Sustainable medicine requires precision, and therefore effective, specific and sensitive diagnosis. This research, by biologists from the Ludwig-Maximilians University of Munich, summarizes these advances as follows: one drop of blood, several diagnoses. New technologies such as infrared light and machine learning now make it possible to comprehensively and efficiently assess human health in just a few minutes. With a saving of time and inconvenience for the patient, but also a considerable reduction in costs and a global and sustainable approach to their health.
These new technologies take on their full meaning with the aging of populations and the increasing prevalence of multimorbidity. A screening tool with which a single drop of blood is enough to provide comprehensive health information and which can detect several health problems with a single measurement therefore meets the needs of
a growing need for holistic diagnosis.
Infrared spectroscopy First of alla technique that uses infrared light to analyze the molecular composition of samples, has become a fundamental tool in chemistry in recent decades. The technology makes it possible to assign molecules as a “fingerprint”, delivered precisely by spectrometry. Applied to complex biofluids such as blood plasma, this physicochemical technique provides detailed data on molecular signals, making it a powerful medical diagnostic tool. However, until recent years, infrared spectroscopy, although more widely used, had not been integrated into the routine of medical diagnosis.
- Large research programs, such as the Kora program, carried out in Germany, have tested these techniques on the blood of thousands of participants, representative of a varied population: more than 5,000 blood plasma samples were evaluated using the infrared spectroscopy.
Machine learning nextallows molecular fingerprints to be analyzed and correlated with specific medical data. These fingerprints contain valuable information that the algorithm is now able to associate with different health conditions, including abnormal blood lipid levels, blood pressure abnormalities, prediabetes or type 2 diabetes. The algorithm is of course also capable of identifying healthy individuals but also those suffering from multiple pathologies and combined in different ways.
The combination of the 2 technologies, infrared spectroscopy and machine learning does not just identify one condition at a time: it accurately identifies a whole range of health problems in the same patient. In addition, the system makes it possible to predict, very early, the development of metabolic syndrome, years before the appearance of symptoms…
Infrared molecular fingerprints soon in routine screening
These technologies, which are more sustainable because they consume less resources, whether from the system or from patients, will soon allow doctors to detect and manage conditions more effectively. A particularly important point for metabolic disorders such as cholesterol abnormalities and diabetes, for which rapid and effective interventions can significantly improve outcomes.
A drop of blood for a growing diagnostic repertoire: as the system is refined, its capabilities expand and more and more comorbidities are added to the diagnostic repertoire. In practice, this will allow personalized health monitoring where soon everyone, from a simple home test, will be able to regularly monitor their condition and become the main actor in their own health.
In conclusion, the combination of infrared spectroscopy and machine learning should transform diagnostics, in a more sustainable and responsible way.
