Costal Cartilage Calcification in a Caucasian Population: Machine Learning Recommendations for Chest CT-Guided Rhinoplasty Planning.
Authors
Affiliations (7)
Affiliations (7)
- Department of Otolaryngology and Head & Neck Surgery, Erasmus Medical Center, Rotterdam, Netherlands.
- Department of Otolaryngology and Head & Neck Surgery, Erasmus MC University Medical Center Rotterdam, Rotterdam, Netherlands.
- Department of Radiology, Erasmus MC University Medical Center Rotterdam, Rotterdam, Netherlands.
- Department of Otolaryngology and Head & Neck Surgery, Erasmus Universiteit Rotterdam, Rotterdam, Netherlands.
- Otorhinolaryngology and Head & Neck Surgery, Erasmus MC University Medical Center Rotterdam, Rotterdam, Netherlands.
- Department of Otolaryngology and Head & Neck Surgery, Erasmus MC, Rotterdam, Netherlands.
- Department of Otolaryngology and Head & Neck Surgery, Erasmus Medical Center Rotterdam, Rotterdam, Netherlands.
Abstract
Autologous costal cartilage calcification (CCC) can impact the course and long-term results of rhinoplasty. Preoperative information about the presence, severity and pattern of CCC helps to assess donor site suitability and rhinoplasty planning. To use machine learning to identify a sex-specific age threshold, beyond which a preoperative chest CT is likely to reveal CCC relevant for rhinoplasty planning. Cross-sectional retrospective study of 662 Caucasian adults. Prevalence, severity and patterns of CCC in ribs 5-8 were assessed by three independent reviewers. A machine learning algorithm was used to predict the age threshold beyond which a chest CT scan is beneficial to rhinoplasty planning. The prevalence of CCC in Caucasian adults was 89.6%. Nearly all individuals over age 50 exhibited some form of CCC. In young females CCC was more severe and prevalent in the central core of ribs 5-8 compared to age-matched males. A chest CT is recommended in females over 23 years and males over 40 years. No data-driven recommendations regarding an upper age limit for costal cartilage use could be determined from the data.