Predicting human chronological age via AI analysis of dorsal hand versus facial images: A study in a cohort of Indian females.

Predicting a person's chronological age (CA) from visible skin features using artificial intelligence (AI) is now commonplace. Often, convolutional neural network (CNN) models are built using images of the face as biometric data. However, hands hold telltale signs of a person's age. To determine the utility of using only hand images in predicting CA, we developed two deep CNNs based on 1) dorsal hand images (H) and 2) frontal face images (F). Subjects (n = 1454) were Indian women, 20-80 years, across three geographic cohorts (Mumbai, New Delhi and Bangalore) and having a broad variation in skin tones. Images were randomised: 70% of F and 70% of H were used to train CNNs. The remaining 30% of F and H were retained for validation. CNN validation showed mean absolute error for predicting CA using F and H of 4.1 and 4.7 years, respectively. In both cases correlations of predicted and actual age were statistically significant (r(F) = 0.93, r(H) = 0.90). The CNNs for F and H were validated for dark and light skin tones. Finally, by blurring or accentuating visible features on specific regions of the hand and face, we identified those features that contributed to the CNN models. For the face, areas of the inner eye corner and around the mouth were most important for age prediction. For the hands, knuckle texture was a key driver for age prediction. Collectively, for AI estimates of CA, CNNs based solely on hand images are a viable alternative and comparable to CNNs based on facial images.

Temporal Variation of the Facial Skin Microbiome: A 2-Year Longitudinal Study in Healthy Adults.

BACKGROUND: The human skin microbiome is highly personalized, depending on, for example, body site, age, gender, and lifestyle factors. The temporal stability of an individual's skin microbiome-its resiliency and robustness over months and years-is also a personalized feature of the microbiome. The authors measured the temporal stability of the facial skin microbiome in a large cohort of subjects. In addition to measuring microbiome dynamics, they tracked facial skin condition using noninvasive, objective imaging and biophysical measures to identify significant facial features associated with temporal changes in microbiome diversity and composition. METHODS: The authors used 16S ribosomal RNA amplicon sequencing to track cheek and forehead skin microbiome diversity and composition annually over a 2-year period (2017-2019) in 115 healthy adult men and women. Skin metadata included facial features, such as wrinkles, hyperpigmentation, porphyrins, and skin color tone, as well as biophysical parameters for stratum corneum barrier function, pH, hydration, and elasticity. RESULTS: Across the subject population, the facial skin microbiome composition and diversity were relatively stable, showing minor variation over the 2-year period. However, for some subjects, composition, diversity, and relative abundance of specific organisms showed substantial changes from one year to the next, and these changes were associated with changes in stratum corneum barrier function and follicular porphyrins. CONCLUSIONS: For healthy people, facial skin microbiome diversity and composition are relatively stable from year to year. Tracking the temporal changes in the microbiome along with skin phenotypic changes allows for a deeper understanding of the skin microbiome's role in health and disease. These results should be helpful in the design of longer-term intervention trials with microbiome-based skin care treatments.

Rosmarinus officinalis L. (Rosemary) Extracts Containing Carnosic Acid and Carnosol are Potent Quorum Sensing Inhibitors of Staphylococcus aureus Virulence.

Staphylococcus aureus is an opportunistic pathogen and a common cause of skin infection. S. aureus also plays a role in the pathogenesis of the chronic inflammatory skin disease, atopic dermatitis. S. aureus virulence involves activation of the quorum sensing agr operon. In this paper, we show that the diterpene carnosic acid, present in R. officinalis L. (rosemary) leaves, is a specific inhibitor of S. aureus agr expression as low as 5 µM. Carnosol and rosmarinic acid are two other phytochemicals present in rosemary leaves. Carnosol, but not rosmarinic acid, is also a potent agr expression inhibitor. Natural rosemary extracts containing carnosic acid and carnosol inhibit S. aureus agr expression, both in luciferase reporter strains and in wild type strains isolated from patients with atopic dermatitis. Specific inhibition of S. aureus virulence using topical formulations of rosemary extract may offer a practical approach to preventing and treating flares of atopic dermatitis.

