Page 51 - Fister jr., Iztok, and Andrej Brodnik (eds.). StuCoSReC. Proceedings of the 2018 5th Student Computer Science Research Conference. Koper: University of Primorska Press, 2018
P. 51
ain by computing their objective (fitness) function analysis: A review. IEEE reviews in biomedical engineering,
minimization ability, robustness in term of standard deviation 2, 147.
(STD.) and computational time. Segmentation quality parameters [6] Irshad, H., Veillard, A., Roux, L., & Racoceanu, D. (2014).
like QILV, PSNR and FSIM are also calculated to judge the Methods for nuclei detection, segmentation, and
segmentation accuracy of the clustering models over cluster classification in digital histopathology: a review—current
number 4, 6 and 8. CS algorithm based clustering model has been status and future potential. IEEE reviews in biomedical
run 40 times for each image. The segmented images engineering, 7, 97-114.
corresponding to Figs.1(a)-(b) are given as Figs.2-3 respectively. [7] Jothi, J. A. A., & Rajam, V. M. A. (2015). Segmentation of
On the other hand, Table 1 represents the average values of nuclei from breast histopathology images using PSO-based
fitness, standard deviation (STD.), computational time and quality Otsu’s multilevel thresholding. In Artificial Intelligence and
parameters over 40 images. Table 1 shows that CS based Evolutionary Algorithms in Engineering Systems (pp. 835-
clustering model provide minimized fitness value within less 843). Springer, New Delhi.
computational time. The standard deviation also proves the [8] Beevi, S., Nair, M. S., & Bindu, G. R. (2016). Automatic
significant robustness of CS algorithm. The resultant segmented segmentation of cell nuclei using Krill Herd optimization
images of CS based model associated with greater QILV, FSIM based multi-thresholding and localized active contour model.
and PSNR over cluster number 4, 6 and 8. Therefore, it can be Biocybernetics and Biomedical Engineering, 36(4), 584-596.
said that CS is the better than traditional K-means algorithm with [9] Shen, P., Qin, W., Yang, J., Hu, W., Chen, S., Li, L., ... &
the MAX_FE based termination condition. Gu, J. (2015). Segmenting multiple overlapping Nuclei in
H&E Stained Breast Cancer Histopathology Images based on
Conclusion an improved watershed.
[10] Veta, M., Van Diest, P. J., Kornegoor, R., Huisman, A.,
A Cuckoo Search (CS) algorithm based clustering model has been Viergever, M. A., & Pluim, J. P. (2013). Automatic nuclei
proposed for the proper segmentation of breast histopathology segmentation in H&E stained breast cancer histopathology
images. The performance of the CS algorithm based clustering images. PloS one, 8(7), e70221.
model has been compared to K-means algorithms in terms of [11] Wienert, S., Heim, D., Saeger, K., Stenzinger, A., Beil, M.,
fitness, computational time and quality parameters. Values of Hufnagl, P., ... & Klauschen, F. (2012). Detection and
quality parameters indicate that CS based clustering model segmentation of cell nuclei in virtual microscopy images: a
provide segmented images with higher QILV, FSIM and PSNR minimum-model approach. Scientific reports, 2, 503.
compares to K-means based clustering model. Analysis of the [12] Qi, X., Xing, F., Foran, D. J., & Yang, L. (2012). Robust
results also shows that K-means also associates with huge segmentation of overlapping cells in histopathology
computational time when number of clusters increases and this specimens using parallel seed detection and repulsive level
time consume problem has been successfully surmounted by using set. IEEE Transactions on Biomedical Engineering, 59(3),
CS algorithm. 754-765.
Several future directions exist of this study such as use of fuzzy [13] Xu, J., Janowczyk, A., Chandran, S., & Madabhushi, A.
logic based clustering, rough set based clustering, formulation of (2011). A high-throughput active contour scheme for
the multi-objective clustering models, and use these clustering segmentation of histopathological imagery. Medical image
models for different kinds of images. analysis, 15(6), 851-862.
[14] Paramanandam, M., Thamburaj, R., Manipadam, M. T., &
REFERENCES Nagar, A. K. (2014, May). Boundary extraction for
imperfectly segmented nuclei in breast histopathology
[1] Ferlay, J., Soerjomataram, I., Ervik, M., Dikshit, R., Eser, S., images–a convex edge grouping approach. In International
Mathers, C., ... & Bray, F. (2015). GLOBOCAN 2012 v1. 0, Workshop on Combinatorial Image Analysis (pp. 250-261).
