Defesa – Dissertação de Arthur Endres Balbão

Aluno(a): Arthur Endres Balbão
Orientador(a): Marcelo Walter

Título: Spatiotemporal Simulation of Androgenetic Alopecia in Digital Humans
Linha de Pesquisa: Computação Gráfica e Visualização de Dados

Data: 10/09/2026
Hora: 09:00
Local: Esta banca ocorrerá de forma remota. Acesso público disponibilizado pelo link https://conferenciaweb.rnp.br/sala/marcelo-walter.

Banca Examinadora:
-Ricardo Guerra Marroquim (TU Delft)
-Eduardo Simões Lopes Gastal (UFRGS)
-Luiz Henrique de Figueiredo (IMPA)

Presidente da Banca: Marcelo Walter

Resumo: Androgenetic alopecia (AGA) is a prevalent condition characterized by distinct, gender-specific patterns of hair loss. Although AGA is well characterized in the biological and clinical literature, simulating progressive hair loss remains a challenge in computer graphics. We present a real-time, biologically grounded implementation of the Follicular Automaton Model (FAM), which simulates successive hair growth cycles for individual follicles. By adapting FAM to three-dimensional scalp geometry, our method enables the visual simulation of AGA progression over time. Our approach introduces a novel segmentation scheme that decomposes hair loss into three primary patterns: vertex, medial, and frontal. By modulating follicular miniaturization probabilities within these zones, we accurately reproduce the complex, clinically observed progression paths of hair loss for both genders and varied hair types. We validate our model primarily through qualitative assessment that demonstrates visual agreement with established clinical classification systems, such as the Norwood-Hamilton scale for men and the Ludwig scale for women. As a complementary consistency check, we assess whether the computational implementation reproduces the expected temporal dynamics of follicular phases. Beyond character rendering, our work offers potential applications in dermatological visualization.

Palavras-Chave: Alopecia, Visual Simulation, Computer Graphics