AI Blood Test Using Infrared Light Detects Cancer with High Precision
Cambridge, Thursday, 10 April 2025.
An AI-powered blood test using infrared light detects cancer ‘fingerprints’ with up to 81% accuracy, marking a breakthrough in early cancer diagnostics.
Groundbreaking Research at Max Planck Institute
Scientists at the Max Planck Institute of Quantum Optics in Germany, led by Dr. Michaela Žigman, have developed a revolutionary blood test that employs electric-field molecular fingerprinting (EMF) technology [1]. The research, published in ACS Central Science on April 9, 2025, analyzed blood plasma samples from 2,533 participants, including patients with various types of cancer [2]. The study demonstrates particularly promising results for lung cancer detection, achieving an accuracy rate of up to 81% [1][3].
Technical Innovation in Cancer Detection
The technology utilizes pulsed infrared light to identify specific molecular patterns in blood plasma, offering a less invasive alternative to traditional tissue biopsies [3]. The AI model was trained using blood samples from over 2,100 individuals, including patients with lung, prostate, breast, and bladder cancer who had not yet undergone treatment [4]. While the results for lung cancer detection are impressive, the accuracy for other cancer types, such as breast cancer, currently stands at approximately 50%, indicating areas for future improvement [4].
Broader Implications for Cancer Diagnostics
This development comes at a time when AI-powered diagnostic tools are showing increasing promise across various medical fields. For context, a 2017 study demonstrated that AI models could outperform dermatologists in skin cancer diagnosis [5]. Dr. Žigman emphasizes that this new framework shows robust potential for disease phenotyping under real-world conditions [2]. The researchers are now planning to expand their studies to include a wider range of cancers and larger patient populations to enhance the technology’s diagnostic capabilities [3][4].
Future Prospects and Limitations
While the current results are promising, the technology still faces some challenges [alert! ‘ongoing validation needed for various cancer types’]. The research team acknowledges the need for improving measurement reproducibility and further validation in additional patient populations [2]. The development of new blood tests for various other cancers, including pancreatic, breast, and stomach cancer, is currently underway [4], suggesting a broader transformation in cancer diagnostics is on the horizon.