Cancer Incidence and Mortality Worldwide: IARC Springer, Cham.
CancerBase No. 11. Lyon, France: International Agency for [15] Su, H., Liu, F., Xie, Y., Xing, F., Meyyappan, S., & Yang, L.
Research on Cancer; 2013. (2015, April). Region segmentation in histopathological
breast cancer images using deep convolutional neural
[2] Elston, C. W., & Ellis, I. O. (1991). Pathological prognostic network. In Biomedical Imaging (ISBI), 2015 IEEE 12th
factors in breast cancer. I. The value of histological grade in International Symposium on (pp. 55-58). IEEE.
breast cancer: experience from a large study with long‐term [16] Naylor, P., Laé, M., Reyal, F., & Walter, T. (2017, April).
follow‐up. Histopathology, 19(5), 403-410. Nuclei segmentation in histopathology images using deep
neural networks. In Biomedical Imaging (ISBI 2017), 2017
[3] Robbins, P., Pinder, S., De Klerk, N., Dawkins, H., Harvey, IEEE 14th International Symposium on (pp. 933-936). IEEE.
J., Sterrett, G., ... & Elston, C. (1995). Histological grading [17] Shafique, S., & Tehsin, S. (2018). Computer-Aided
of breast carcinomas: a study of interobserver agreement. Diagnosis of Acute Lymphoblastic
Human pathology, 26(8), 873-879. Leukaemia. Computational and mathematical methods in
medicine, 2018.
[4] Meijer, G. A., Beliën, J. A., Van Diest, P. J., & Baak, J. P. [18] Tosta, T. A. A., Neves, L. A., & do Nascimento, M. Z.
(1997). Origins of... image analysis in clinical pathology. (2017). Segmentation methods of H&E-stained histological
Journal of clinical pathology, 50(5), 365. images of lymphoma: a review. Informatics in Medicine
Unlocked, 9, 35-43.
[5] Gurcan, M. N., Boucheron, L., Can, A., Madabhushi, A.,
Rajpoot, N., & Yener, B. (2009). Histopathological image
StuCoSReC Proceedings of the 2018 5th Student Computer Science Research Conference 53
Ljubljana, Slovenia, 9 October
minimization ability, robustness in term of standard deviation 2, 147.
(STD.) and computational time. Segmentation quality parameters [6] Irshad, H., Veillard, A., Roux, L., & Racoceanu, D. (2014).
like QILV, PSNR and FSIM are also calculated to judge the Methods for nuclei detection, segmentation, and
segmentation accuracy of the clustering models over cluster classification in digital histopathology: a review—current
number 4, 6 and 8. CS algorithm based clustering model has been status and future potential. IEEE reviews in biomedical
run 40 times for each image. The segmented images engineering, 7, 97-114.
corresponding to Figs.1(a)-(b) are given as Figs.2-3 respectively. [7] Jothi, J. A. A., & Rajam, V. M. A. (2015). Segmentation of
On the other hand, Table 1 represents the average values of nuclei from breast histopathology images using PSO-based
fitness, standard deviation (STD.), computational time and quality Otsu’s multilevel thresholding. In Artificial Intelligence and
parameters over 40 images. Table 1 shows that CS based Evolutionary Algorithms in Engineering Systems (pp. 835-
clustering model provide minimized fitness value within less 843). Springer, New Delhi.
computational time. The standard deviation also proves the [8] Beevi, S., Nair, M. S., & Bindu, G. R. (2016). Automatic
significant robustness of CS algorithm. The resultant segmented segmentation of cell nuclei using Krill Herd optimization
images of CS based model associated with greater QILV, FSIM based multi-thresholding and localized active contour model.
and PSNR over cluster number 4, 6 and 8. Therefore, it can be Biocybernetics and Biomedical Engineering, 36(4), 584-596.
said that CS is the better than traditional K-means algorithm with [9] Shen, P., Qin, W., Yang, J., Hu, W., Chen, S., Li, L., ... &
the MAX_FE based termination condition. Gu, J. (2015). Segmenting multiple overlapping Nuclei in
H&E Stained Breast Cancer Histopathology Images based on
Conclusion an improved watershed.
