EyeModelHub
Home› Subspecialties› Oculoplastics› Model card
§Model Card

Facial Age & Attractiveness AI Platform

Automated estimation of facial attractiveness (1–10 scale) and apparent facial age from pre- and postoperative photographs

Balel Y. · 2023 · Turkey · doi:10.58600/eurjther1648
Open DOI
Measurement/Regression CNN External photo Internally validated
▸ More attributes
Anatomy · Periorbital soft tissueAnatomy · Facial aestheticConditions · BlepharoptosisSingle-centern = 541 patients · 1082 images images
§01Overview

Brief overview

This exploratory study used a deep evolutionary neural network for facial age estimation in patients undergoing orthognathic surgery. The model was trained on 66,200 images from the UTK, APPA, and IMDB datasets and applied to standardized pre- and postoperative facial photographs from 50 patients undergoing LeFort I and/or bilateral sagittal split osteotomy. The AI system estimated facial age from full-face and lateral-profile images, while human evaluators additionally scored facial aesthetics. The model achieved 52% categorical accuracy on its test set. Analyses explored relationships between cephalometric changes, perceived age, and aesthetic outcomes following surgery. No external validation, calibration analysis, code release, or clinical deployment was reported.

§02Performance
Headline performance

Age model categorical accuracy ≈51% on an internal test set.

Full metrics — as reported
Age model categorical accuracy ≈51% on an internal test set.
Aggregate postoperative changes: −1.9 years apparent age and +0.43 attractiveness points; improvements found in ~80% of images.
Calibration · Uncertainty
Not reported.
Bias · Fairness
Not reported.
§03Data & Validation
Study designRetrospective cohort
Center typeSingle-center
Patients541
Images / eyes1082 images
SplitNot reported
Reporting frameworks None stated
Age range>18
Sex (% female)Not reported
Race / ethnicity reportedN
Subgroup performanceN
§04Deployment
Deployment maturity
  1. Concept
  2. Prototype
  3. Internal val.
  4. External val.
  5. Regulator
  6. Clinical
Regulatory status
None No FDA / Health Canada / CE clearance.
Artifacts & license
○Code repository none
○Pretrained weights none
○Live demo none
License Not specified
§05Quality Review · APPRAISE-AI
Total score
50 /100 Moderate
ClinicalDataMethodRobust.Report.Repro.
Sub-domain scores
Clinical relevance 4/4 100%
Data quality 11/24 46%
Methodological conduct 12/20 60%
Robustness of results 7/20 35%
Reporting quality 11/12 92%
Reproducibility 5/20 25%
▸ Why this score?

The APPRAISE-AI tool grades AI/ML medical studies across 24 items in 6 subdomains, scored independently by two reviewers. The total reflects the methodological transparency of the publication — not the underlying clinical value of the model.

How to read this: high scores in clinical relevance and reporting with low scores in methodological conduct, robustness, or reproducibility is the dominant pattern in ophthalmic AI today.

Quality bands: Low <40 · Moderate 40–59 · High 60–79 · Very High 80+. Most published ophthalmic AI sits in the Low–Moderate band; studies that break into High typically combine multi-center datasets, prospective evaluation, and open-source release.

§06Full Metadata
Citation & identification
Bibliographic record
▸
Model IDoculoplastics_balel_2023
Full author listBalel Y.
JournalEur J Ther
DOI10.58600/eurjther1648
ConditionsBlepharoptosis
Architecture
Model internals
▸
Architecture familyCNN
Architecture detailCustom CNN models
Funding & conflicts
Disclosures
▸
FundingNot reported
Conflicts of interestNo conflicts
Card lifecycle
Status, version, reviewers
▸
Card statusPublished
Card versionv1.0.0
First added2026-04-28
Last reviewed2026-04-28
Reviewersattribution opt-in pending
Update history
Card changelog
▸
  • 2026-04-28 initial publication (v1.0.0)
← Back to Oculoplastics