When surgeons operate on brain tumors, they have to decide how much healthy brain tissue to remove along with the tumor. Removing more tissue gives a better chance of getting all of the tumor out. But it also takes away a more healthy brain.

Surgeons want to remove the tumor but spare the healthy brain. To do this, they need to know exactly what type of tumor it is. That helps them know how aggressive it is and how much tissue they need to remove.

The New Solution

Scientists in the Netherlands created an A.I. system called Sturgeon. It uses a patient’s tumor DNA to diagnose the tumor type and subtype. This gives the surgeon details about the tumor while the patient is still in surgery. The study was published in a reputable science journal NATURE.

The system looks at chemical modifications in the tumor DNA. It uses deep learning to recognize patterns that reveal the tumor type. This all happens in less than 90 minutes.

How Well It Works

Sturgeon was tested on 50 frozen tumor samples first. It accurately diagnosed 45 of them in under 40 minutes.

It was then tested during 25 actual surgeries, mostly on children. It gave 18 correct diagnoses. For 7 tumors there wasn’t enough data for Sturgeon to be confident.

The Benefits

• Gives a diagnosis during surgery, unlike standard tests that take weeks
• Helps the surgeon decide how aggressive to be and how much tissue to remove
• May guide better treatment tailored to the specific tumor

Sturgeon allows detailed diagnosis during surgery when it’s most useful. This can lead to better outcomes for patients.

Conclusion

The new A.I. tool Sturgeon can diagnose brain tumors during surgery. This helps surgeons remove the tumor while saving healthy brain tissue. Faster diagnosis means surgeons can be more precise. It also allows for better treatment matched to each tumor type. Overall, Sturgeon has the potential to improve outcomes for brain tumor patients. This shows how A.I. can provide real benefits during medical procedures.

FAQ’s

How AI is used in brain tumor detection?

AI-based brain tumor detection involves three main steps. First is image enhancement to improve visual quality. The second step locates tumors using a clustering method. This identifies areas that could contain tumors. The third step is feature extraction. A recent deep-learning approach uses two pre-trained models to extract image features as vectors.

How accurate is AI diagnosis?

AI methods like machine learning (ML) and deep learning (DL) can improve the accuracy of diagnosing genetic disorders. In psychiatry, ML approaches have achieved 48% to 95% accuracy in diagnosis from genetic data. This demonstrates the potential for AI to reliably predict disorders based on genetics. With further development, ML and DL models may continue improving diagnosis accuracy across medical fields.

Which AI is best for medical diagnosis?

Top AI Tools for Medical Diagnosis

• Twill – Extracts key medical data to help analyze patient information.
• Merative – Uses ML to analyze records and suggest diagnoses.
• Viz.ai – Detects signs of stroke in CT scans through deep learning.
• Enlitic – Applies deep learning to medical images to find abnormalities.
• Regard – NLP system that reads doctor notes to recommend diagnoses.