Detecting Asymmetry With Artificial Intelligence: Consistency Evaluation of the Looksmaxxing Application in Photo-Based Facial Analysis
试验速览
- 阶段
- 不适用
- 状态
- 进行中（未招募）
- 发起方
- 入组人数
- 90
- 试验地点
- 1
- 主要终点
- Accuracy of Looksmaxxing AI in Detecting Chin Deviation
研究概览
简要总结
This study aims to evaluate whether an artificial intelligence application called Looksmaxxing AI will be able to correctly identify chin deviation (chin asymmetry) from standard frontal facial photographs. A total of 540 photographs will be included in the study. The eye areas will be covered to protect identity. Each photo will be analyzed by the AI, and its answers will be compared with clinical reality. The accuracy of two versions of the software (Looksmaxxing 4o and 5) will be assessed. The results may help show whether simple photo-based analysis can support early detection of chin asymmetry, especially in areas with limited access to orthodontic examination.
详细描述
This observational study will investigate the accuracy of the Looksmaxxing AI application in detecting chin deviation based on standardized frontal facial photographs. A total of 540 images will be included. To protect identity, the eye areas of all participants will be covered. The AI will be asked to determine whether chin deviation is present relative to the midline of the face (excluding the nose).
The study population will consist of two equally sized groups:
270 individuals with clinically observed chin deviation (laterognathia)
270 individuals with clinically normal chin position
The responses provided by Looksmaxxing AI will be compared with the clinical reality and recorded for accuracy. Both versions of the software, Looksmaxxing 4o and Looksmaxxing 5, will be analyzed to evaluate their level of consistency with clinical findings. Results will be reported as accuracy percentages. In addition, prediction accuracy will be compared between individuals with and without chin deviation, and the statistical significance of the difference will be tested using the Chi-square method.
研究设计
- 研究类型
- Observational
- 观察模型
- Cohort
- 时间视角
- Retrospective
入排标准
- 年龄范围
- 10 Years 至 40 Years（Child, Adult）
- 性别
- All
- 接受健康志愿者
- 是
入选标准
- •Absence of any craniofacial anomaly
- •No major wound or scar in the facial or neck region
- •No history of previous orthodontic treatment
- •For male participants: absence of a long beard that could affect the appearance of the chin
- •Availability of standardized frontal facial photographs taken in natural head position
排除标准
- •Photographs in which the patient's face appears slightly angled or turned sideways
- •Blurred or low-quality photographs with reduced clarity
结局指标
主要结局
Accuracy of Looksmaxxing AI in Detecting Chin Deviation
时间窗: At study completion (October 2025)
The proportion of correct classifications (presence or absence of chin deviation) made by the Looksmaxxing AI application compared to the clinical gold standard, expressed as a percentage.
次要结局
未报告次要终点
研究者
Seda Sağoğlu
Research assistant
Konya Necmettin Erbakan Üniversitesi