K-Means Clustering Algorithm

A Descriptive Study of
K-Means Clustering for Image Segmentation

In Bioinformatics
By-
Divyesh Saglani
Introduction
➢ Genetic Algorithms are one of the most applied class of algorithms for solving
global/multi-modal optimization problems and have been extensively studied for solving
NP-hard optimization problems.
➢ Graph clustering is an efficient methodology to detect groups of entities in a network or
graph, possessing certain similar characteristics.
➢ The genome projects are generating large amounts of biological data and so the data can
be perplexing and confusing. Data clustering approaches can group similar data into
clusters, which will reveal important biological meanings.
➢ This clustering is widely used in image segmentation.
Introduction
➢ Image segmentation is used to classify the pixels of an image correctly in a decision
oriented application.
➢ One of the most used clustering algorithm is k-means clustering algorithm.
➢ It is simple and computationally faster than the hierarchical clustering. And it can also
work for large number of variable.
➢ It is required to initialize the k number of centroid. Different value of initial centroid
would result different cluster.
➢ Image segmentation becomes one of important tool in medical area where it is used to
extract or region of interest from the background. So medical images are segmented using
different technique and process outputs are used for the further analysis in medical.
Related Work
1. Alan Jose, S. Ravi and M. Sambath proposed Brain Tumor Segmentation using K-
means Clustering and Fuzzy C-means Algorithm and its area calculation
(Dhanachandra, 2015 & Küçükkülahli, 2016).
2. Genetic algorithms are used for detecting money-laundering by finding the clusters in
a graph, constructed using financial and customer data (AlsedÃ, 2012).
3. K-Means Clustering Algorithms used for clustering the pixels of an image for image
color quantization (Hamerly, 2015).
4. K-Means Clustering Algorithms used for post-processing to decide the memberships
in spectral clustering (Hamerly, 2015).
5. Genetic Algorithms have been used for pattern recognition ( Tang, 1996).
Related Work
6. Newman and Girvan highlighted the relevance of the community structure in
complex networks and proposed a method to detect it (AlsedÃ, 2012).
7. Madhu Yedla, Srinivasa Rao Pathakota, T. M. Srinivasa proposed Enhancing K-
means clustering algorithm with improved initial center ( Hamerly, 2015).
8. K. A. Abdul Nazeer, M. P. Sebastian proposed an enhanced algorithm to improve the
accuracy and efficiency of the k-means clustering algorithm ( Hamerly, 2015).
9. The implementation of Lloyd’s k-means clustering algorithm and Progressive Greedy
k-means clustering algorithm was done and the algorithms were compared by
Gregory A. Wilkin and Xiuzhen Huang (Wilkin, 2007).
A lot of other works have already been done but the researches are still being conducted to
find out if these algorithms can be improved in some ways ( Küçükkülahli, 2016).
Research Methodology
Clustering
Genetic Algorithms Algorithms (GAC) K-Means Clustering
and (GACN)
Brain Tumor and Image

MRI using image Segmentation
segmentation and using k-means
k-means clustering clustering
1. Genetic Algorithms: - It is a metaheuristic inspired by the process of natural selection
that belongs to the larger class of evolutionary algorithms and are commonly used to
generate high-quality solutions to optimization and search problems by relying on bio-
inspired operators such as mutation, crossover and selection.
2. Clustering Algorithms: - These are types of genetic algorithms and here have been
used to find the best fit cluster. Iteratively divide the graph in two best fit union of
clusters at every step, discard the group of clusters which does not contain the desired
node and repeat this process in the remaining subgraph. Are of 2 types: -
a. GAC - Genetic Clustering Algorithm
b. GACN - Genetic Algorithm : Cluster of node
3. K- means Clustering: - Clustering is a method to divide a set of data into a specific
number of groups. It’s one of the popular method is k-means clustering. It is popular
due to its clarity, simplicity, and intuitive optimization function.
4. Image Segmentation using k-means clustering: - From the above concepts of genetic
clustering algorithms and k-means clustering, and algorithm was formed and was
implemented on images to see how different segments of images get affected due to k-
means clustering algorithm.
5. Brain Tumor and MRI using image segmentation and k-means clustering: - The above
concepts were then used to implement in real life problems,i .e. ,finding the tumor area
in brain using image segmentation and k -means clustering.
Results
Original Image Fig1 with k = 5 Fig2 with k = 10

