Kidney Tumor Segmentation Using Deep Learning

Prashant Jadiya

Abstract: Proper segmentation of Kidney Tumors can assist doctors to detect and diagnose diseases. When it comes to segmentation, U-Net is arguably the most successful segmentation architecture in the medical domain. In this paper, we address the challenge of simultaneous semantic segmentation of kidney tumor and proposed multidimensional 3D U-Net with an attempt to improve it with V-Net and Auto-Encoder architecture. Due to marginally higher dice scores, it was very difficult to choose architecture that will be accurate for segmentation. We did experiment for training of 190 cases that is provided by Kidney Tumor Segmentation Challenge database.

Keywords: Medical Segmentation, Kidney Tumor Nephrometry Segmentation, Multidimensional images, Neural Network, Auto-Encoder

How to Cite?: Prashant Jadiya, "Kidney Tumor Segmentation Using Deep Learning", Volume 8 Issue 12, December 2019, International Journal of Science and Research (IJSR), Pages: 1746-1749, https://www.ijsr.net/getabstract.php?paperid=ART20203792, DOI: https://dx.doi.org/10.21275/ART20203792