Artificial Intelligence-assisted Multimodal Quantification of Facial Pore Reduction After Microneedling With...

Enlarged facial pores are a common cosmetic concern, exacerbated by age-related collagen degradation and sebaceous hyperactivity. Despite clinical demand, objective quantification methods for pore improvement remain limited. This study used 3 independent analytical platforms-QuantifiCare 3-dimensional (3D) imaging, Gwyddion-based surface roughness analysis, and artificial intelligence (AI)-assisted image segmentation-to evaluate the efficacy of microneedling combined with adipose-derived stem cell exosome (ASCE+) therapy.

Thirty subjects underwent 3 sessions of microneedling with topical ASCE+ application. Standardized high-resolution facial images were obtained via the LifeViz Mini system under controlled lighting and positioning. QuantifiCare software generated regional pore scores across 6 facial zones. Gwyddion software quantified 3D surface roughness (R _z, µm) in 4 subunits using paired anatomical landmarks. AI-based contour segmentation assessed pore count and mean pore area (pixels²) in forehead, nose, malar, and buccal zones. Paired t tests were used for statistical comparisons.

QuantifiCare pore scores significantly increased across all regions (P < 0.001), reflecting improved skin texture. Gwyddion R _z values decreased by 43%-61%, indicating smoother surface topography. AI analysis showed marked pore size reduction (forehead, -58.2%; nose, -56.1%; malar region, -36.4%; P < 0.001), and modest reduction in the buccal zone (-10.7%, P = 0.0132). These multimodal findings consistently support enhanced dermal remodeling and reduced pilosebaceous prominence.

This is the first study to integrate 3D imaging, roughness mapping, and AI segmentation for pore analysis after microneedling and exosome therapy. The combined modality demonstrates significant improvement in pore architecture and skin texture, highlighting its utility in aesthetic dermatology.