New Insights into the Intrinsic and Extrinsic Factors That Shape the Human Skin Microbiome.

Despite recognition that biogeography and individuality shape the function and composition of the human skin microbiome, we know little about how extrinsic and intrinsic host factors influence its composition. To explore the contributions of these factors to skin microbiome variation, we profiled the bacterial microbiomes of 495 North American subjects (ages, 9 to 78 years) at four skin surfaces plus the oral epithelium using 16S rRNA gene amplicon sequencing. We collected subject metadata, including host physiological parameters, through standardized questionnaires and noninvasive biophysical methods. Using a combination of statistical modeling tools, we found that demographic, lifestyle, and physiological factors collectively explained 12 to 20% of the variability in microbiome composition. The influence of health factors was strongest on the oral microbiome. Associations between host factors and the skin microbiome were generally dominated by operational taxonomic units (OTUs) affiliated with the Clostridiales and Prevotella A subset of the correlations between microbial features and host attributes were site specific. To further explore the relationship between age and the skin microbiome of the forehead, we trained a Random Forest regression model to predict chronological age from microbial features. Age was associated mostly with two mutually coexcluding Corynebacterium OTUs. Furthermore, skin aging variables (wrinkles and hyperpigmented spots) were independently correlated to these taxa.IMPORTANCE Many studies have highlighted the importance of body site and individuality in shaping the composition of the human skin microbiome, but we still have a poor understanding of how extrinsic (e.g., lifestyle) and intrinsic (e.g., age) factors influence its composition. We characterized the bacterial microbiomes of North American volunteers at four skin sites and the mouth. We also collected extensive subject metadata and measured several host physiological parameters. Integration of host and microbial features showed that the skin microbiome was predominantly associated with demographic, lifestyle, and physiological factors. Furthermore, we uncovered reproducible associations between chronological age, skin aging, and members of the genus Corynebacterium Our work provides new understanding of the role of host selection and lifestyle in shaping skin microbiome composition. It also contributes to a more comprehensive appreciation of the factors that drive interindividual skin microbiome variation.

Race, rather than skin pigmentation, predicts facial hair growth in women.

BACKGROUND: The Ferriman-Gallwey hirsutism score is the currently accepted standard for assessing excess hair growth that may indicate hyperandrogenicity. The score was originally based on 60 Caucasian women, and recent studies suggest that it may need modification to be used in other populations. OBJECTIVES: To investigate ethnic, racial, and pigmentary variations in hair growth of the upper lip in diverse multinational populations. RESULTS: Variations in hair growth of the upper lip were significantly related to self-reported ethnicity and race. In a logistic regression with racial groups and skin lightness, all racial groupings (African American, Hispanic, Asian Indian, and East Asian) were predictive of hair growth (all p<0.0001), but skin lightness was not (all p>0.05). CONCLUSION: The observed differences in constitutive hair growth illustrate the need to develop an ethnically stratified visual scoring method to more accurately characterize the severity of excess hair growth.

Sunlight exposure may be a risk factor of hearing impairment: a community-based study in Japanese older men and women.

BACKGROUND: Oxidative stress is an important factor in the pathology of age-related hearing loss. Recent animal studies have reported that ultraviolet radiation in sunlight is related to systemic induction of oxidative stress. Chronic sun exposure leads to photodamaged skin, which is manifested as facial skin wrinkling and hyperpigmentation. We hypothesized that sunlight exposure, as assessed by the severity of facial skin photodamage, might be associated with hearing impairment through an oxidative stress mechanism. To examine this, we performed a cross-sectional analysis by using the baseline data from a community-based cohort study of older Japanese. METHODS: A total of 805 residents (342 men and 463 women) aged 65 years or older living in Kurabuchi Town, Gunma prefecture, Japan, were examined between 2005 and 2006. Facial skin condition was quantified by image analysis of standardized facial images. Hearing impairment was defined as a failure to hear a 30-dB signal at 1 kHz and a 40-dB signal at 4 kHz in the better ear in pure-tone audiometric tests. RESULTS: In men, facial wrinkle was positively associated with hearing impairment (for highest vs lowest: multivariate-adjusted odds ratio, 2.16; 95% confidence interval, 1.00-4.66; p for trend = .01). Stratified analysis by age, educational level, smoking status, sunscreen or foundation use, and diabetes showed results similar to those for men as a whole. This association was particularly pronounced in men with the low levels of antioxidants and without occupational noise exposure. We observed no apparent association in women. CONCLUSIONS: The results support the hypothesis that chronic sun exposure is a risk factor of hearing impairment.

