Quantification of Efflorescences in Pustular Psoriasis Using Deep Learning
2
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Objectives: Pustular psoriasis (PP) is one of the most severe and chronic skin conditions. Its treatment is difficult, and measurementsof its severity are highly dependent on clinicians’ experience. Pustules and brown spots are the main efflorescencesof the disease and directly correlate with its activity. We propose an automated deep learning model (DLM) to quantifylesions in terms of count and surface percentage from patient photographs. Methods: In this retrospective study, two dermatologistsand a student labeled 151 photographs of PP patients for pustules and brown spots. The DLM was trained andvalidated with 121 photographs, keeping 30 photographs as a test set to assess the DLM performance on unseen data. We alsoevaluated our DLM on 213 unstandardized, out-of-distribution photographs of various pustular disorders (referred to as thepustular set), which were ranked from 0 (no disease) to 4 (very severe) by one dermatologist for disease severity. The agreementbetween the DLM predictions and experts’ labels was evaluated with the intraclass correlation coefficient (ICC) for thetest set and Spearman correlation (SC) coefficient for the pustular set. Results: On the test set, the DLM achieved an ICC of0.97 (95% confidence interval [CI], 0.97–0.98) for count and 0.93 (95% CI, 0.92–0.94) for surface percentage. On the pustularset, the DLM reached a SC coefficient of 0.66 (95% CI, 0.60–0.74) for count and 0.80 (95% CI, 0.75–0.83) for surface percentage. Conclusions: The proposed method quantifies efflorescences from PP photographs reliably and automatically, enabling aprecise and objective evaluation of disease activity.
Abstract
Objectives: Pustular psoriasis (PP) is one of the most severe and chronic skin conditions. Its treatment is difficult, and measurementsof its severity are highly dependent on clinicians’ experience. Pustules and brown spots are the main efflorescencesof the disease and directly correlate with its activity. We propose an automated deep learning model (DLM) to quantifylesions in terms of count and surface percentage from patient photographs. Methods: In this retrospective study, two dermatologistsand a student labeled 151 photographs of PP patients for pustules and brown spots. The DLM was trained andvalidated with 121 photographs, keeping 30 photographs as a test set to assess the DLM performance on unseen data. We alsoevaluated our DLM on 213 unstandardized, out-of-distribution photographs of various pustular disorders (referred to as thepustular set), which were ranked from 0 (no disease) to 4 (very severe) by one dermatologist for disease severity. The agreementbetween the DLM predictions and experts’ labels was evaluated with the intraclass correlation coefficient (ICC) for thetest set and Spearman correlation (SC) coefficient for the pustular set. Results: On the test set, the DLM achieved an ICC of0.97 (95% confidence interval [CI], 0.97–0.98) for count and 0.93 (95% CI, 0.92–0.94) for surface percentage. On the pustularset, the DLM reached a SC coefficient of 0.66 (95% CI, 0.60–0.74) for count and 0.80 (95% CI, 0.75–0.83) for surface percentage. Conclusions: The proposed method quantifies efflorescences from PP photographs reliably and automatically, enabling aprecise and objective evaluation of disease activity.
