liver ct image segmentation with an optimum threshold using measure of fuzziness srge
TRANSCRIPT
Liver CT Image Segmentation with an Optimum Threshold using Measure of
Fuzziness
The 5th International Conference on Innovations in Bio-Inspired Computing and Applications. June 23-25, 2014
Abder-Rahman Ali
Overview
Motivation Proposed approach Measure of fuzziness Calculating the optimum threshold based on the
measure of fuzzinessOptimum threshold and ambiguous pixels
Results Conclusions
The 5th International Conference on Innovations in Bio-Inspired Computing and Applications. June 23-25, 2014
Motivation
The 5th International Conference on Innovations in Bio-Inspired Computing and Applications. June 23-25, 2014
We use an optimum threshold, calculated using measure of fuzziness, in order to reveal the
ambiguous pixels, which are eventually assigned to the appropriate clusters
Proposed Appraoch (FCM-t)
The 5th International Conference on Innovations in Bio-Inspired Computing and Applications. June 23-25, 2014
Measure of Fuzziness
The 5th International Conference on Innovations in Bio-Inspired Computing and Applications. June 23-25, 2014
• Linear index of fuzziness (used to calculate the optimum threshold)
Measure of Fuzziness (cont…)
The 5th International Conference on Innovations in Bio-Inspired Computing and Applications. June 23-25, 2014
• Linear index of fuzziness (used to calculate the optimum threshold)
Calculating the Optimum Threshold based on the Measure of Fuzziness
The 5th International Conference on Innovations in Bio-Inspired Computing and Applications. June 23-25, 2014
• Linear index of fuzziness (used to calculate the optimum threshold)
Optimum Threshold and Ambiguous Pixels
The 5th International Conference on Innovations in Bio-Inspired Computing and Applications. June 23-25, 2014
• The optimum threshold is used to reveal the ambiguous pixels
• Pixels with membership values greater than or equal to the threshold will be assigned to the appropriate clusters (identied as 1 and 2 )
• Pixels with membership values less than the threshold will be marked as ambiguous, and assigned to the appropriate clusters, calculated by rounding to the nearest integer the average of the cluster values in the 3 x3 neighbourhood of that uncertain pixel
Results
The 5th International Conference on Innovations in Bio-Inspired Computing and Applications. June 23-25, 2014
original image
groundtruth
FCM
FCM-t
Results (Jaccard Index)
The 5th International Conference on Innovations in Bio-Inspired Computing and Applications. June 23-25, 2014
Results (CPU Processing Time)
The 5th International Conference on Innovations in Bio-Inspired Computing and Applications. June 23-25, 2014
Conclusions
The 5th International Conference on Innovations in Bio-Inspired Computing and Applications. June 23-25, 2014
Compared to the traditional Fuzzy C-Means, the proposed approach showed significantly better results in terms of Jaccard Index, although that was at the cost of some processing power
From a visual perspective, the proposed approach in some cases was able to show the ground truth more clearly
Thanks and Acknowledgement
The 5th International Conference on Innovations in Bio-Inspired Computing and Applications. June 23-25, 2014
http://www.egyptscience.net
Authors: Abder-Rahman Ali, Micael Couceiro, Ahmed M. Anter, Abul Ella Hassenian, Mohamed F. Tolba, and Vaclav Snasel