Capturing skin properties from dynamic mechanical analyses.

BACKGROUND/PURPOSE: Existing skin mechanical testing devices focus on measuring skin elasticity and are not tailored to assess the dynamic behavior of skin. Additionally, the mathematical techniques used in existing devices are often not optimal. METHODS: A new dynamic mechanical device that measures the linear dynamics of skin was developed and tested. The mechanical properties of skin were evaluated in experiments in which stiffness and damping parameters were measured (i) at different locations on the arm and hand, (ii) when stratum corneum hydration was varied by controlled changes in environmental humidity, and (iii) following the application of film-forming polymers. Parallel measurements were made with the Cutometer so that the two devices could be compared. RESULTS: The findings revealed that reliable and valid measurements of skin mechanical properties can be obtained from the device. The stiffness of the skin was shown to vary significantly as a function of skin site, changes in stratum corneum hydration and following application of the polymer films. Changes in the damping parameter were less consistently associated with varying the condition of the skin. CONCLUSION: The high reliability and speed of measurement make this device and analytic procedure an attractive option for testing skin mechanics.

Acne vulgaris in women: prevalence across the life span.

BACKGROUND: Acne vulgaris is a common skin disease with a large quality of life impact, characterized by comedones, inflammatory lesions, secondary dyspigmentation, and scarring. Although traditionally considered a disease of adolescence, reports suggest it is also a disease of adults, especially adult women. Our objectives were to determine acne prevalence in a large, diverse group of women and to examine acne by subtype and in relation to other skin findings, measurements, and lifestyle factors. METHODS: We recruited 2895 women aged 10-70 from the general population. Photographs were graded for acne lesions, scars, and dyspigmentation. Measurements were taken of sebum excretion and pore size, and survey data were collected. RESULTS: Of the women studied, 55% had some form of acne: 28% had mild acne, and 27% had clinical acne, 14% of which was primarily inflammatory and 13% of which was primarily comedonal. Acne peaked in the teenage years, but 45% of women aged 21-30, 26% aged 31-40, and 12% aged 41-50 had clinical acne. Women with inflammatory acne were younger than those with comedonal acne (p≤0.001), and postmenopausal women had less acne than age-matched peers (p<0.0001). Acne was associated with facial hirsutism (p=0.001), large pores (p=0.001), and sebum excretion (p=0.002). Smokers had more, primarily comedonal, acne than nonsmokers. CONCLUSIONS: The cross-sectional design precludes conclusions about progression of acne with age. Participation was restricted to women. The photographic nature of the study imposes general limitations. Techniques used in this study were not sufficiently sensitive to identify cases of subclinical acne. More than a quarter of women studied had acne, which peaked in the teens but continued to be prevalent through the fifth decade.

Lifestyle factors and visible skin aging in a population of Japanese elders.

BACKGROUND: The number of studies that use objective and quantitative methods to evaluate facial skin aging in elderly people is extremely limited, especially in Japan. Therefore, in this cross-sectional study we attempted to characterize the condition of facial skin (hyperpigmentation, pores, texture, and wrinkling) in Japanese adults aged 65 years or older by using objective and quantitative imaging methods. In addition, we aimed to identify lifestyle factors significantly associated with these visible signs of aging. METHODS: The study subjects were 802 community-dwelling Japanese men and women aged at least 65 years and living in the town of Kurabuchi (Takasaki City, Gunma Prefecture, Japan), a mountain community with a population of approximately 4800. The facial skin condition of subjects was assessed quantitatively using a standardized facial imaging system and subsequent computer image analysis. Lifestyle information was collected using a structured questionnaire. The association between skin condition and lifestyle factors was examined using multivariable regression analysis. RESULTS: Among women, the mean values for facial texture, hyperpigmentation, and pores were generally lower than those among age-matched men. There was no significant difference between sexes in the severity of facial wrinkling. Older age was associated with worse skin condition among women only. After adjusting for age, smoking status and topical sun protection were significantly associated with skin condition among both men and women. CONCLUSIONS: Our study revealed significant differences between sexes in the severity of hyperpigmentation, texture, and pores, but not wrinkling. Smoking status and topical sun protection were significantly associated with signs of visible skin aging in this study population.

