IntroductionAssessing facial aesthetic surgery outcomes often relies on expert opinion, which inherently introduces bias. This study examines the use of artificial intelligence (AI) to objectively compare novel and standard brow lift techniques. MethodsFifty-nine patients underwent brow lift surgery between 2006 and 2020 and were divided into two groups: Group A (n=47) underwent a standard brow lift, while Group B (n=12) received our novel technique. Patients were excluded if they underwent a simultaneous rhytidectomy or if their images were unreadable by the software. AI analysis examined emotion and action unit functioning, and then was correlated to manual measurements of brow position. ResultsGroup A demonstrated an increase in happiness from 1.6% to 9.5% (p=0.004), a decrease in sadness from 14.9% to 8.0% (p=0.06), and a decrease in anger from 14.8% to 5.7% (p=0.007). Group B experienced an increase in happiness from 5.2% to 12.6% (p=0.18), a decrease in sadness from 18.9% to 10.3% (p=0.1), and a decrease in anger from 26.4% to 10.9% (p=0.07). Group A exhibited elevation in brow position by 2.3-3.7mm (p<0.05) and a decrease in brow lowerer action unit functioning from 17.8% to 3.8% (p=0.0005). Group B exhibited elevation in brow position by 2.78-5.05mm (p<0.05) and a decrease in brow lowerer action unit functioning from 35% to 8.4% (p=0.011). ConclusionThis study demonstrated similar efficacy of the examined brow lift techniques for causing positive emotional changing while elevating the brow. This objective comparison demonstrates the utility of artificial intelligence in assessing facial surgery.
Read full abstract