Figures for implementation of k-means clustering for image segmentation
# Number of Number of Run Time (GAC) Run time (GACN)
Nodes clusters seconds seconds
1 11 2 0.15±0.056 0.54±.029
2 21 3 0.74±.0223 2.07±0.409
3 61 4 12.53±3.161 4.37±0.104
4 101 5 34.97±9.478 6.96±0.0541
5 301 15 430.90±110.5 25.53±0.508
6 401 10 643.23±170.032 37.61±0.625
Error-plot of the logarithmic run-time; GAC (red), GACN (blue) Results of GAC and GACN (runtime ± standard deviation)
Results
Figure 2: Output image of tumor detection in brain using k-means clustering algorithm
Conclusion
● Both algorithms, GAC and GACN were compared and it was found that GACN was
better if number of nodes were high and this can be used money-laundering
detection.
● Image Segmentation using k-means clustering was then implemented and the
resultant images were compared with different values of k, and the changes in
segmentation of different clusters were noticed with change in value of k.
● Then a real life example for image segmentation using k-means clustering was
implemented, which was to check the development of tumor in brain and it was
found that the stage of tumor can be calculated from this segmented image and also
the shape can position can for found with least possible error.
References
1. AlsedÃ , L., Awasthi, A., & LÃ¤ssig, J. (2012). Genetic Clustering Algorithms for Detecting Money-
Laundering.
2. Hamerly, G., & Drake, J. (2015). Accelerating Lloyd’s algorithm for k-means clustering. In Partitional
clustering algorithms (pp. 41-78). Springer, Cham.
3. Dhanachandra, N., Manglem, K., & Chanu, Y. J. (2015). Image segmentation using K-means
clustering algorithm and subtractive clustering algorithm. Procedia Computer Science, 54(2015),
764-771.
4. Wilkin, G. A., & Huang, X. (2007, August). K-means clustering algorithms: Implementation and
comparison. In Computer and Computational Sciences, 2007. IMSCCS 2007. Second International
Multi-Symposiums on (pp. 133-136). IEEE.
5. Selvakumar, J., Lakshmi, A., & Arivoli, T. (2012, March). Brain tumor segmentation and its area
calculation in brain MR images using K-mean clustering and Fuzzy C-mean algorithm. In Advances
in Engineering, Science and Management (ICAESM), 2012 International Conference on (pp. 186-
190). IEEE.
References
6. Küçükkülahli, E., Erdoğmuş, P., & Polat, K. (2016). Brain MRI segmentation based on different
clustering algorithms. Brain, 155(3).
7. Dhanachandra, N., & Chanu, Y. J. (2015). Image Segmentation Method Using K-Means Clustering
Algorithm for Color Image. Advanced Research in Electrical and Electronic Engineering, 2.
8. Man, K. F., Tang, K. S., & Kwong, S. (1996). Genetic algorithms: concepts and applications [in
engineering design]. IEEE transactions on Industrial Electronics, 43(5), 519-534.
9. Bioinformatic Algorithms - Cambridge Computer Laboratory
10. Wu, M. N., Lin, C. C., & Chang, C. C. (2007, November). Brain tumor detection using color-based
k-means clustering segmentation. In Intelligent Information Hiding and Multimedia Signal
Processing, 2007. IIHMSP 2007. Third International Conference on (Vol. 2, pp. 245-250). IEEE.

K-Means Clustering Algorithm

Uploaded by

Copyright:

Available Formats

K-Means Clustering Algorithm

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

K-Means Clustering Algorithm

Uploaded by

Copyright:

Available Formats

A Descriptive Study of

K-Means Clustering for Image Segmentation

Brain Tumor and Image

Original Image Fig1 with k = 5 Fig2 with k = 10

4 101 5 34.97±9.478 6.96±0.0541

5 301 15 430.90±110.5 25.53±0.508

6 401 10 643.23±170.032 37.61±0.625

You might also like