Vithanage Erandi Kawshalya Madhushani Jade Times Staff
V.E.K. Madhushani is a Jadetimes news reporter covering Innovation.
Using AI to Combat Ovarian Cancer: A Deadly and Underfunded Challenge
Ovarian cancer, often referred to as “rare, underfunded, and deadly,” poses a significant challenge to early detection and treatment. By the time symptoms manifest, the cancer has often spread beyond the ovaries, dramatically reducing survival rates. Audra Moran, head of the Ovarian Cancer Research Alliance (Ocra), emphasizes the urgency of detecting ovarian cancer at least five years before symptoms arise to improve mortality rates.
New advancements in artificial intelligence (AI) are offering hope for earlier detection. Researchers are now leveraging AI in combination with innovative blood tests to identify subtle biomarkers, enabling detection of ovarian cancer in its earliest stages.
Breakthrough Technology: Nanotubes and AI for Early Detection
Dr. Daniel Heller, a biomedical engineer at Memorial Sloan Kettering Cancer Center, is at the forefront of this innovation. His team uses nanotubes microscopic tubes of carbon 50,000 times smaller than a human hair—capable of emitting fluorescent light. These nanotubes interact with molecules in blood samples to emit unique light signals, which AI decodes to identify patterns indicative of ovarian cancer.
“The patterns these sensors detect are far too subtle for the human eye to interpret,” says Dr. Heller. Machine-learning algorithms trained on data from patients with ovarian cancer analyze these signals, identifying patterns that suggest the presence of cancer.
While the early results are promising, a major challenge is the limited availability of data. Ovarian cancer is rare, and much of the necessary data is siloed in hospital systems, making it difficult for researchers to access. Despite training algorithms on data from just a few hundred patients, Dr. Heller’s system outperformed current biomarkers, demonstrating the immense potential of AI-driven diagnostics.
Dr. Heller’s ultimate goal is to create a tool that can triage gynecological diseases, helping doctors quickly differentiate between cancer and other conditions. He estimates that such a tool could be available within three to five years.
AI Speeds Up Testing for Deadly Infections
Beyond cancer detection, AI is also transforming how doctors diagnose and treat infections like pneumonia, which can be fatal for cancer patients. Pneumonia can be caused by over 600 pathogens, necessitating numerous tests to identify the cause an expensive and time-consuming process.
California-based Karius has developed an AI-powered blood test capable of identifying the specific pathogen causing pneumonia within 24 hours. By comparing patient samples to a vast database of microbial DNA containing billions of data points, the AI quickly pinpoints the infection, enabling faster and more accurate treatment.
“Previously, a pneumonia patient would undergo 15 to 20 different tests in their first week of hospitalization, costing about $20,000,” says Alec Ford, CEO of Karius. With the AI-driven test, hospitals can reduce costs and improve patient outcomes.
AI and the Quest to Decode Complex Biomarker Patterns
Researchers are also using AI to identify patterns in biomarkers for a wide range of diseases. Dr. Slavé Petrovski, a researcher at AstraZeneca, developed an AI platform called Milton, which analyzes biomarkers from the UK Biobank to accurately identify 120 diseases with a success rate of over 90%.
“These patterns are incredibly complex,” Dr. Petrovski explains. “Often, it’s not about one biomarker but the interplay of multiple biomarkers that only AI can decipher.”
Dr. Heller echoes this sentiment in his work on ovarian cancer, noting that while his nanotube sensors respond to proteins and molecules in the blood, the specific markers linked to cancer remain unknown. AI’s ability to uncover these patterns is transforming how diseases are diagnosed.
The Data Sharing Dilemma
Despite these advances, the lack of accessible data remains a significant barrier. Many hospitals and institutions are reluctant to share patient data, limiting the ability of researchers to train their algorithms effectively.
Ocra is addressing this challenge by funding a large scale patient registry that includes electronic medical records from patients who have consented to share their data for research purposes.
“It’s still early days for AI in medical research,” says Moran. “We’re navigating uncharted territory, but the potential is extraordinary.”
The Future of AI in Medicine
As AI continues to advance, it holds the promise of revolutionizing early disease detection and treatment. From ovarian cancer to deadly infections, these technologies are poised to save lives by enabling faster, more accurate diagnostics. While challenges such as data accessibility and algorithm refinement remain, the breakthroughs achieved so far suggest a bright future for AI driven healthcare.