참고문헌 (30)
- Gooderham MJ, "An update on generalized pustular psoriasis", Expert Rev Clin Immunol, 15(9), pp.907~919 (2019)[DOI]
- Puzenat E, "What are the best outcome measures for assessing plaque psoriasis severity? A systematic review of the literature", J Eur Acad Dermatol Venereol, 24(Suppl 2), pp.10~16 (2010)[DOI]
- Bhushan M, "Oral liarozole in the treatment of palmoplantar pustular psoriasis : a randomized, double-blind, placebo-controlled study", Br J Dermatol, 145(4), pp.546~553 (2001)[DOI]
- Kolios AG, "Detection of small changes in psoriasis intensity with PrecisePASI", Dermatology, 230(4), pp.314~317 (2015)[DOI]
- 윤상웅, "Reduction of Inter-Rater and Intra-Rater Variability in Psoriasis Area and Severity Index Assessment by Photographic Training", Annals of Dermatology, 27(5), pp.557~562 (2015), 대한피부과학회
- Esteva A, "Dermatologist-level classification of skin cancer with deep neural networks", Nature, 542(7639), pp.115~118 (2017)[DOI]
- Meienberger N, "Observer-independent assessment of psoriasis-affected area using machine learning", J Eur Acad Dermatol Venereol, 34(6), pp.1362~1368 (2020)[DOI]
- Andermatt S, "Brainlesion: glioma, multiple sclerosis, stroke and traumatic brain injuries", pp.228~238 (2018), Springer
- Furger F, "Artificial neural networks in pattern recognition", pp.187~199 (2020), Springer
- Ronneberger O, "Medical image computing and computer-assisted intervention –MICCAI 2015", pp.234~241 (2015), Springer
- He K, "Deep residual learning for image recognition", Proceedings of the IEEE conference on computer vision and pattern recognition, pp.770~778 (2016)
- Deng J, "ImageNet:a large-scale hierarchical image database", Proceedings of 2009 IEEE Conference on Computer Vision and Pattern Recognition, pp.248~255 (2009)
- "A disciplined approach to neural network hyper-parameters: Part 1--learning rate, batch size, momentum, and weight decay" (2018)[URL]
- Yeung M, "Unified focal loss : generalising dice and cross entropy-based losses to handle class imbalanced medical image segmentation", Comput Med Imaging Graph, 95, pp.102026~ (2022)[DOI]
- Lin TY, "Focal loss for dense object detection", IEEE Trans Pattern Anal Mach Intell, 42(2), pp.318~327 (2020)[DOI]
- Zhu W, "AnatomyNet : deep learning for fast and fully automated whole-volume segmentation of head and neck anatomy", Med Phys, 46(2), pp.576~589 (2019)[DOI]
- El Jurdi R, "High-level prior-based loss functions for medical image segmentation : a survey", Comput Vis Image Underst, 210, pp.103248~ (2021)[DOI]
- Paszke A, "PyTorch : an imperative style, high-performance deep learning library", Adv Neural Inf Process Syst, 32, pp.8024~8035 (2019)
- Howard J, "Fastai : a layered API for deep learning", Information, 11(2), pp.108~ (2020)[DOI]
- van Stralen KJ, "Measuring agreement, more complicated than it seems", Nephron Clin Pract, 120(3), pp.c162~c167 (2012)[DOI]
- Schaap MJ, "Image-based automated psoriasis area severity index scoring by convolutional neural networks", J Eur Acad Dermatol Venereol, 36(1), pp.68~75 (2022)[DOI]
- Wu X, "Medical image computing and computer assisted intervention – MICCAI 2021", pp.185~194 (2021), Springer
- Pal A, "OR 2.0Context-aware operating theaters, computer assisted robotic endoscopy, clinical image-based procedures, and skin image analysis", pp.252~259 (2018), Springer
- Cazzolato MT, "The UTrack framework for segmenting and measuring dermatological ulcers through telemedicine", Comput Biol Med, 134, pp.104489~ (2021)[DOI]
- Zhao C, "Segmentation of dermoscopy images based on deformable 3D convolution and ResU-NeXt +", Med Biol Eng Comput, 59(9), pp.1815~1832 (2021)[DOI]
- Goyal M, "Skin lesion segmentation in dermoscopic images with ensemble deep learning methods", IEEE Access, 8, pp.4171~4181 (2019)[DOI]
- Schnurle S, "On using support vector machines for the detection and quantification of hand eczema", Proceedings of the 9th International Conference on Agents and Artificial Intelligence(ICAART), pp.75~84 (2017)
- Raj R, "Deep learning based multi-segmentation for automatic estimation of psoriasis area score", Proceedings of 2021 8th International Conference on Signal Processing and Integrated Networks (SPIN), pp.1137~1142 (2021)
- Liu Y, "A survey and performance evaluation of deep learning methods for small object detection", Expert Syst Appl, 172, pp.114602~ (2021)[DOI]
- Finnane A, "Proposed technical guidelines for the acquisition of clinical images of skin-related conditions", JAMA Dermatol, 153(5), pp.453~457 (2017)[DOI]