Evaluation of surgical wound segmentation using quantitative analysis

To provide assistance to medical practitioners in wound care, numerous studies have been developing clinical decision support tools for wound monitoring and tracing. Image segmentation is the foundation of image analysis. A superior segmentation could address further image analysis and accurately interpret an image even with background noise. However, the segmentation performance of the wound using objective quantitative analysis has not been well-documented. In this study, we evaluated the segmentation algorithm based on the Canny edge detector and genetic algorithm along with a quantitative objective analysis based on empirical objective and precision-recall measure. According to a majority vote among three surgeons as the gold standard for over 100 wound images, the results revealed that the proposed segmentation algorithm achieved 89% accuracy (95% confidence interval: 87.5--90.6) for wound region detection. In conclusion, the proposed segmentation detector is robust and reliable enough to perform further image analysis. The findings provide further support to evaluate image segmentation as a part of image processing.

[1]  Anuradha Godavarty,et al.  Wound size measurement of lower extremity ulcers using segmentation algorithms , 2016, SPIE BiOS.

[2]  Edward J. Delp,et al.  Optimum color spaces for skin detection , 2001, Proceedings 2001 International Conference on Image Processing (Cat. No.01CH37205).

[3]  L. DiPietro,et al.  Factors Affecting Wound Healing , 2010, Journal of dental research.

[4]  Bo Song,et al.  Automated wound identification system based on image segmentation and Artificial Neural Networks , 2012, 2012 IEEE International Conference on Bioinformatics and Biomedicine.

[5]  Qian Huang,et al.  Quantitative methods of evaluating image segmentation , 1995, Proceedings., International Conference on Image Processing.

[6]  Feipei Lai,et al.  Surgical wound segmentation based on adaptive threshold edge detection and genetic algorithm , 2017, International Conference on Graphic and Image Processing.

[7]  Y. J. Zhang,et al.  A survey on evaluation methods for image segmentation , 1996, Pattern Recognit..

[8]  Aurélio J. C. Campilho,et al.  Performance Evaluation of Image Segmentation , 2006, ICIAR.

[9]  Metin Nafi Gürcan,et al.  Segmentation and automated measurement of chronic wound images: probability map approach , 2014, Medical Imaging.

[10]  Rafael Marcos Luque Baena,et al.  Wound image evaluation with machine learning , 2015, Neurocomputing.

[11]  Hazem Wannous,et al.  Efficient SVM classifier based on color and texture region features for wound tissue images , 2008, SPIE Medical Imaging.

[12]  Hazem Wannous,et al.  Robust tissue classification for reproducible wound assessment in telemedicine environments , 2010, J. Electronic Imaging.

[13]  Jong Gwan Lim,et al.  Empirical Validation of Objective Functions in Feature Selection Based on Acceleration Motion Segmentation Data , 2015 .

[14]  Pietro Salvo,et al.  Correlation Between Wound Temperature Obtained With an Infrared Camera and Clinical Wound Bed Score in Venous Leg Ulcers. , 2015, Wounds : a compendium of clinical research and practice.

[15]  Lee J. Markowitz,et al.  Wounds measured from digital photographs using photodigital planimetry software: validation and rater reliability. , 2011, Wounds : a compendium of clinical research and practice.

[17]  Wookyung Sun,et al.  Segmentation of wounds using gradient vector flow , 2015, 2015 International Conference on Intelligent Informatics and Biomedical Sciences (ICIIBMS).

[18]  Xin Zhao,et al.  Improved image segmentation method based on optimized threshold using Genetic Algorithm , 2008, 2008 IEEE/ACS International Conference on Computer Systems and Applications.

[19]  J. Satheesh Kumar,et al.  Performance Evaluation of Image Segmentation using Objective Methods , 2016 .

[20]  T. K. Hunt,et al.  Human skin wounds: A major and snowballing threat to public health and the economy , 2009, Wound repair and regeneration : official publication of the Wound Healing Society [and] the European Tissue Repair Society.