Artificial intelligence (AI) is transforming medical research by analyzing blood tests to detect diseases early. This innovation offers hope for diagnosing deadly illnesses like ovarian cancer and speeding up other critical tests.
Tackling Ovarian Cancer with AI
Ovarian cancer is rare, underfunded, and lethal. Audra Moran, head of the Ovarian Cancer Research Alliance (Ocra), emphasizes that early detection is crucial. Most ovarian cancers start in the fallopian tubes and may spread before symptoms appear. Detecting it five years before symptoms could significantly impact survival rates.
AI-powered blood tests are emerging to identify ovarian cancer in its early stages. These tests analyze complex patterns in blood that humans cannot detect. Dr. Daniel Heller, a biomedical engineer at Memorial Sloan Kettering Cancer Center, leads research using nanotubes to enhance detection. Nanotubes—50,000 times smaller than a human hair—emit specific light wavelengths based on molecules in blood samples.
Machine-learning algorithms decode these subtle patterns, identifying cancer markers with greater accuracy than existing methods. However, challenges remain. Ovarian cancer’s rarity limits available training data, and hospitals often silo patient data. Despite these hurdles, AI systems are improving, with potential for wider adoption after further studies.
Dr. Heller envisions a future where AI tools quickly assess all gynecological diseases. He estimates practical applications could emerge within three to five years.
Speeding Up Blood Tests for Deadly Infections
AI isn’t limited to cancer detection. It also accelerates tests for infections like pneumonia. Pneumonia, often fatal for cancer patients, is caused by over 600 pathogens, requiring numerous diagnostic tests. California-based company Karius uses AI to identify the exact pathogen within 24 hours, selecting the right antibiotic.
According to Alec Ford, Karius’ CEO, traditional testing costs about $20,000 in the first week of hospitalization. AI-enabled tests compare patient samples against a vast microbial DNA database, streamlining diagnosis. Such rapid identification would have been impossible without AI.
Dr. Slavé Petrovski, a researcher at AstraZeneca, has also developed an AI platform called Milton. Using biomarkers from the UK Biobank, Milton identifies 120 diseases with over 90% accuracy. This pattern recognition is uniquely suited to AI, which can detect complex relationships in massive datasets.
Challenges for AI in Medical Research & Future Potential
Despite AI’s promise, challenges remain. Researchers don’t fully understand the connections AI identifies between biomarkers and diseases. Moreover, data sharing is limited. Ms. Moran from Ocra notes, “People aren’t sharing their data, or there’s not a mechanism to do it.”
To address this, Ocra funds a large-scale patient registry with electronic medical records for algorithm training. As Ms. Moran puts it, “We’re still in the wild west of AI.”
AI’s ability to decode complex data patterns promises to revolutionize medicine, from early cancer detection to faster infection diagnoses. With continued research and collaboration, these breakthroughs could soon transform patient care worldwide.
