AI can diagnose brain tumors better, faster and earlier than doctors can

‘Game-changing’ AI can diagnose brain tumors better, faster and earlier than doctors can, study suggests

  • Researchers trained AI to recognize the 10 most common types of brain tumors
  • Tumor specimens were split and analyzed by the AI and a team of doctors 
  • The algorithm was 94.6% accurate at correctly diagnosing the tumors compared to doctors who were 93.9% accurate

Artificial intelligence can detect brain tumors during surgery better than trained doctors, a new study claims. 

Researchers developed a computer algorithm that could diagnose the cancerous masses with almost 95 percent accuracy.

The results analyzed by neuropathologists were slightly less accurate at around 94 percent.

The team, from New York University Langone Health, says the technology will increase both the speed and accuracy of diagnosis so physicians can focus on making sure their patients receive the proper care. 

A new study from NYU Langone Health has found that artificial intelligence can correctly diagnose brain tumors with 94.6% accuracy compared to 93.9% accuracy by doctors (file image)

‘As surgeons, we’re limited to acting on what we can see,’ said senior author Dr Daniel Orringer, an associate professor of neurosurgery at NYU Grossman School of Medicine.

‘This technology allows us to see what would otherwise be invisible, to improve speed and accuracy in the [operating room], and reduce the risk of misdiagnosis.’

For the study, published in Nature Medicine, the team used an imaging technique called stimulated Raman histology.

Developed by Dr Orringer, it finds tumors in human tissue with scattered laser light, which illuminates features that tradition scans typically can’t pick up.

The microscopic images are then processed by the AI in about two-and-a-half minutes – about a tenth of the time pathologists would need to identify a tumor from diagnostic scans. 

To train the algorithm, researchers used more than 2.5 million samples from 415 patients with the 10 most common types of brain cancer.

Next, they recruited 278 brain tumor patients undergoing surgery at three different hospitals or medical centers. 

Brain tumor specimens were biopsied from the patients, split into two and assigned to either doctors or the AI. 

The AI diagnosed the tumors more precisely and consistently than doctors did, with 94.6 percent accuracy compared to 93.9 percent. 

Researchers say not only could the algorithm be faster, cheaper and diagnose tumors earlier on in their progressions than traditional methods, but the technology could also be useful to centers that don’t have expert neuropathologists.

‘With…this game-changing technology, we’re now even better equipped to provide safe surgeries and quality outcomes for the most complex brain tumor cases,’ said Dr John Golfinos, chair of the department of neurosurgery at NYU Langone.  

The study comes on the heels of a growing body of research that has found that artificial intelligence can improve upon the work of physicians.

Last year, two separate studies from the Mayo Clinic and Stanford University found that AI was able to identify irregular heart rhythms and the early stages of heart failure.

Additionally, in 2019, researchers at the National Cancer Institute and Global Good developed a computer algorithm that can identify precancerous changes in a woman’s cervix up to 1.3 times better than standard tests. 

Read more at DailyMail.co.uk