Advertisement

Convolutional neural network-based skin image segmentation model to improve classification of skin diseases in conventional and non-standardized picture images

      Highlights

      • Auto intelligences (AI) trained by conventional photo images show less performance.
      • We trained DeepLabv3 + -based CNN segmentation model segmenting skin and lesion area.
      • The CNN segmentation model can generate image datasets suitable for AI diagnosis.
      • AI diagnosis in the CNN-segmented images was better than the original images.
      • Our CNN segmentation model will improve AI classification in skin disease images.

      Abstract

      Background

      For dermatological practices, non-standardized conventional photo images are taken and collected as a mixture of variable fields of the image view, including close-up images focusing on designated lesions and long-shot images including normal skin and background of the body surface. Computer-aided detection/diagnosis (CAD) models trained using non-standardized conventional photo images exhibit lower performance rates than CAD models that detect lesions in a localized small area, such as dermoscopic images.

      Objective

      We aimed to develop a convolutional neural network (CNN) model for skin image segmentation to generate a skin disease image dataset suitable for CAD of multiple skin disease classification.

      Methods

      We trained a DeepLabv3 + -based CNN segmentation model to detect skin and lesion areas and segmented out areas that satisfy the following conditions: more than 80% of the image will be the skin area, and more than 10% of the image will be the lesion area.

      Results

      The generated CNN-segmented image database was examined using CAD of skin disease classification and achieved approximately 90% sensitivity and specificity to differentiate atopic dermatitis from malignant diseases and complications, such as mycosis fungoides, impetigo, and herpesvirus infection. The accuracy of skin disease classification in the CNN-segmented image dataset was almost equal to that of the manually cropped image dataset and higher than that of the original image dataset.

      Conclusion

      Our CNN segmentation model, which automatically extracts lesions and segmented images of the skin area regardless of image fields, will reduce the burden of physician annotation and improve CAD performance.

      Abbreviations:

      CAD (computer-aided diagnosis), AD (atopic dermatitis), CNN (convolutional neural network), NSDD (National Skin Disease Database), AI (artificial intelligence), MF (mycosis fungoides), IM (impetigo), HI (herpesvirus infection), AUC (area under the curve)
      To read this article in full you will need to make a payment

      Purchase one-time access:

      Academic & Personal: 24 hour online accessCorporate R&D Professionals: 24 hour online access
      One-time access price info
      • For academic or personal research use, select 'Academic and Personal'
      • For corporate R&D use, select 'Corporate R&D Professionals'

      Subscribe:

      Subscribe to Journal of Dermatological Science
      Already a print subscriber? Claim online access
      Already an online subscriber? Sign in
      Institutional Access: Sign in to ScienceDirect

      References

      1. Ramlakhan, K. and Y. Shang. A Mobile Automated Skin Lesion Classification System. in 2011 IEEE 23rd International Conference on Tools with Artificial Intelligence. 2011.

        • González-Díaz I.
        DermaKNet: incorporating the knowledge of dermatologists to convolutional neural networks for skin lesion diagnosis.
        IEEE J. Biomed. Health Inform. 2019; 23: 547-559
        • Yu L.
        • et al.
        Automated melanoma recognition in dermoscopy images via very deep residual networks.
        IEEE Trans. Med. Imaging. 2017; 36: 994-1004
        • Gessert N.
        • et al.
        Skin lesion classification using CNNs with patch-based attention and diagnosis-guided loss weighting.
        IEEE Trans. Biomed. Eng. 2020; 67: 495-503
        • Zhang J.
        • et al.
        Attention residual learning for skin lesion classification.
        IEEE Trans. Med. Imaging. 2019; 38: 2092-2103
        • Wu J.
        • et al.
        Skin lesion classification using densely connected convolutional networks with attention residual learning.
        Sens. (Basel, Switz. ). 2020; 20: 7080
        • Liu Y.
        • et al.
        A deep learning system for differential diagnosis of skin diseases.
        Nat. Med. 2020; 26: 900-908
        • Thomsen K.
        • et al.
        Deep learning for diagnostic binary classification of multiple-lesion skin diseases.
        Front. Med. 2020; 7574329
        • Han S.S.
        • et al.
        Deep neural networks show an equivalent and often superior performance to dermatologists in onychomycosis diagnosis: Automatic construction of onychomycosis datasets by region-based convolutional deep neural network.
        PLoS One. 2018; 13e0191493
        • Sun X.
        • et al.
        A Benchmark for Automatic Visual Classification of Clinical Skin Disease Images. in Computer Vision – ECCV 2016.
        Springer International Publishing, Cham2016
        • Thomsen K.
        • et al.
        Systematic review of machine learning for diagnosis and prognosis in dermatology.
        J. Dermatol. Treat. 2020; 31: 496-510
        • Takeuchi S.
        • Esaki H.
        • Furue M.
        Epidemiology of atopic dermatitis in Japan.
        J. Dermatol. 2014; 41: 200-204
        • Yamasaki K.
        • Gallo R.L.
        Antimicrobial peptides in human skin disease.
        Eur. J. Dermatol. 2008; 18: 11-21
        • Hata T.R.
        • et al.
        Administration of oral vitamin D induces cathelicidin production in atopic individuals.
        J. Allergy Clin. Immunol. 2008; 122: 829-831
        • Fukui R.
        • et al.
        Chronological changes of skin eruptions toward cold abscess formation in hyper-immunoglobulin E syndrome.
        J. Dermatol. 2021; 48: e316-e317
        • Mehrany K.
        • et al.
        Cutaneous T-cell lymphoma and atopy: is there an association?.
        Br. J. Dermatol. 2003; 149: 1013-1017
        • Zhou H.
        • et al.
        Folliculotropic mycosis fungoides associated with atopic dermatitis.
        Austral J. Dermatol. 2018; 59: e143-e145
        • Fujimoto A.
        • et al.
        Deep neural network for early image diagnosis of stevens-johnson syndrome/toxic epidermal necrolysis.
        J. Allergy Clin. Immunol. Pr. 2022; 10: 277-283
        • Chen L.-C.
        • et al.
        Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation. in Computer Vision – ECCV 2018.
        Springer International Publishing, Cham2018
      2. Ramlakhan, K. and Y. Shang, A Mobile Automated Skin Lesion Classification System, in 2011 IEEE 23rd International Conference on Tools with Artificial Intelligence. 2011. p. 138–141.

        • Kittler H.
        • et al.
        Diagnostic accuracy of dermoscopy.
        Lancet Oncol. 2002; 3: 159-165
      3. UMSkinCheck, 〈https://www.uofmhealth.org/patient%20and%20visitor%20guide/my-skin-check-app〉.

      4. MoleMapper, 〈https://molemapper.org/〉.

      5. Miiskin, 〈https://miiskin.com/app/?gclid=EAIaIQobChMIxayqrIKF5AIVibbtCh2uRAgBEAAYASAAEgJs2vD_BwE〉.

      6. SkinVision, 〈https://www.skinvision.com/〉.

        • Muñoz-López C.
        • et al.
        Performance of a deep neural network in teledermatology: a single-centre prospective diagnostic study.
        J. Eur. Acad. Dermatol. Venereol. 2021; 35: 546-553