Development of a digital imaging system for objective measurement of hyperpigmented spots on the face.

BACKGROUND/AIMS: There are few available methods that can be used to quantify hyperpigmented spots on a wide area of the face. The objective of this study was to develop such a method through the use of specialized image analysis technologies. METHODS: This imaging system was composed of a source of illumination whose light intensity was controlled with a dimmer, a 3-CCD video camera connected to a computer, and a positioning device used to correctly align the subject's face. This system was calibrated by adjusting the light intensity, the camera position, and white balance of the camera in order to acquire reproducible images. Using a specific algorithm for the image analysis, this system enabled us to measure both the total area of hyperpigmented spots (mm2) and the averaged skin colour tone (quasi L*a*b*) excluding the area of those hyperpigmented spots in a wide area of the face. The accuracy and reproducibility of the system was validated using a mannequin head with six standard colour chips obtained from the GretagMacbeth ColorChecker, and brown-coloured patches that simulated hyperpigmented spots whose colour and area were both known. The correlation between CIE L*a*b* and quasi L*a*b* values was examined by conducting simultaneous measurements of the facial skin colour of 187 subjects with a tristimulus colourimeter (Minolta Chromameter) and our imaging system. RESULTS: The measurement errors in quasi L*a*b* values of colour chips and the area of brown patches were less than 2 and 5%, respectively, unless these chips or patches were located in the peripheral zone of the mannequin head. The variation in quasi L*a*b* values and the area of hyperpigmented spots (mm2) in five repeated measurements performed once every hour was less than 2%. There was an excellent correlation between the CIE L*a*b* and quasi L*a*b* values, and the Pearson's correlation coefficient between CIE L* and quasi L* value, for instance, was 0.908. CONCLUSIONS: : As long as the region to be evaluated is limited to the cheek and periorbital areas, this system enables automatic detection of hyperpigmented spots in a wide area of the face, as well as the correct measurement of those areas and determination of skin colours.

Utilization of a high-resolution digital imaging system for the objective and quantitative assessment of hyperpigmented spots on the face.

BACKGROUND/AIMS: The aim of this study was to quantify and confirm the efficacy of cosmetic formulations for hyperpigmented spots over a wide area of the face using a high quality digital imaging system that we developed. METHODS: A total of 120 Japanese female volunteers aged 25-60 years with solar lentigines were treated for 6 months with a skin lightening moisturizer (SLM, thereafter) containing 3% magnesium ascorbyl phosphate on one side of the face and vehicle on the other side. During the course of the study, facial images were collected by the image analysis to measure facial skin colour and the total area of hyperpigmented spots. The evaluation was also conducted by visual grading. Measurements were made before and 1, 3, and 6 months after starting the application, and again 6 months after discontinuing the treatment. Three similar clinical studies using the same protocol were repeated for up to one-month to confirm the reproducibility of the results and to examine seasonal variation. RESULTS: SLM significantly reduced the total area of hyperpigmented spots (P < 0.005) after one month of treatment compared to the vehicle, with no significant variation in facial skin colour tone in the areas outside the hyperpigmented spots. The results of the visual grading were consistent with those obtained by image analysis. The total area of hyperpigmented spots 6 months after discontinuing the treatment had returned to pre-treatment levels. The reproducibility of these clinical results was demonstrated in three follow-up studies. CONCLUSIONS: A high-resolution digital imaging method, combined with a split-face clinical protocol is sensitive enough to prove that SLM readily reduces hyperpigmented spots, while maintaining normal facial skin colour.