Intersection of cosmetic technology and wellness
~Utilizing skin research techniques for stress management
Podium 55
Presented by: Tomonori Motokawa
Introduction
Recently, partially due to COVID19, wellness and wellbeing have become increasingly important parts of daily life. One major aspect is stress management, since stress can manifest as physical or mental illness and have negative effects on not only performance and feelings but also skin conditions. If people are aware of their own stress, they can manage their own wellbeing through rest or mental self-care. However, when they are in a chronic or high stress state, self-awareness is difficult, so self-care actions are not taken.
Routine stress status monitoring is desirable for stress management. However, there are various types of symptoms of stress (e.g. accumulation of mental and physical fatigue, stress markers in the blood, and changes in the autonomic nervous system), and the way they appear varies greatly depending on the individual, and it is difficult to know own stress state precisely. People often monitor their health status through health checks and questionnaires. However, the stress state cannot be easily comprehensively checked due to cost and time so the health check is only once or twice a year. Therefore, an easy, cost-effective, comprehensive stress status monitoring technology is needed.
For dealing with these issues, we wanted to leverage the knowledge and experience of the beauty industry since there are many reports that suggest correlations between stress state and facial/skin conditions. Therefore, in this study, we focused on using the facial/skin information as indicators of stress to develop a unique technique for assessing the stress state using deep learning technology.
With our technology, people will be aware of their stress in an easy and interesting way and self-care actions will be taken more often before severe health problems occur.
Methods
As facial/skin information, we collected data generally used by cosmetics researchers, such as subjective evaluation of the skin, macroscopic images of faces, and stratum corneum cell images. For various symptoms of stress, autonomic nervous system disorders, mental and physical fatigue accumulation, oxidative maker accumulation in blood and urine, and insomnia were set as prediction targets. We attempted to develop a technology to estimate these signs of stress accumulation using common facial and skin information. Information was collected from Japanese men and women from their 20s to 50s, with each group size varying depending on the skin data collected.
Furthermore, we developed a smartphone application that incorporates analytical technologies and suggests solutions that promote daily self-care, and then verified the utility of the system.
Results
As a result of the examination, we constructed the assessment technology from the following indicators with accuracy of ≥70%: (a) Estimation of autonomic nervous state of sympathetic/parasympathetic nerve balance and heart beat rate from facial image, (b) Estimation of Chalder Fatigue Scale, Mental Fatigue Scale and Physical Fatigue Scale result from subjective evaluation of skin, (c) Evaluation and correlation the stratum corneum cell image status with the Chalder Fatigue Scale and with the following blood and urine markers: Oxidative stress index, isoprostane, homovanillic acid.
Next, using technology (a) and (b), we developed a smartphone application prototype to assess stress symptoms, give recommendation of self-care action, and create customized self-care relaxation solutions (consisted of music, voice guide, vibration and picture). The prototype was tested for various lengths of time (one week or one month) and by both performing the easy stress status monitoring and experiencing the customized solution, not only improvements in positive mindset for self-care, self-compassion, reduction of stress states but also better skin conditions were observed.
Discussion/Conclusion
In this study, we identify a method of comprehensive stress monitoring by combining machine learning and facial/skin analysis technology that is commonly used in the cosmetics field. Our knowledge of stress relief also contributed to the development of the self-care solutions as well as the skin improvement. This is of great significance in the sense that it shows that cosmetics industry technology can be used for solutions for issues beyond just beauty.
Currently, we are continuing to create an application for smartphones that utilizes some of these estimation technologies to easily identify and manage stress conditions in an affordable and accessible manner. A new world, where we can perform not only skin care but also unbiased health monitoring and wellness self-care by using skin analyze technology, is emerging.
Recently, partially due to COVID19, wellness and wellbeing have become increasingly important parts of daily life. One major aspect is stress management, since stress can manifest as physical or mental illness and have negative effects on not only performance and feelings but also skin conditions. If people are aware of their own stress, they can manage their own wellbeing through rest or mental self-care. However, when they are in a chronic or high stress state, self-awareness is difficult, so self-care actions are not taken.
Routine stress status monitoring is desirable for stress management. However, there are various types of symptoms of stress (e.g. accumulation of mental and physical fatigue, stress markers in the blood, and changes in the autonomic nervous system), and the way they appear varies greatly depending on the individual, and it is difficult to know own stress state precisely. People often monitor their health status through health checks and questionnaires. However, the stress state cannot be easily comprehensively checked due to cost and time so the health check is only once or twice a year. Therefore, an easy, cost-effective, comprehensive stress status monitoring technology is needed.
For dealing with these issues, we wanted to leverage the knowledge and experience of the beauty industry since there are many reports that suggest correlations between stress state and facial/skin conditions. Therefore, in this study, we focused on using the facial/skin information as indicators of stress to develop a unique technique for assessing the stress state using deep learning technology.
With our technology, people will be aware of their stress in an easy and interesting way and self-care actions will be taken more often before severe health problems occur.
Methods
As facial/skin information, we collected data generally used by cosmetics researchers, such as subjective evaluation of the skin, macroscopic images of faces, and stratum corneum cell images. For various symptoms of stress, autonomic nervous system disorders, mental and physical fatigue accumulation, oxidative maker accumulation in blood and urine, and insomnia were set as prediction targets. We attempted to develop a technology to estimate these signs of stress accumulation using common facial and skin information. Information was collected from Japanese men and women from their 20s to 50s, with each group size varying depending on the skin data collected.
Furthermore, we developed a smartphone application that incorporates analytical technologies and suggests solutions that promote daily self-care, and then verified the utility of the system.
Results
As a result of the examination, we constructed the assessment technology from the following indicators with accuracy of ≥70%: (a) Estimation of autonomic nervous state of sympathetic/parasympathetic nerve balance and heart beat rate from facial image, (b) Estimation of Chalder Fatigue Scale, Mental Fatigue Scale and Physical Fatigue Scale result from subjective evaluation of skin, (c) Evaluation and correlation the stratum corneum cell image status with the Chalder Fatigue Scale and with the following blood and urine markers: Oxidative stress index, isoprostane, homovanillic acid.
Next, using technology (a) and (b), we developed a smartphone application prototype to assess stress symptoms, give recommendation of self-care action, and create customized self-care relaxation solutions (consisted of music, voice guide, vibration and picture). The prototype was tested for various lengths of time (one week or one month) and by both performing the easy stress status monitoring and experiencing the customized solution, not only improvements in positive mindset for self-care, self-compassion, reduction of stress states but also better skin conditions were observed.
Discussion/Conclusion
In this study, we identify a method of comprehensive stress monitoring by combining machine learning and facial/skin analysis technology that is commonly used in the cosmetics field. Our knowledge of stress relief also contributed to the development of the self-care solutions as well as the skin improvement. This is of great significance in the sense that it shows that cosmetics industry technology can be used for solutions for issues beyond just beauty.
Currently, we are continuing to create an application for smartphones that utilizes some of these estimation technologies to easily identify and manage stress conditions in an affordable and accessible manner. A new world, where we can perform not only skin care but also unbiased health monitoring and wellness self-care by using skin analyze technology, is emerging.