Evaluation of artificial intelligence-driven mobile and open-source applications for facial growth assessment: A retrospective observational study

Background Facial growth and development are influenced by intrinsic and extrinsic factors, and predicting these changes is valuable in forensic science and clinical dentistry. While age-progression tools powered by Artificial Intelligence (AI), such as mobile applications, Remini AI, and open-source software, Style-Based Age Manipulation (SAM), offer innovative solutions, their clinical applicability remains underexplored. Objective To evaluate the accuracy and clinical feasibility of AI-driven age progression tools, using a mobile application and an open-source software. Methods This retrospective observational study used the Face and Gesture Recognition Network (FG-NET) aging database and personal photographs. Each tool processed 18 original photographs, generating 180 age-progressed outputs for target ages of 5, 11, 15, 18, and 26 years. Subjective assessments by dental experts included reordering, categorization, and accuracy tasks. Objective assessments used an age estimation model Multi Input VOLO (MiVolo), to estimate the age portrayed by the AI-generated images. Statistical analyses included t-tests and chi-square tests with a significance level of 5%. Results Remini AI achieved higher categorization accuracy (55%, P < .001), whereas SAM generated more realistic age progressions between 11 and 18 years old, with a lower mean absolute error (2.7 years vs. 3.7 years, P < .001). Neither tool performed effectively for the youngest (age 5) or oldest (age 26) groups. SAM generated images significantly faster than Remini AI, averaging 9 seconds versus 5 minutes per image (P < .001). Conclusions Remini AI excels in categorization, while SAM demonstrates greater accuracy and efficiency for mid-childhood and young adulthood. Both require refinement for younger and older populations, highlighting AI’s potential in forensic investigations and clinical dentistry.