Morphological Analysis of Pressure Injury Images

This paper details the morphological analysis of pressure injuries when using a novel methodology developed to process medical images, intended to de-compose the contained information in different contrast levels. This technique was developed using parametric surface equations based on a toroidal geometry, which decompose the image vertically and horizontally by using synthetic frequencies calculated directly from the image. In conjunction with Otsu’s threshold algorithm, this technique allows for development of a methodology that is similar to level set, where the objective is to calculate the curve that describes the contours from objects within an image without the need to solve a differential equation. After the decomposition and image binarization, morphological operations are used to obtain the entire domain of the desired object within an image. A data set with 37 images of pressure injuries, provided by the Centre IGURCO, were synthetized using the proposed methodology. In order to evaluate the performance, digital image correlation was used as a measure, where the segments obtained are compared with manual segmentation made by professionals. The results present percentage errors below the 9% for the used parameters with an average image correlation of 0.92 ± 0.036. The proposed methodology is therefore considered a good estimation using a consistent non-invasive technique for the assessment of the healing process of the wound.

[1]  Stanley Osher,et al.  REVIEW ARTICLE: Level Set Methods and Their Applications in Image Science , 2003 .

[2]  Pina Marziliano,et al.  Automatic knee cartilage segmentation from multi-contrast MR images using support vector machine classification with spatial dependencies. , 2013, Magnetic resonance imaging.

[3]  Daniel Sierra-Sosa,et al.  Pressure ulcer image segmentation technique through synthetic frequencies generation and contrast variation using toroidal geometry , 2017, Biomedical engineering online.

[4]  Hui Zhang,et al.  Probe into Image Segmentation Based on Sobel Operator and Maximum Entropy Algorithm , 2012, 2012 International Conference on Computer Science and Service System.

[5]  Milta O. Little,et al.  Pressure Ulcer Scale for Healing ( PUSH ) PUSH Tool 3 . 0 , 2012 .

[6]  V. Rajinikanth,et al.  Otsu based optimal multilevel image thresholding using firefly algorithm , 2014 .

[7]  Ricardo Navarro Navarro,et al.  Ulceras por presión , 2008 .

[8]  Jun Tang,et al.  A color image segmentation algorithm based on region growing , 2010, 2010 2nd International Conference on Computer Engineering and Technology.

[9]  Tony F. Chan,et al.  Active contours without edges , 2001, IEEE Trans. Image Process..

[10]  S. Bramble.  Image analysis for the biological sciences , 1996 .

[11]  Demetri Terzopoulos,et al.  Snakes: Active contour models , 2004, International Journal of Computer Vision.

[12]  Wang Zhan-kai.  Application of the OTSU Method in Image Processing , 2006 .

[13]  B. Turner,et al.  A Non-Contact Imaging-Based Approach to Detecting Stage I Pressure Ulcers , 2006, 2006 International Conference of the IEEE Engineering in Medicine and Biology Society.

[14]  J. Sethian.  1 Advancing Interfaces : Level Set and Fast Marching Methods , 1999 .

[15]  S. Osher,et al.  Level set methods: an overview and some recent results , 2001 .

[16]  Joan Dalmau,et al.  Úlceras por presión , 2004 .

[17]  R. Guadagnin,et al.  An image mining based approach to detect pressure ulcer stage , 2014, Pattern Recognition and Image Analysis.

[18]  Benjamin Pfaff,et al.  Handbook Of Image And Video Processing , 2016 .

[19]  O. Basset,et al.  3D Ultrasound Elastography for Early Detection of Lesions. Evaluation on a Pressure Ulcer Mimicking Phantom , 2007, 2007 29th Annual International Conference of the IEEE Engineering in Medicine and Biology Society.

[20]  B Nila Mankar,et al.  COMPARISION OF DIFFERENT IMAGING TECHNIQUES USED FOR CHRONIC WOUNDS , 2013 .

[21]  Yilong Yin,et al.  Non-locally regularized segmentation of multiple sclerosis lesion from multi-channel MRI data. , 2014, Magnetic resonance imaging.

[22]  Nader Jafarnia Dabanloo,et al.  Automatic classification of speech dysfluencies in continuous speech based on similarity measures and morphological image processing tools , 2016, Biomed. Signal Process. Control..

[23]  Jiu-lun Fan,et al.  Image segmentation based on weak fuzzy partition entropy , 2015, Neurocomputing.