Abstract—Descriptive features of nuclei positions from tumor tissue samples are studied to construct mathematical models for
tumor growth. Extracted from kidney cancer patients, tumorous tissue pieces are implanted in the flanks of mice to
measure the course of tumor mass, which are then sampled on
glass slides. H&E slides are digitized under light microscope and analyzed to identify the structure of nuclei positions. Using k-means clustering method, the nuclei locations of each H&E
slide are evaluated. The cluster features are used as inputs to our artificial intelligence based personalized tumor growth parameter computation method, called PReP-C. The
exponential linear tumor growth model parameters and the
corresponding growth curves computed by PReP-C are
compared to the preclinical tumor volume measurements. The
correlation between the computed results and the
measurements from 14 H&E pathology slides is encouraging to
build personalized mathematical models for tumor growth.
Index Terms—H&E slide, tumor growth models, kidney
cancer, exponential linear model, k-means clustering.
A. Saribudak and M. U. Uyar are with Dept. of Electrical Eng., The City College of the City University of New York, USA (e-mail: {asaribudak,uyar}@ccny.cuny.edu).
Y. Dong and J. Hsieh are with Memorial Sloan Kettering Cancer Center, New York, USA (e-mail: {dongy,hsiehj}@mskcc.org).
The initial research used in this work was supported by U.S. Army Communications-Electronics RD&E Center contracts W15P7T-09-CS021 and W15P7T-06-C-P217, and by the National Science Foundation grants ECCS-0421159, CNS-0619577 and IIP-1265265. The contents of this
document represent the views of the authors and are not necessarily the official views of, or are endorsed by, the U.S. Government, Department of Defense, and Department of the Army or the U.S. Army Communications
Electronics RD&E Center.
[PDF]
Cite: Aydin Saribudak, Yiyu Dong, James Hsieh, and M. Umit Uyar, "Modeling Tumor Growth for Kidney Cancer Based on
Nuclei Clusters of Pathology Slides," International Journal of Engineering and Technology vol. 8, no. 5, pp. 375-379, 2016.