Connect with us

AI Research

AI just detected tiny brain tumors often missed by MRIs

Published

on


AI can now detect tiny brain tumors often missed by MRI scans. This breakthrough research, which was conducted by the Netherlands Cancer Institute and Robovision Healthcare, showcases that AI continues to be a game-changer when it comes to pushing medical innovation to new levels.

One reason this breakthrough is so notable is because brain metastases are by far the most common type of tumors found in the central nervous system. They can affect up to 17% of adult cancer patients, according to Robovision’s statement on the accomplishment. Despite being so common, early detection of these tumors is often difficult, as they grow fast and can be smaller than 3mm in most cases. This makes them hard to spot in the various MRI slides that are taken during a single scan, especially in high volume cases.

How AI is improving cancer detection

However, that could all change very soon. According to a new study published in the journal Radiology, researchers testing BrainMets.ai, the new AI from Robovision Healthcare, were able to achieve a 97.4% lesion-level sensitivity. The results vary from there, though, depending on the size of the lesions present.

In lesions that were 12mm or greater, the AI detected 100% of the brain scans with lesions present. Lesions 6-12mm in size were detected accurately 98% of the time, while those between 3-6mm were detected 97.9% of the time. Finally, for lesions smaller than 3mm, the AI was able to accurately detect them in 93% of the brain scans put through the test.

This accomplishment is certainly encouraging, especially when tied together with other recent advances in medical AI. Previously, scientists managed to create an AI to detect signs of Alzheimer’s development just by listening to voice recordings. Other research has proven that AI can detect certain kinds of cancer 99% of the time, too.

The goal of these advancements isn’t to replace doctors outright. Instead, it’s meant to improve how we detect these deadly diseases, so we can start combating them earlier. Combining human power with the power of AI can help us process important information more quickly and efficiently. But it does require having the right components in place. Robovision says that its AI is only as capable as it is because of how it was trained.

It’s all about having the right ingredients and developers in place to bring the AI tools to life. This helps avoid false detections and other issues like hallucinating. Of course, it’s unlikely AI will ever be infallible. But with humans also helping along the way, it doesn’t need to be. It just needs to speed up how we detect cancer—which it is doing—by finding the patients most likely to need a deeper look.



Source link

AI Research

The new frontier of medical malpractice

Published

on


Although the beginnings of modern artificial intelligence (AI) can be traced
as far back as 1956, modern generative AI, the most famous example of which is
arguably ChatGPT, only began emerging in 2019. For better or worse, the steady
rise of generative AI has increasingly impacted the medical field. At this time, AI has begun to advance in a way that creates
potential liability…



Source link

Continue Reading

AI Research

Pharmaceutical Innovation Rises as Global Funding Surges and AI Reshapes Clinical Research – geneonline.com

Published

on



Pharmaceutical Innovation Rises as Global Funding Surges and AI Reshapes Clinical Research  geneonline.com



Source link

Continue Reading

AI Research

Radiomics-Based Artificial Intelligence and Machine Learning Approach for the Diagnosis and Prognosis of Idiopathic Pulmonary Fibrosis: A Systematic Review – Cureus

Published

on



Radiomics-Based Artificial Intelligence and Machine Learning Approach for the Diagnosis and Prognosis of Idiopathic Pulmonary Fibrosis: A Systematic Review  Cureus



Source link

Continue Reading

Trending