Presented by: Justin Ziemba MD
University of Pennsylvania

Introduction

Limited information exists on the spatial distribution of stones within the kidney. Spatial mapping of kidney stones offers a quantitative and visual tool to assess the location and frequency of kidney stones in individuals and populations. This study evaluated the feasibility of creating a statistical spatial mapping atlas through CT scan analysis using image segmentation and registration algorithms.

Materials

Non-contrast CT scans of 122 patients with kidney stones were retrospectively identified. A deep learning (DL) image segmentation model was built to segment kidneys automatically. Kidney stones were automatically identified as outlier voxels with Hounsfield Unit (HU) larger than mean plus 5-sigma of HUs of all voxels within the kidney, and a kidney stone image was generated to have unit intensity value at locations of stone voxels and zero otherwise. Bilateral kidney atlases were created by registering all individual images to a common image space using affine-transformation and diffeomorphic deformable image registration algorithms subsequently. The kidney stone image of each kidney was spatially transformed to the corresponding left or right kidney atlas with the same image deformation information obtained for creating the atlases. Statistical spatial maps of the kidney stones were generated for the bilateral kidneys separately to quantify frequency of the stones at every voxel of the kidney atlases.

Results

,

The DL image segmentation model segmented the kidneys with an average dice value of 0.96. The automatic stone identification algorithm detected individual stones with 100% sensitivity. Statistical spatial maps of the kidney stones quantified spatial frequency of the kidney stones across subjects, with frequency up to 6%.

Conclusion

The statistical spatial mapping of kidney stones provides a quantitative and visual means to characterize frequency of stones within the kidneys at both the individual and population level. Precisely characterizing stones in terms of size and location can facilitate effective treatment planning and ensure no stones are “missed” for an individual patient and assess patients’ stone location as compared to the group atlases. In the future we hope to incorporate stone registration and spatial mapping into clinical care as well as better evaluate its performance and utility in this setting.

Funding

NIDDK P20 CHOP/ Penn Center for Machine Learning in Urology (P20DK127488) AUA Care Foundation and SPU Sushil Lacy Research Scholar Award (KMF)

Lead Authors

Jiong Wu, PhD
University of Pennsylvania

Co-Authors

Yuemeng Li, MS
University of Pennsylvania

Benjamin Schurhamer, MD
University of Pennsylvania

Axel Largent, PhD
University of Pennsylvania

Joey Logan, BS
Children's Hospital of Philadelphia

Abhay Singh, BA
Children's Hospital of Philadelphia

Joanie Garrat, MD
University of Pennsylvania

Justin Ziemba, MD
University of Pennsylvania

Gregory Tasian, MD, MSc, MSCE
Children's Hospital of Philadelphia

Yong Fan, PhD
University of Pennsylvania