Lesion segmentation using 3D scan and deep learning for the evaluation of facial portwine stain birthmarks

Cheng Ke; Yuanbo Huang; Jun Yang; Yunjie Zhang; Huiqi Zhan; Chunfa Wu; Mingye Bi; Zheng Huang

doi:10.1016/j.pdpdt.2024.104030

Lesion segmentation using 3D scan and deep learning for the evaluation of facial portwine stain birthmarks

Photodiagnosis Photodyn Ther. 2024 Feb 28:46:104030. doi: 10.1016/j.pdpdt.2024.104030. Online ahead of print.

Authors

Cheng Ke¹, Yuanbo Huang², Jun Yang², Yunjie Zhang³, Huiqi Zhan¹, Chunfa Wu¹, Mingye Bi², Zheng Huang⁴

Affiliations

¹ MOE Key Laboratory of Medical Optoelectronics Science and Technology, School of Optoelectronics and Information Engineering, Fujian Normal University, Fuzhou 350100, PR China.
² Department of Dermatology, Wuxi People's Hospital, Wuxi 214000, PR China.
³ Department of Dermatology, Beijing Puxiang Hospital of Traditional Chinese Medicine, Beijing 100080, PR China.
⁴ MOE Key Laboratory of Medical Optoelectronics Science and Technology, School of Optoelectronics and Information Engineering, Fujian Normal University, Fuzhou 350100, PR China. Electronic address: zheng_huang@msn.com.

PMID: 38423233
DOI: 10.1016/j.pdpdt.2024.104030

Abstract

Background: Portwine stain (PWS) birthmarks are congenital vascular malformations. The quantification of PWS area is an important step in lesion classification and treatment evaluation.

Aims: The aim of this study was to evaluate the combination of 3D scan with deep learning for automated PWS area quantization.

Materials and methods: PWS color was measured using a portable spectrophotometer. PWS patches (29.26-45.82 cm²) of different color and shape were generated for 2D and 3D PWS model. 3D images were acquired by a handheld 3D scanner to create texture maps. For semantic segmentation, an improved DeepLabV3+ network was developed for PWS lesion extraction from texture mapping of 3D images. In order to achieve accurate extraction of lesion regions, the convolutional block attention module (CBAM) and DENSE were introduced and the network was trained under Ranger optimizer. The performance of different backbone networks for PWS lesion extraction were also compared.

Results: IDeepLabV3+ (Xception) showed the best results in PWS lesion extraction and area quantification. Its mean Intersection over Union (MIou) was 0.9797, Mean Pixel Accuracy (MPA) 0.9908, Accuracy 0.9989, Recall 0.9886 and F1-score 0.9897, respectively. In PWS area quantization, the mean value of the area error rate of this scheme was 2.61 ± 2.33.

Conclusions: The new 3D method developed in this study was able to achieve accurate quantification of PWS lesion area and has potentials for clinical applications.

Keywords: 3D scan; Area quantization; Deep learning; DeepLabV3+; Portwine stains; Spectrophotometer.