Tiny brain cancers that MRI scans frequently miss can now be found using AI. The Netherlands Cancer Institute and Robovision Healthcare collaborated on this ground-breaking study, which demonstrates how AI is still revolutionizing medical innovation.
The fact that brain metastases are the most prevalent kind of tumors in the central nervous system is one factor contributing to the significance of this discovery. “They can impact as many as 17% of adult cancer patients,” Robovision said in a statement on the achievement. Even though these tumors are quite prevalent, they can frequently be difficult to identify early since they develop quickly and are often less than 3 mm. This makes it challenging to identify them in the several MRI slides collected in a single scan, particularly in cases with a high volume.
How AI is enhancing the diagnosis of cancer
But it can all change in the near future. In a recent study, researchers tested Robovision Healthcare’s new AI, BrainMets.ai, and found that it achieved a lesion-level sensitivity of 97.4%, as reported in the journal Radiology. But from then, the outcomes differ according on how big the lesions are.
100% of brain images with lesions were identified by the AI when the lesions were 12 mm or larger. Accurate detection of lesions 6–12 mm was 98% of the time, whereas 97.9% of the time, lesions 3-6 mm were recognized. Last but not least, in 93% of the brain scans that were tested, the AI successfully identified lesions that were smaller than 3 mm.
This achievement is undoubtedly positive, particularly in light of other recent developments in medical artificial intelligence. In the past, researchers were able to design an AI that could identify Alzheimer’s symptoms based just on speech recordings. According to other studies, AI is also 99% accurate in detecting some types of cancer.
The purpose of these developments is not to completely replace physicians. In order to begin fighting these dangerous diseases early, it is intended to enhance our ability to recognize them. Our ability to digest critical information more rapidly and effectively can be enhanced by combining human and artificial intelligence capabilities. However, it does necessitate having the appropriate elements in place. According to Robovision, the AI’s capabilities are limited by the training process.
