We don't see colors, we don't hear sounds, and yet the molecules in our blood speak a secret language. It is that of the imperceptible vibrations that occur when they are hit by flashes of infrared light. A language that could tell the most important story: that of cancer screening. Recently some scientists of theMax Planck Institute for Quantum Optics in Germany they have demonstrated that These “molecular fingerprints” can distinguish lung cancer patients from healthy subjects with an accuracy of up to 81%. A discovery that could inaugurate a new chapter in oncological diagnostics, faster and less invasive than current tissue biopsies.
Cancer Screening and the Secret Language of Molecules
When blood samples are exposed to pulses of infrared light from a laser, molecules in the plasma they start to vibrate. It's not poetry, it's pure physics: different components of the molecules absorb or reflect the energy of the light pulses and, as a result, emit their own characteristic light pattern that can be recorded and read as an “infrared molecular fingerprint.” The amazing thing? This fingerprint differs between cancer patients and healthy subjects. It fascinates me how science can exploit invisible phenomena to reveal such important truths.
A systematic review published ACS Central Science the 9 April opens up promising scenarios. The infrared-based cancer screening system joins a series of new blood tests currently under development to diagnose a plethora of cancers, including pancreatic, breast, and gastric cancers. It is true that we are still in the early stages of development, but the goal is clear: to detect cancer earlier than currently available methods, and to do so in a less invasive way than traditional tissue biopsies.
Cancer Screening, the Role of Artificial Intelligence and Future Challenges
To create this new blood test, Michaela Zhigman and colleagues first trained an AI model to detect differences in the fingerprint of plasma taken from more than 2.100 people. The sample included patients with lung, prostate, breast, or bladder cancer who had not yet been treated for their disease. The comparison? For each cancer patient, the researchers showed the AI a blood sample from a person of the same sex and similar age without cancer.
“Laser-based infrared molecular fingerprinting detects cancer, demonstrating its potential for clinical diagnostics,” said study co-author Michaela Zhigman.
After training their AI model with this data, the researchers tested how accurate it was at identifying cancer signatures in the plasma of about 430 people who were not included in the initial data. The result? The model achieved up to 81% accuracy in distinguishing plasma samples from lung cancer patients versus those without the disease. I can’t help but think about how many lives could be saved if we could perfect this technology.
Current limitations and prospects
The moon always has a dark side (I say this ironically, but the metaphor holds): the model has revealed itself much less effective at detecting the other three types of cancer included in the study. For example, he identified only about 50% of cases of breast cancer. Given these limitations, much more research will be needed before it can ever be used in clinical practice. Still, researchers are optimistic about its prospects. They now plan to train the model on a wider range of tumors, using data from more patients, to see if they can improve its accuracy in detecting cancer.
I like this mix of caution and hope that characterizes scientific research. Every great discovery always begins with small, uncertain steps. Cancer screening based on infrared molecular fingerprints may one day become routine in hospitals, or it may turn out to be a dead end. But each attempt brings us closer to understanding this complex disease. And in the meantime, we will continue to watch in wonder at that secret language that flows in our veins, just waiting to be deciphered.
