AI Tumor Diagnosis in Brain Cancer

AI Tumor Diagnosis in Brain Cancer


AI Tumor Diagnosis in Brain Cancer When surgeons reach the periphery of a brain tumor, a challenging decision presents itself: should they remove a portion of healthy brain tissue to ensure complete tumor removal or avoid healthy tissue and risk leaving some malignant cells behind? Now, researchers in the Netherlands are harnessing the power of artificial intelligence to provide surgeons with critical information that can aid in making this crucial choice.

Thank you for reading this post, don’t forget to subscribe!

The innovative method, recently detailed in a study published in the journal Nature, involves a computer system scanning segments of a tumor’s DNA, pinpointing specific chemical modifications that can yield a detailed diagnosis of the tumor type, and even its subtype.

This real-time diagnosis, obtained during the early stages of lengthy brain surgery, offers invaluable guidance to surgeons on how aggressively to operate. Moreover, in the future, this method may facilitate tailored treatments for specific tumor subtypes.

AI Tumor Diagnosis in Brain Cancer

AI Tumor Diagnosis in Brain Cancer

“It’s imperative that the tumor subtype is known at the time of surgery,” emphasizes Jeroen de Ridder, an associate professor at the Center for Molecular Medicine at UMC Utrecht, a Dutch hospital, who led the study. “What we have now uniquely enabled is to allow this very fine-grained, robust, detailed diagnosis to be performed already during the surgery.”

Their deep learning system, aptly named “Sturgeon,” was initially tested on frozen tumor samples from previous brain cancer operations. Impressively, it accurately diagnosed 45 out of 50 cases within just 40 minutes of commencing genetic sequencing. In the remaining five cases, it refrained from providing a diagnosis due to unclear information.

AI Tumor Diagnosis in Brain Cancer 90 Mints

Subsequently, the system underwent testing during 25 live brain surgeries, most of which were performed on children, alongside the standard method of examining tumor samples under a microscope. The new approach yielded 18 correct diagnoses and fell short of the necessary confidence threshold in the other seven cases. Remarkably, it provided diagnoses in less than 90 minutes, enabling real-time decision-making during surgery.

Currently, in addition to examining brain tumor samples under a microscope, doctors can send them for more comprehensive genetic sequencing. However, not every hospital has access to this technology, and even for those that do, results can take several weeks to arrive, according to Dr. Alan Cohen, the director of the Johns Hopkins Division of Pediatric Neurosurgery and a cancer specialist.

“We have to start treatment without knowing what we’re treating,” Dr. Cohen laments.

The new method employs a faster genetic sequencing technique, focusing on a small slice of the cellular genome, ensuring results are available before the surgeon begins operating on the tumor’s margins.

Dr. de Ridder emphasizes the model’s strength in delivering a diagnosis with sparse genetic data, akin to recognizing an image based on just one percent of its pixels and from an unknown portion of the image. “It can figure out itself what it’s looking at and make a robust classification,” he notes.

However AI Tumor Diagnosis in Brain Cancer, some tumors remain challenging to diagnose. The samples obtained during surgery are roughly the size of a kernel of corn, and if they contain healthy brain tissue, the deep learning system may struggle to identify enough tumor-specific markers.

To address this, pathologists examining samples under a microscope were asked to identify those with the most tumor for sequencing, as explained by Marc Pagรจs-Gallego, a bioinformatician at UMC Utrecht and a co-author of the study.

Additionally, variations within a patient’s tumor cells mean that the sequenced segment may not represent the entire tumor. Some less common tumors might not correspond to previously classified types, and some tumors are easier to classify than others.

Other medical centers have already begun using the new method on surgical samples, underscoring its potential for widespread application.

However, Dr. Sebastian Brandner, a professor of neuropathology at University College London, cautions that sequencing and classifying tumor cells often require significant expertise in bioinformatics and capable individuals to operate and maintain the technology. “Implementation itself is less straightforward than often suggested,” he notes.

It’s worth noting that brain tumors are particularly well-suited to classification based on the chemical modifications that this new method analyzes, as not all cancers can be diagnosed this way.

This method is AI Tumor Diagnosis in Brain Cancer part of a broader movement aimed at bringing molecular precision to tumor diagnosis, potentially leading to the development of targeted treatments that are less harmful to the nervous system. However, translating this deeper understanding of tumors into new therapies remains a formidable challenge.

“We’ve made some gains,” Dr. Cohen reflects, “but not as many in the treatment as in the understanding of the molecular profile of the tumors.”

Read More Latest Technology brings AI in Brain Aneurysms Treatment