федеральное государственное автономное образовательное учреждение высшего образования
«Самарский национальный исследовательский университет имени академика С.П. Королева»
Lasers and Neural Networks to Predict Severity of Acne and Other Diseases

Lasers and Neural Networks to Predict Severity of Acne and Other Diseases

Самарский университет

Solution by Samara scientists will help assess skin condition in less than 10 minutes

08.12.2022 2022-12-14
Within a year, scientists at Samara National Research University will develop a technology for noninvasive automatic analysis of skin condition and design a portable unit that can quickly and painlessly detect serious pathologies, such as neoplasms, and diagnose and analyze problems usually considered cosmetic — acne, post-acne, rosacea, skin dehydration, and others. 

The project has won the All-Russian UMNIK-Photonics competition and received financial support from the Innovation Promotion Foundation. The development is carried out by employees of the Department of Laser and Biotechnical Systems as part of a research group led by Professor Valery Zakharov, Doctor in Physical and Mathematical Science. Consultative support is provided by specialists of Samara State Medical University clinics and the Samara Oncology Center.
 
“In the future, the system will enable biochemical analysis of the patient’s skin, similar to the way medical labs now perform biochemical analysis of blood. Whereas currently skin analysis requires taking a sample, a scrape, or a biopsy, the future procedure will be non-invasive, without disturbing the skin integrity. The patient comes in, the doctor measures the spectrum of a certain skin area, the program analyzes the input and issues a printout with concentrations of skin components and a preliminary diagnosis. This can be used both for diagnosing neoplasms and in cosmetology, for determining general characteristics of skin condition (skin type, moisture saturation) and diagnosis of various pathological conditions (acne, post-acne, rosacea, hyperpigmentation, dehydration, and so on). The analysis accuracy will be from 75 to 99 %,” says project author Irina Matveeva, postgraduate student at the Institute of Informatics and Cybernetics at Samara University.
 
The technology will be based on Raman spectroscopy, which has proven high accuracy in the study of epidermis and upper layers of the dermis. The laser beam is directed at the patient’s skin; changes in the scattered emission spectrum correspond to the presence of certain substances in the skin.
 
Previously this method was used by scientists to diagnose skin neoplasms and kidney diseases. Samara scientists want to expand the field of application of Raman spectroscopy and introduce this method into therapeutic practice. Processing of the results of spectroscopy and diagnosis can be made with the help of convolutional neural networks and appropriate software.
 
“Each peak in the Raman spectrum corresponds to a specific chemical compound. Thus, by studying the recorded spectrum, one can obtain a rather comprehensive assessment of biochemical composition of the upper layer of the skin to the depth of a few millimeters where optical radiation with a wavelength of 785 nm can penetrate. Based on the information about biochemical composition of the epidermis and upper layers of the dermis, one can draw conclusions about the presence of pathologies, just as a doctor draws conclusions about diseases from a blood test. You can even predict, forecast the degree of severity of the disease by the concentration of certain substances in the skin,” emphasizes Irina Matveeva.
 
The project’s author also says that skin is the largest human organ by area prone to many different diseases and pathologies. According to the World Health Organization, residents of developed countries now visit a dermatologist 40 % more often than a decade ago. However, despite the increased demand, the existing methods in dermatology and cosmetology still do not allow for a quick and high-quality diagnosis. 
 
Most often the skin condition is diagnosed through visual examination by a dermatologist. The accuracy of this method is not high — only 60 to 75 %, depending on the qualifications and experience of the doctor. In addition, visual inspection does not allow to assess the skin biochemistry. With a dermatoscope, one can examine the structure of the skin more closely, almost like under a microscope, but unfortunately only the outside, not the inside. In addition, many pathological conditions have a visually very similar clinical picture, or even go unnoticed.
 
The gold standard of diagnosis is histopathological examination. However, high accuracy comes at a high price: the price of a complete set of histological equipment starts at several million rubles, and the cost of analysis is also high, since consumables are required. A histological examination takes several hours or even days. Another important aspect of the procedure for the patient is its invasiveness, that is, a piece of tissue is excised and taken. Therefore, the development of a new technology which will allow to analyze the biochemical composition of the patient’s skin quickly, accurately, painlessly and inexpensively, remains relevant.
 
“Our technology will produce the results more quickly — the examination will take less than 10 minutes, which means that the diagnosis can be done immediately, in just one appointment with a doctor. Besides, the skin integrity is not disturbed in any way, and the analysis does not require the use of any chemical agents and therefore has a low cost. As part of the project, we plan to develop software for analyzing Raman spectra, make a prototype of a portable spectroscopic unit, and test it,” says Irina Matveeva.
 
The unit size allows to easily place it on the table in the dermatologist’s office. For biochemical skin analysis, the unit has to be connected to a computer or laptop with the appropriate software developed by the project team. In the future, this skin composition biochemical analysis technology will be patented.
 
For reference
 
* Irina Matveeva is the author and co-author of more than 20 publications in leading Russian and international journals. In 2020, she became a finalist of the VII season of INRADEL international competition of innovative projects within the framework of the Microelectronics forum, and also took 3rd place in the III Sino-Russian competition of industrial innovations. In spring 2022, she won the competition for the assignment of monetary payments to young scientists and engineers working in the Samara region. In autumn 2022, she won a grant competition of Samara University for the development of startup projects and the regional Young Scientist competition. 
 
** All-Russian UMNIK-Photonics competition is organized to support projects of young scientists in the field of photonics, radiophotonics, optoelectronics and related fields in the interests of organizations of the Consortium of the Competence Center of the Photonics National Technological Initiative, as well as to encourage young scientists and specialists to create small innovative enterprises, necessary for commercialization of scientific development results.
 
*** Raman spectroscopy is a type of spectroscopy that uses inelastic scattering of monochromatic light (laser radiation) in the molecules under study. The spectrum of scattered radiation reveals spectral lines that are absent in the spectrum of primary (excitation) light. The number and location of the new spectral lines depend on the molecular structure of the substance under research, which enables detecting certain changes in the biological tissues.