours. An initial ER visit resulted in a diagnosis of acid reflux and a prescription for antacids.
Still in pain, he consulted Grok, which flagged a possible perforated ulcer or appendicitis and urged immediate CT scanning. Upon returning to the hospital, tests confirmed a severely inflamed, near-ruptured appendix. Emergency surgery followed, leading to a full recovery.
