A new application that uses artificial intelligence could revolutionize the way scientists study hair and could lead to the development of health diagnoses based solely on hair.
The artificial intelligence model accelerates and simplifies the hair quantification process, allowing the microscope to scan slides and collect images of hundreds of hairs at a time. In seconds, it can capture large amounts of high-resolution data and then process it using deep learning algorithms to collect the color, shape, width and length of each hair. Researchers tested it using mouse fur, but it can be applied to the hair of any species, including humans.
The research behind the app was conducted and developed by scientists at the Washington State University School of Veterinary Medicine and is published in the Journal of Dermatology Research.
“In many ways, a person’s hair reflects health in some way, and if you start separating them with tweezers, as many hair scientists do, you can make some very interesting discoveries, but you did it manually, right under a microscope,” said Ryan Driskell, associate professor and lead researcher on the study. “So, our idea was: What would happen if you could get a computer program to do this for you?”
The concept for the application was proposed by Jasson Makkar, a graduate student in molecular biology at Washington State University, whose task was the monotonous task of manually separating thousands of hairs in various research projects in Driskell’s laboratory that focused on hair and skin.
To turn this idea into reality, Makkar used WSU’s high-performance computing cluster Kamiak to train an AI computer vision model to recognize hair. With additional help from the Aperio GT450 microscope at the Washington Laboratory for Animal Disease Diagnosis, high-resolution imaging of hair fibers has been automated.
Makkar said the application has many implications, including forensic science and the hair products industry, but allowing scientists to assess the health of a person or animal through hair may be the most important.
He said that by identifying longitudinal data points for healthy hair for each species, a scale could be created for human doctors and veterinarians to rate overall health based on hair. Different conditions, such as hormonal imbalances or nutritional deficiencies, can alter hair growth in ways that can be detected and may be used for diagnosis.
The new technology not only identifies the species of hair, but also reveals the age, health and ethnicity of humans, which could help criminal investigations.
“Law enforcement agencies have this method, using hair fiber classification as a forensic tool in criminal investigations,” Dricker said. “This method is somewhat controversial because most of the work is done by forensic technicians who visually identify hair types found at crime scenes and then cross-reference them with a limited database of all mammalian hair types.”
Driscoll added that the technology not only allows scientists to cross-reference hair fibers with high accuracy in an impartial manner, but also generates a database large enough to accurately quantify hair types in different individuals and possible anatomical locations.
Makkar said using these same tools to evaluate the impact of various hair care products on hair is another feature brought by the app.
“Take a piece of hair, apply the makeup you are testing, and then use our deep hair phenotyping tool to observe it and see how it changes,” Marca said.
The data generated for this study is available through interactive web tools at Skinregenesis. org.
For more content, you can click on the link below the video
Thank you for watching this video. If you like it, please subscribe and like it. thank
Original text:https://medicalxpress.com/news/2024-09-ai-hair-analysis-method-health.html
More information: Jasson Makkar et al.,”In-depth publication of typology: Impact on Endocrinology, Development, and Aging”, Journal of Dermatology Research (2024). DOI：10.1016/j.jid.2024.08.014
Journal Information: Journal of Dermatology Research
Provided by Washington State University
Oil tubing: