New NTU Singapore Studies Advance Cancer Screening and Early Detection

Researchers at Nanyang Technological University (NTU) in Singapore have published two new studies that could significantly improve cancer screening and early detection, potentially saving thousands of lives by identifying tumors at earlier, more treatable stages. The findings, published in recent scientific journals, highlight innovative approaches to liquid biopsy techniques and AI-driven imaging analysis, both of which address critical gaps in current diagnostic methods.

Advancing Liquid Biopsy for Early Cancer Detection

The first study, published in Nature Cancer, introduces a breakthrough in liquid biopsy technology—a method that detects cancer by analyzing circulating tumor DNA (ctDNA) in blood samples. The NTU team developed a highly sensitive assay capable of identifying multiple cancer types, including lung, breast and colorectal cancers, with a detection rate of over 90% in early-stage patients.

Current liquid biopsy methods often struggle with sensitivity, particularly in early-stage cancers where tumor DNA fragments are scarce. The NTU researchers overcame this challenge by combining single-molecule sequencing with machine learning algorithms to amplify and analyze minuscule amounts of ctDNA. According to the study, their technique achieved a false-positive rate of less than 5%, making it far more reliable than existing tests.

“This is a game-changer for cancer screening,” said Dr. Lim Wei, lead author of the study and a senior researcher at NTU’s Lee Kong Chian School of Medicine. “Early detection is the key to improving survival rates, and our method could enable population-wide screening programs that are both affordable and accurate.”

The study also demonstrated the assay’s ability to distinguish between different cancer types, which could help clinicians tailor treatment plans more effectively. While the research is still in the validation phase, early results suggest it could be deployed in clinical settings within the next two to three years.

AI-Powered Imaging for High-Risk Tumor Identification

The second study, published in Radiology, focuses on an artificial intelligence (AI) model designed to analyze medical imaging—such as mammograms, CT scans, and MRIs—to identify high-risk tumors with greater precision than human radiologists. The NTU team trained the AI on a dataset of over 50,000 imaging studies, including cases from multiple global institutions, to recognize subtle patterns associated with aggressive cancers.

Traditional imaging often relies on radiologists’ subjective assessments, which can lead to missed diagnoses or unnecessary biopsies. The NTU AI model, however, achieved a 94% accuracy rate in identifying malignant tumors in breast cancer cases and an 89% accuracy rate in lung cancer detection. Importantly, it also reduced false positives by 30%, minimizing patient anxiety and unnecessary procedures.

Dr. Tan Mei Ling, co-author of the study and a radiology specialist at NTU, emphasized the model’s potential to address healthcare disparities. “In regions with limited access to specialized radiologists, AI-assisted imaging could bridge the gap and ensure more people receive timely and accurate diagnoses,” she said.

The AI tool is currently undergoing clinical trials in collaboration with Singapore’s National Cancer Centre and could be integrated into hospital workflows within the next 12 to 18 months, pending regulatory approval.

Broader Implications for Public Health

Both studies underscore the growing role of technology in transforming cancer care. Liquid biopsy methods could revolutionize screening programs, particularly for high-risk populations, while AI-driven imaging may reduce diagnostic errors and improve early intervention rates. Experts note that these advancements are particularly critical in low-resource settings, where access to specialized diagnostic tools is limited.

However, challenges remain. The liquid biopsy assay, while promising, requires further validation in diverse patient populations to ensure its effectiveness across different ethnicities and cancer subtypes. Similarly, the AI imaging model must be tested in real-world clinical environments to confirm its reliability in routine practice.

Dr. Ong Choon Nam, chair of NTU’s Cancer Science Institute, highlighted the importance of collaboration in advancing these technologies. “These innovations are not just about scientific breakthroughs—they’re about translating research into real-world impact,” he said. “We’re working closely with policymakers and healthcare providers to ensure these tools can be implemented equitably.”

What’s Next?

The NTU research teams are now focusing on scaling up their technologies for clinical use. The liquid biopsy assay is expected to enter Phase III trials later this year, while the AI imaging model will undergo expanded testing in multi-center studies. If successful, these tools could redefine cancer diagnostics globally, offering new hope for early detection and better patient outcomes.

For now, experts urge caution, emphasizing that while these studies are promising, they represent early steps in a long journey toward widespread adoption. Public health officials and clinicians will need to carefully evaluate their integration into existing healthcare systems to maximize their benefits.