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.

Human Skin Gloss Perception Based on Texture Statistics.

We propose objective, image-based techniques for quantitative evaluation of facial skin gloss that is consistent with human judgments. We use polarization photography to obtain separate images of surface and subsurface reflections, and rely on psychophysical studies to uncover and separate the influence of the two components on skin gloss perception. We capture images of facial skin at two levels, macro-scale (whole face) and meso-scale (skin patch), before and after cleansing. To generate a broad range of skin appearances for each subject, we apply photometric image transformations to the surface and subsurface reflection images. We then use linear regression to link statistics of the surface and subsurface reflections to the perceived gloss obtained in our empirical studies. The focus of this paper is on within-subject gloss perception, that is, on visual differences among images of the same subject. Our analysis shows that the contrast of the surface reflection has a strong positive influence on skin gloss perception, while the darkness of the subsurface reflection (skin tone) has a weaker positive effect on perceived gloss. We show that a regression model based on the concatenation of statistics from the two reflection images can successfully predict relative gloss differences.

A comparative study of the effects of retinol and retinoic acid on histological, molecular, and clinical properties of human skin.

BACKGROUND: All-trans retinol, a precursor of retinoic acid, is an effective anti-aging treatment widely used in skin care products. In comparison, topical retinoic acid is believed to provide even greater anti-aging effects; however, there is limited research directly comparing the effects of retinol and retinoic acid on skin. OBJECTIVES: In this study, we compare the effects of retinol and retinoic acid on skin structure and expression of skin function-related genes and proteins. We also examine the effect of retinol treatment on skin appearance. METHODS: Skin histology was examined by H&E staining and in vivo confocal microscopy. Expression levels of skin genes and proteins were analyzed using RT-PCR and immunohistochemistry. The efficacy of a retinol formulation in improving skin appearance was assessed using digital image-based wrinkle analysis. RESULTS: Four weeks of retinoic acid and retinol treatments both increased epidermal thickness, and upregulated genes for collagen type 1 (COL1A1), and collagen type 3 (COL3A1) with corresponding increases in procollagen I and procollagen III protein expression. Facial image analysis showed a significant reduction in facial wrinkles following 12 weeks of retinol application. CONCLUSIONS: The results of this study demonstrate that topical application of retinol significantly affects both cellular and molecular properties of the epidermis and dermis, as shown by skin biopsy and noninvasive imaging analyses. Although the magnitude tends to be smaller, retinol induces similar changes in skin histology, and gene and protein expression as compared to retinoic acid application. These results were confirmed by the significant facial anti-aging effect observed in the retinol efficacy clinical study.