[10] Veta, M., Van Diest, P. J., Kornegoor, R., Huisman, A.,
A Cuckoo Search (CS) algorithm based clustering model has been Viergever, M. A., & Pluim, J. P. (2013). Automatic nuclei
proposed for the proper segmentation of breast histopathology segmentation in H&E stained breast cancer histopathology
images. The performance of the CS algorithm based clustering images. PloS one, 8(7), e70221.
model has been compared to K-means algorithms in terms of [11] Wienert, S., Heim, D., Saeger, K., Stenzinger, A., Beil, M.,
fitness, computational time and quality parameters. Values of Hufnagl, P., ... & Klauschen, F. (2012). Detection and
quality parameters indicate that CS based clustering model segmentation of cell nuclei in virtual microscopy images: a
provide segmented images with higher QILV, FSIM and PSNR minimum-model approach. Scientific reports, 2, 503.
compares to K-means based clustering model. Analysis of the [12] Qi, X., Xing, F., Foran, D. J., & Yang, L. (2012). Robust
results also shows that K-means also associates with huge segmentation of overlapping cells in histopathology
computational time when number of clusters increases and this specimens using parallel seed detection and repulsive level
time consume problem has been successfully surmounted by using set. IEEE Transactions on Biomedical Engineering, 59(3),
CS algorithm. 754-765.
Several future directions exist of this study such as use of fuzzy [13] Xu, J., Janowczyk, A., Chandran, S., & Madabhushi, A.
logic based clustering, rough set based clustering, formulation of (2011). A high-throughput active contour scheme for
the multi-objective clustering models, and use these clustering segmentation of histopathological imagery. Medical image
models for different kinds of images. analysis, 15(6), 851-862.
[14] Paramanandam, M., Thamburaj, R., Manipadam, M. T., &
REFERENCES Nagar, A. K. (2014, May). Boundary extraction for
imperfectly segmented nuclei in breast histopathology
[1] Ferlay, J., Soerjomataram, I., Ervik, M., Dikshit, R., Eser, S., images–a convex edge grouping approach. In International
Mathers, C., ... & Bray, F. (2015). GLOBOCAN 2012 v1. 0, Workshop on Combinatorial Image Analysis (pp. 250-261).
Cancer Incidence and Mortality Worldwide: IARC Springer, Cham.
CancerBase No. 11. Lyon, France: International Agency for [15] Su, H., Liu, F., Xie, Y., Xing, F., Meyyappan, S., & Yang, L.
Research on Cancer; 2013. (2015, April). Region segmentation in histopathological
breast cancer images using deep convolutional neural
[2] Elston, C. W., & Ellis, I. O. (1991). Pathological prognostic network. In Biomedical Imaging (ISBI), 2015 IEEE 12th
factors in breast cancer. I. The value of histological grade in International Symposium on (pp. 55-58). IEEE.
breast cancer: experience from a large study with long‐term [16] Naylor, P., Laé, M., Reyal, F., & Walter, T. (2017, April).
follow‐up. Histopathology, 19(5), 403-410. Nuclei segmentation in histopathology images using deep
neural networks. In Biomedical Imaging (ISBI 2017), 2017
[3] Robbins, P., Pinder, S., De Klerk, N., Dawkins, H., Harvey, IEEE 14th International Symposium on (pp. 933-936). IEEE.
J., Sterrett, G., ... & Elston, C. (1995). Histological grading [17] Shafique, S., & Tehsin, S. (2018). Computer-Aided
of breast carcinomas: a study of interobserver agreement. Diagnosis of Acute Lymphoblastic
Human pathology, 26(8), 873-879. Leukaemia. Computational and mathematical methods in
medicine, 2018.
[4] Meijer, G. A., Beliën, J. A., Van Diest, P. J., & Baak, J. P. [18] Tosta, T. A. A., Neves, L. A., & do Nascimento, M. Z.
(1997). Origins of... image analysis in clinical pathology. (2017). Segmentation methods of H&E-stained histological
Journal of clinical pathology, 50(5), 365. images of lymphoma: a review. Informatics in Medicine
Unlocked, 9, 35-43.
[5] Gurcan, M. N., Boucheron, L., Can, A., Madabhushi, A.,
Rajpoot, N., & Yener, B. (2009). Histopathological image
StuCoSReC Proceedings of the 2018 5th Student Computer Science Research Conference 53
Ljubljana, Slovenia, 9 October
