We use cookies to understand how you use our site and to improve your experience. This includes personalizing content and advertising. To learn more, click here. By continuing to use our site, you accept our use of cookies. Cookie Policy.

Features Partner Sites Information LinkXpress hp
Sign In
Advertise with Us
LGC Clinical Diagnostics

Download Mobile App




AI Model Identifies Sex-Specific Risks Associated with Brain Tumors

By LabMedica International staff writers
Posted on 10 Oct 2024
Print article
Images show regions of glioblastoma tumors in females (top) and males (bottom). AI models predict presence of relatively higher risk (red) and lower risk (blue) areas (Photo courtesy of Tiwari Lab/UW–Madison)
Images show regions of glioblastoma tumors in females (top) and males (bottom). AI models predict presence of relatively higher risk (red) and lower risk (blue) areas (Photo courtesy of Tiwari Lab/UW–Madison)

For years, cancer researchers have observed that men are more likely than women to develop glioblastoma, a deadly and aggressive form of brain cancer with a median survival of just 15 months after diagnosis. Additionally, these tumors tend to be more aggressive in men. However, identifying specific characteristics that could help predict which tumors will grow faster has remained a challenge. Now, researchers are turning to artificial intelligence (AI) to uncover these risk factors and explore how they differ between men and women.

Scientists at the University of Wisconsin–Madison (Madison, WI, USA) are utilizing AI's computational capabilities to analyze large datasets of medical images, aiming to find patterns that could assist oncologists in making more informed decisions for their patients. Their goal is to address the entire range of challenges that cancer patients face, from diagnosis and prognosis to assessing treatment response. In this study, the researchers focused on digital images of pathology slides—thin sections of tumor samples—in an effort to detect patterns that might predict how fast a tumor could grow and, consequently, how long a patient might survive. Accurate prognosis is critical, as it influences treatment decisions and impacts patients’ quality of life after diagnosis.

To tackle this problem, the researchers developed an AI model capable of detecting subtle patterns in pathology slides that might be imperceptible to the human eye. They trained the model using data from over 250 glioblastoma studies, teaching it to recognize distinctive tumor features such as the abundance of certain cell types and the extent of the tumor’s invasion into nearby healthy tissue. Additionally, the model was trained to identify correlations between these features and patients' survival times, while also considering their sex. Through this approach, the team created an AI model that can identify risk factors for more aggressive tumors, with distinct patterns associated with each sex.

For women, higher-risk characteristics identified by the AI model included tumors that were infiltrating into healthy tissue. In men, the presence of pseudopalisading cells—cells that surround dying tissue—was linked to more aggressive tumors. The researchers' initial findings, published in Science Advances, revealed that the model also detected tumor traits associated with poorer prognoses for both sexes. The team is now extending their work to MRI data and has started using AI to analyze other cancers, such as pancreatic and breast cancers, with the goal of improving patient outcomes. This study could pave the way for more personalized treatment approaches for glioblastoma patients.

“There’s a ton of data collected in a cancer patient’s journey,” said radiology and biomedical engineering professor Pallavi Tiwari. “Right now, unfortunately, it’s usually studied in a siloed fashion, and this is where AI has huge potential. By uncovering these unique patterns, we hope to inspire new avenues for personalized treatment and encourage continued inquiry into the underlying biological differences seen in these tumors.”

New
Gold Member
Thyroid Stimulating Hormone Assay
TSH EIA 96 Test
Antipsychotic TDM AssaysSaladax Antipsychotic Assays
New
Food Allergens Assay Kit
Allerquant 14G A
New
Human Immunodeficiency Virus Assay
RealLine HIV Quantitative Kit

Print article

Channels

Clinical Chemistry

view channel
Image: The new saliva-based test for heart failure measures two biomarkers in about 15 minutes (Photo courtesy of Trey Pittman)

POC Saliva Testing Device Predicts Heart Failure in 15 Minutes

Heart failure is a serious condition where the heart muscle is unable to pump sufficient oxygen-rich blood throughout the body. It ranks as a major cause of death globally and is particularly fatal for... Read more

Hematology

view channel
Image: QScout CBC will give a complete blood count in 2 minutes from fingerstick or venous blood (Photo courtesy of Ad Astra Diagnostics)

Next Gen CBC and Sepsis Diagnostic System Targets Faster, Earlier, Easier Results

Every hour is critical in protecting patients from infections, yet there are currently limited tools to assist in early diagnosis before patients reach a hospital. The complete blood count (CBC) is a common... Read more

Immunology

view channel
Image: An immune response is initiated when an antigen-presenting cell (pink) presents foreign material to a T-cell (blue) (Photo courtesy of JAX)

Advanced Imaging Method Maps Immune Cell Connections to Predict Cancer Patients Survival

A growing tumor is influenced not only by the tumor cells themselves but also by the surrounding tissue, which alters its biology. Immune cells communicate by transferring vital signaling proteins to their... Read more

Microbiology

view channel
Image: The InfectoSynovia test has the potential to revolutionize the diagnosis of periprosthetic joint infection (Photo courtesy of 123RF)

High-Accuracy Bedside Test to Diagnose Periprosthetic Joint Infection in Five Minutes

Periprosthetic joint infection (PJI) represents a significant global issue that is worsening as the number of joint replacements increases due to aging populations. In the United States alone, the anticipated... Read more
Copyright © 2000-2024 Globetech Media. All rights reserved.