Previously submitted to: JMIR mHealth and uHealth (no longer under consideration since Jun 02, 2025)
Date Submitted: Jan 6, 2025
Warning: This is an author submission that is not peer-reviewed or edited. Preprints - unless they show as "accepted" - should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.
Evaluating AI-Driven Mobile Applications and Open-Source Software for Assessing Facial Growth and Development: A Retrospective Observational Study
ABSTRACT
Background:
Facial growth and development are dynamic processes influenced by intrinsic and extrinsic factors. Predicting these changes is essential in forensic science and clinical dentistry. While AI-driven age progression tools like Remini AI (a mobile application) and SAM (open-source software) may 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 (Remini AI) and an open-source software (SAM).
Methods:
This retrospective observational study used the FG-NET aging database and additional personal images. Remini AI and SAM processed 18 images each, generating 180 age-progressed images in total for predefined age groups (5, 11, 15, 18, and 26 years). Subjective assessments involved dental experts performing reordering, categorization, and accuracy tasks, while objective assessments used an age estimation model (MiVolo) to estimate the age of the AI-generated images. Statistical analyses included t-tests and chi-square tests with a significance level of 5%.
Results:
Remini AI achieved higher results for the categorization task (55%, P < .001), whereas SAM demonstrated superior accuracy in generating 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 younger (age 5) or older (age 26) age groups. SAM outperformed Remini AI in processing speed, generating images in an average of 9 seconds versus 5 minutes per image (P < .001).
Conclusions:
Remini AI excels in categorization tasks, while SAM demonstrates greater accuracy and efficiency for realistic age progressions, particularly in mid-childhood and young adulthood. Both tools require refinement for younger and older populations. These findings emphasize the potential of AI-driven tools in forensic investigations and clinical dentistry, including orthodontic planning and reconstructive procedures. Clinical Trial: This study is retrospective and observational, and no trial registration is applicable.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.