AI Speeds Up Brain Tumor Classification with 95% Accuracy: Study

Australian and US researchers highlight their artificial intelligence (AI) technology that can identify brain cancers faster and more precisely.

The Australian National University and the National Cancer Institute in the US have developed a deep learning model to predict DNA methylation, which controls gene activity, and classify brain tumors into 10 major categories, according to a report from VOA.

DEPLOY, a US-Australian system, analyzes histopathology images of patient tissue, initially reported in Nature Medicine.

AI Speeds Up 'Very Slow' Brain Tumor Testing, Treatment

Many hospitals globally do not provide DNA methylation-based profiling, which takes weeks to obtain findings. Australian National University research fellow Danh-Tai Hoang believes that the AI-powered technology will speed up diagnostics.

Hoang noted that the current method used in identifying brain tumors is "very time-consuming" as test results become available after two to three weeks.

He and fellow researchers noted that the DEPLOY technology can save the lives of many brain tumor patients, especially those who have high-grade conditions.

With a 95% accuracy rate, the researchers verified their AI model on large datasets of 4,000 US and European patients.

This development in the use of AI in brain tumor treatment comes after a recent TechTimes report on how a woman got her voice back with the help of the technology.

Last summer, a life-threatening brain tumor removed Alexis "Lexi" Bogan, known for her lively voice, and she lost her ability to talk effectively.

Surgery removed the tumor in August, but Bogan struggled to talk a month later once the breathing tube was withdrawn. After months of therapy, her speech remained affected, making conversation difficult.

An AI program was used to restore Lexi's speech in April, reconstructing her voice from brief adolescent audio, sounding quite natural.

The AI was trained by the Rhode Island Lifespan Hospital network medical team with a 15-second recording via a voice cloning tool developed by OpenAI.

Bogan uses the app 40 times a day to give future patients feedback. It helps her communicate in preschool, at stores, and with her father.

While Bogan's medical team aims to spread this technology worldwide to help millions of patients suffering from stroke, throat cancer, and neurological conditions, OpenAI remains cautious with its voice cloning tool, which has not yet been made public.

AI-Generated Drugs Coming Soon?

In separate news on AI's impact on healthcare, A recent study underlined the effectiveness of AI-generated compounds, notably at Eli Lilly, which impressed experts.

Experts at the firm say AI's capacity to investigate millions of compounds in minutes eclipses wet laboratories' annual output. Because of its efficiency, medical research should utilize AI's full potential.

AlphaFold, an AI program developed by DeepMind, showed how AI is crucial to biology, according to Nvidia vice president of health care Kimberly Powell. She noted that the breakthrough can train "transformer models with very large data sets."

Generative AI explores the unimaginable potential to accelerate medication creation and optimization. Powell refers to this AI as "hallucinating" novel medications, which speed up chemical analysis.

With the technology, experts can quickly screen enormous chemical libraries and predict protein structures with remarkable precision, which can significantly help in developing life-saving medications.

Despite this advancement in AI, human trials remain essential, Powell noted.