Classification of pressure ulcer tissues with 3D convolutional neural network
A 3D convolution neural network (CNN) of deep learning architecture is supplied with essential visual features to accurately classify and segment granulation, necrotic eschar, and slough tissues in pressure ulcer color images. After finding a region of interest (ROI), the features are extracted from both the original and convolved with a pre-selected Gaussian kernel 3D HSI images, combined with first-order models of current and prior visual appearance. The models approximate empirical marginal probability distributions of voxel-wise signals with linear combinations of discrete Gaussians (LCDG). The framework was trained and tested on 193 color pressure ulcer images. The classification accuracy and robustness were evaluated using the Dice similarity coefficient (DSC), the percentage area distance (PAD), and the area under the ROC curve (AUC). The obtained preliminary DSC of 92%, PAD of 13%, and AUC of 95% are promising.
KeywordsPressure ulcer 3D convolution neural network (CNN) Tissue classification Linear combinations of discrete Gaussians (LCDG)
The authors thank Prof. Dr. Georgy Gimel’farb, Department of Computer Science, University of Auckland, Auckland, New Zealand, for his help in revising the paper. In addition, the authors want to thank Sofia Zahia, Connor Burns, and Daniel Sierra-Sosa for their support in summarizing the related work and preparing the masks for the GT images.
The grants that have contributed with partial funding of the study are IT − 905 − 16 to eVIDA research group from the Basque Government, JC2015 − 00305 Josè Castillejo Research Stay Grant from the Spanish Ministry, and ACM2017_09 from the University of Deusto.
- 1.Mukherjee R, Manohar DD, Das DK, Achar A, Mitra A, Chakraborty C (2014) Automated tissue classification framework for reproducible chronic wound assessment. Biomed Res Int 2014:1–9Google Scholar
- 2.Fauzi MFA, Khansa I, Catignani K, Gordillo G, Sen CK, Gurcan MN (2015) Computerized segmentation and measurement of chronic wound images. Comput Biol Med 60:74–85. [Online]. Available: http://www.sciencedirect.com/science/article/pii/S0010482515000645 CrossRefGoogle Scholar
- 3.Sen CK, Gordillo GM, Roy S, Kirsner R, Lambert L, Hunt TK, Gottrup F, Gurtner GC, Longaker MT (2009) Human skin wounds: a major and snowballing threat to public health and the economy. Wound Repair Regen 17(6):763–771. [Online]. Available: https://doi.org/10.1111/j.1524-475X.2009.00543.x CrossRefGoogle Scholar
- 4.Cuddigan J, Berlowitz DR, Ayello EA (2001) Pressure ulcers in america: prevalence, incidence, and implications for the future. Adv Skin Wound Care 14(4):208–215Google Scholar
- 5.Deprez JF, Cloutier G, Schmitt C, Gehin C, Dittmar A, Basset O, Brusseau E (2007) 3d ultrasound elastography for early detection of lesions. evaluation on a pressure ulcer mimicking phantom. In: 2007 29th annual international conference of the ieee engineering in medicine and biology society , pp 79–82Google Scholar
- 6.Agostini JV, Baker DI, Bogardus ST Making health care safer: A critical analysis of patient safety practices. Agency for Healthcare Research and Quality, U.S. Department of Health and Human Services, 2001, no. 27, ch. Prevention of Pressure Ulcers in Older Patients, pp 301–306Google Scholar
- 7.Leachtenauer J, Kell S, Turner B, Newcomer C, Lyder C, Alwan M (2006) A non-contact imaging-based approach to detecting stage i pressure ulcers. In: the proceedings of the 2006 international conference of the IEEE engineering in medicine and biology society, pp 6380–6383Google Scholar
- 8.Prado A, Andrades P, Benítez S Cirugía Plástica Esencial. Hospital Clinico Universidad De Chile, 2005, ch. Úlceras por presión, pp 111–126Google Scholar
- 11.Dorileo AG, Frade MAC, Rangayyan RM, Azevedo-Marques PM (2010) Segmentation and analysis of the tissue composition of dermatological ulcers. In: CCECE 2010, pp 1–4Google Scholar
- 13.Azevedo-Marques PM, Pereira SM, Frade MAC, Rangayyan RM (2013) Segmentation of dermatological ulcers using clustering of color components. In: 2013 26th IEEE Canadian conference on electrical and computer engineering (CCECE) , pp 1–4Google Scholar
- 16.Wang C, Yan X, Smith M, Kochhar K, Rubin M, Warren SM, Wrobel J, Lee H (2015 ) A unified framework for automatic wound segmentation and analysis with deep convolutional neural networks. In: 2015 37th annual international conference of the ieee engineering in medicine and biology society (EMBC), pp 2415–2418Google Scholar
- 17.Kawahara J, Hamarneh G (2016) Multi-resolution-Tract CNN with hybrid pretrained and skin-lesion trained layers. Springer International Publishing, Cham, pp 164–171Google Scholar
- 24.Thomas S (2017) Medetec wound database, http://www.medetec.co.uk/files/medetec-image-databases.html
- 26.Powers DMW (2011) Evaluation: from precision, recall and f-measure to roc, informedness, markedness & correlation. J Mach Learn Technol 2(1):37–63Google Scholar
- 27.Yang Y, Huang S (2007) Image segmentation by fuzzy c-means clustering algorithm with a novel penalty term. Computing and informatics 26:17–31Google Scholar