Reproducibility of Three-dimensional Area Measurement of Periocular Tumors
Author: Xincen Hou
Base Hospital / Institution: Department of Ophthalmology, University of Cologne, Faculty of Medicine and University Hospital Cologne, Cologne, Germany.
Rapid fire oral presentation
Abstract ID: 24-302
Purpose
In this study, we utilized three-dimensional (3D) stereophotogrammetry to quantify the surface area of diverse tumors in the periocular region, assessing the reproducibility of these measurements.
Methods
We collected 3D facial images from 150 patients diagnosed with periocular tumors, amounting to 175 tumors. The tumors underwent classification according to their diameter, shape, color, location, surface texture, distance from the eyelid margin, clarity of boundaries, and histological characteristics. Following this, the surface area of the tumor models was quantified, and intra- and interrater reproducibility was assessed.
Results
Firstly, the method presented overall reliable measurements with intra- and interrater correlation coefficients(ICC) of 0.998 and 0.974, mean absolute differences(MAD) of 0.006 and 0.004 mm2, relative error measurements(REM) of 1.94% and 1.22%, technical error of measurement(TEM) of 0.023 and 0.078 mm2, and relative technical error of measurement(rTEM) of 6.95% and 23.76%, respectively. We then screened four variables that had an impact on measurement reliability among the nine variables we included, namely tumor diameter, color, location, and whether the border was clear or not. Further analysis of each factor revealed that tumors located at the outer canthus, with a diameter greater than 5mm, clear boundaries, and dark-brown color exhibited higher reliability. Conversely, tumors located at the inner canthus, with a diameter less than 5mm, unclear boundaries, and yellow color demonstrated lower reliability in periocular area.
Conclusion
This study provides the first confirmation of the high reliability of using three-dimensional (3D) stereophotogrammetry to measure the actual surface area of periocular tumors. Furthermore, it identifies several variables that influence measurements, resulting in more precise outcomes and demonstrating the feasibility of its application in measuring eyelid tumors. Consequently, it presents a novel and reliable option for recording and follow-up of different eyelid tumors in clinical practice.
Additional Authors
| First name | Last name | Base Hospital / Institution |
|---|---|---|
| Wanlin | Fan | Department of Ophthalmology, University of Cologne, Faculty of Medicine and University Hospital Cologne, Cologne, Germany. |
| Alexander C. | Rokohl | Department of Ophthalmology, University of Cologne, Faculty of Medicine and University Hospital Cologne, Cologne, Germany. |
| Ludwig M. | Heindl | Department of Ophthalmology, University of Cologne, Faculty of Medicine and University Hospital Cologne, Cologne, Germany. |
