SPIE Medical Imaging 2010 Graph-based Pigment Network Detection in Skin Images Maryam Sadeghi1,3, Majid Razmara1, Martin Ester1, Tim K. Lee1,2,3 and M. Stella Atkins1 1: School of Computing Science, Simon Fraser University 2: Department of Dermatology and Skin Science, University of British Columbia 3:Cancer Control Research, BC Cancer Agency 1 Skin cancer and melanoma Skin cancer : most common of all cancers Melanoma : leading cause of mortality Early detection: significantly reduces mortality Basal cell carcinoma 2 Combined nevus [ Images courtesy of “Dermoscopy of pigmented skin lesions” ] Melanoma Dermoscopy 3 Pigment Network Detection Present (Typical or Atypical Pigment Network ) Typical: “light to dark-brown network with small, uniformly spaced network holes and thin network lines distributed more or less regularly throughout the lesion and usually thinning out at the periphery” Atypical: “black, brown or gray network with irregular holes and thick lines“ Absent: There is no typical or atypical pigment network 4 Present 5 Absent 6 Problem Statement and Motivation Problem: Pigment network detection in dermoscopy images Motivation: Skin texture analysis for computer-aided diagnosis Pigment Network Visualisation for Training purposes 7 Algorithm overview Given a dermoscopy image Original 8 Algorithm overview Pre-processing. Using LoG sharp changes of colors are detected Original 9 Laplacian of Gaussian Algorithm overview Converting the result of the pre-processing to a graph. Original 10 Laplacian of Gaussian Image to Graph Algorithm overview Converting the result of the pre-processing to a graph. Original 11 Laplacian of Gaussian Image to Graph Cyclic Subgraphs Algorithm overview Converting the result of the pre-processing to a graph. Original 12 Laplacian of Gaussian Pigment Network Image to Graph Cyclic Subgraphs Algorithm overview Converting the result of the pre-processing to a graph. Original Laplacian of Gaussian Image to Graph Present 13 Classification Pigment Network Cyclic Subgraphs Given Image 14 Filtered by Laplacian of Gaussian 15 Binary Image to Graph Conversion A binary image has some connected components Each of them is converted to a graph (G) Each pixel a node of G A unique label according to its coordinate Graph Gi |V|= size of the connected component i in pixels 7x7 |E|=Number of edges connecting the white pixels |V|=17 |E|=17 3 8 Iterative Loop Counting Algorithm. Graph Gi 48 16 Connected Component i Removing Undesired Cycles Labels of nodes coordinates in the image Mean intensity of meshes in the original image Globules and dots: Inside color darker than outside color Inside Color Outside Color Extended area by 2 pixels Tuning the Thresholds 17 18 19 Pigment Network Graph A new graph representing the Pigment Network Centers of the detected cycles ( green holes in the image) are determined as nodes For each center the distance from all nodes is computed According to the size of the lesion and the average size of the net holes, Maximum Distance Threshold (MDT) is set Two nodes are connected together if they are closer than 20 MDT 21 Image Classification Density Ratio of the detected pigment network Density E V * log( LesionSize ) Lesion Size: Size of the area of the image that is inspected for finding the pigment network Density Threshold Density Ratio ≥ Threshold => Present 22 Density Ratio < Threshold => Absent Experimental Results Original Image LOG Edge Detector Cyclic Subgraphs Present 23 Original Image LOG Edge Detector Cyclic Subgraphs Absent Evaluation Data Set and Result: A set of 100 dermoscopic images used for tuning the parameters and thresholds of the method 500 images of size 768x512 are used to test the performance of the method Taken from Argenziano et al.’s Interactive Atlas of Dermoscopy Each image is labeled as ‘Absent’ or ‘Present {typical, atypical} Accuracy: 92.6% 24 Future Work Features of pigment networks Color, regularity, thickness, spatial arrangement Extending the classification to 3 classes of Absent, Typical, and Atypical Color of the of surrounding network in blue channel Thickness and irregularity of the network Modifying the method to find other dermoscopic structures and patterns 25 Questions? Thank you 26 Conclusion A novel graph-based method for classifying and visualizing pigment networks. Evaluating its ability to classify and visualize real dermoscopic images The accuracy of the method is 92.6% (classifying images to Absent and Present) 27 Previous Work Comparing our results to previous methods: Anantha et al. “Detection of pigment network in dermatoscopy images using texture analysis” , 2004, Accuracy: 80% Betta et al. “Dermoscopic image-analysis system: estimation of atypical pigment network and atypical vascular pattern”, 2006, Recall:50% , Precision: 100%, F-measure: 66.66% Our method: Accuracy: 92.6% 28 Pre-processing: 2D edge detection Gaussian 29 derivative of Gaussian is the Laplacian operator: Laplacian of Gaussian Graph-based Pigment Network Detection 30 Absent 31 Present 32 Original Laplacian of Gaussian Image to Graph Pigment Network Cyclic Subgraphs Present Classification 33 Filtered by Laplacian of Gaussian 34 35 36 Experimental Results(2) 37 Experimental Results(2) 38 Experimental Results(2) 39 Present 40 Absent 41 Pigment Network Graph 42 Pigment Network Graph 43 44 45