Machines see the future for patients diagnosed with brain tumors

March 13, 2018, Emory University
Credit: CC0 Public Domain

For patients diagnosed with glioma, a deadly form of brain tumor, the future can be very uncertain. While gliomas are often fatal within two years of diagnosis, some patients can survive for 10 years or more. Predicting the course of a patient's disease at diagnosis is critical in selecting the right therapy and in helping patients and their families to plan their lives.

Researchers at Emory and Northwestern Universities recently developed artificial intelligence (AI) software that can predict the survival of diagnosed with glioma by examining data from tissue biopsies. The approach, described in Proceedings of the National Academy of Sciences, is more accurate than the predictions of doctors who undergo years of highly-specialized training for the same purpose.

Doctors currently use a combination of genomic tests and microscopic examination of tissues to predict how a patient's disease will behave clinically or respond to therapy. While genomic testing is reliable, these tests do not completely explain patient outcomes, and so microscopic examination is used to further refine prognosis. Microscopic examination, however, is very subjective, with different pathologists often providing different interpretations of the same case. These interpretations can impact critical decisions like whether a patient enrolls in an experimental clinical trial or receives radiation therapy as part of their treatment.

"Genomics have significantly improved how we diagnose and treat gliomas, but remains subjective. There are large opportunities for more systematic and clinically meaningful data extraction using computational approaches," says Daniel J. Brat, MD, PhD, the lead neuropathologist on the study, who began developing the software while at Emory University and the Winship Cancer Institute. Brat currently is chair of pathology at Northwestern University Feinberg School of Medicine.

The researchers used an approach called deep-learning to train the software to learn visual patterns associated with patient survival using microscopic images of brain tumor tissue samples. The breakthrough resulted from combining this advanced technology with more conventional methods that statisticians use to analyze . When the software was trained using both images and genomic data, its predictions of how long patients survive beyond diagnosis were more accurate than those of human pathologists. The study used public data produced by the National Cancer Institute's Cancer Genome Atlas project to develop and evaluate the algorithm.

"The eventual goal is to use this software to provide doctors with more accurate and consistent information. We want to identify patients where treatment can extend life," says Lee A.D. Cooper, PhD, the study's lead author, a professor of biomedical informatics at Emory University School of Medicine and member of the Winship Cancer Institute. "What the pathologists do with a microscope is amazing. That an algorithm can learn a complex skill like this was an unexpected result. This is more evidence that AI will have a profound impact in medicine, and we may experience this sooner than expected."

The researchers also demonstrated that the software learns to recognize many of the same structures and patterns in the tissues that pathologists use when performing their examinations. "Validation remains a barrier to using these algorithms in patient care. Being able to explain why an algorithm works is an important step towards clinical implementation."

The researchers are looking forward to future studies to evaluate whether the can be used to improve outcomes for newly diagnosed patients.

Explore further: Machine learning techniques show promise for supporting medical decisions

More information: Pooya Mobadersany et al. Predicting cancer outcomes from histology and genomics using convolutional networks, Proceedings of the National Academy of Sciences (2018). DOI: 10.1073/pnas.1717139115

Related Stories

Machine learning techniques show promise for supporting medical decisions

February 28, 2018
Several studies being presented at the American College of Cardiology's 67th Annual Scientific Session demonstrate how the computer science technique known as machine learning can be used to accurately predict clinical outcomes ...

Computers trounce pathologists in predicting lung cancer type, severity, researchers find

August 16, 2016
Computers can be trained to be more accurate than pathologists in assessing slides of lung cancer tissues, according to a new study by researchers at the Stanford University School of Medicine.

Scientists target glioma cancer stem cells, which could improve patient survival

February 5, 2018
Glioblastoma multiforme is the most common and aggressive primary brain tumor and has one of the worst survival rates of all cancers. Despite surgery, radiation and chemotherapy, these tumors virtually always become resistant ...

Study redefines brain tumor diagnosis and treatment

June 10, 2015
Not all brain cancers are the same but together they represent a deadly disease that has been difficult to identify and treat. Scientists at multiple institutions have found a new way of classifying brain cancers that could ...

Will 'AI' be part of your health-care team?

December 12, 2017
(HealthDay)—Artificial intelligence is assuming a greater role in many walks of life, with research suggesting it may even help doctors diagnose disease.

Detroit patients' contributions to national study re-define low-grade brain tumor diagnosis

June 24, 2015
Sixty-seven patients from the Hermelin Brain Tumor Center at Henry Ford Hospital and their families made important contributions to a national cancer study that proposes a change in how some brain tumors are classified - ...

Recommended for you

A bad influence—the interplay between tumor cells and immune cells

October 16, 2018
Research at Huntsman Cancer Institute (HCI) at the University of Utah (U of U) yielded new insights into the environment surrounding different types of lung tumors, and described how these complex cell ecosystems may in turn ...

Student develops microfluidics device to help scientists identify early genetic markers of cancer

October 16, 2018
As anyone who has played "Where's Waldo" knows, searching for a single item in a landscape filled with a mélange of characters and objects can be a challenge. Chrissy O'Keefe, a Ph.D. student in the Department of Biomedical ...

Technique to 'listen' to a patient's brain during tumour surgery

October 16, 2018
Surgeons could soon eavesdrop on a patient's brain activity during surgery to remove their brain tumour, helping improve the accuracy of the operation and reduce the risk of impairing brain function.

Researchers elucidate roles of TP63 and SOX2 in squamous cell cancer progression

October 16, 2018
Squamous cell carcinomas (SCCs) are aggressive malignancies arising from the squamous epithelium of various organs, such as the esophagus, head and neck, lungs, and skin. Previous studies have demonstrated that two master ...

Function of neutrophils during tumor progression unraveled

October 15, 2018
Researchers at The Wistar Institute have characterized the function of neutrophils, a type of white blood cells, during early stages of tumor progression, showing that they migrate from the bone marrow to distant sites and ...

Delving where few others have gone, leukemia researchers open new path

October 15, 2018
A Wilmot Cancer Institute study uncovers how a single gene could be at fault in acute myeloid leukemia (AML), one of the deadliest cancers. The breakthrough gives researchers renewed hope that a gene-targeted therapy could ...

0 comments

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.