Back to 2024 Abstracts
One Profile to Rule Them All? A Neural Network Investigation of the Homogenizing Effect of Primary Rhinoplasty
Kristina L. Khaw
*1, Stephen M. Lu
21Cooper Medical School of Rowan University, Plainsboro, NJ; 2University of Pittsburgh Medical Center, Mechanicsburg, PA
Background: A goal of aesthetic rhinoplasty is to improve nasal aesthetics while preserving individuality. To the contrary, we hypothesized that primary rhinoplasty has a homogenizing effect on a patient's appearance and in a distinct manner for a given rhinoplasty surgeon.
Methods: Standardized and unedited pre- and postoperative photos, including frontal, lateral, and oblique views, of 145 Caucasian female primary rhinoplasty patients from the public websites of three independent, experienced rhinoplasty surgeons were included (Surgeon A: n=50, Surgeon B: n=50, Surgeon C: n=45). Using Amazon Rekognition, similarity between photos were calculated by comparing all the preoperative photos to each other and all postoperative photos to each other. Similarity differences between pre- and postoperative views were calculated intra- and intercohort. Pre- and postoperative changes were analyzed with a paired t-test. A simple t-test assessed the magnitude of change between cohorts.
Results: Patients were significantly more similar post surgery (0.89%, p<0.001) with the greatest effect in the lateral (1.13%, p<0.001), oblique (0.80%, p<0.001), and then the frontal (0.63%, p<0.001) views. In the subgroups, Surgeon A similarity increased 1.13% (front: 0.78%, lateral: 1.66%, oblique: 1.05%), while Surgeon B had a 0.89% increase (front: 0.46%, lateral: 0.90%, oblique: 0.78%), and Surgeon C a 0.64% increase (front: 0.56%, lateral: 0.75%, oblique: -0.12%). The intercohort similarity increase (0.56%, p=0.003) was significantly less than the intracohort increase.
Conclusion: This study illustrates that primary rhinoplasty patients have increased homogeneity postoperatively. Patients within cohorts were more alike in appearance than between cohorts. We believe that this is the first study that quantifies the homogenizing effect of rhinoplasties using neural networks as well as shows the quantifiable "signature" effect of a specific surgeon on their patients' appearances as a population. This study may impact how surgeons set patients' expectations and how patients select surgeons.
Back to 2024 Abstracts
