ASN's Mission

To create a world without kidney diseases, the ASN Alliance for Kidney Health elevates care by educating and informing, driving breakthroughs and innovation, and advocating for policies that create transformative changes in kidney medicine throughout the world.

learn more

Contact ASN

1401 H St, NW, Ste 900, Washington, DC 20005

email@asn-online.org

202-640-4660

The Latest on Twitter

Kidney Week

Abstract: SA-PO001

Computational Segmentation of Glomeruli to Align Histomorphology With Spatial Transcriptomic Signature

Session Information

  • Bioengineering
    November 05, 2022 | Location: Exhibit Hall, Orange County Convention Center‚ West Building
    Abstract Time: 10:00 AM - 12:00 PM

Category: Bioengineering

  • 300 Bioengineering

Authors

  • Ginley, Brandon, University at Buffalo, Buffalo, New York, United States
  • Melo ferreira, Ricardo, Indiana University, Bloomington, Indiana, United States
  • Balis, Ulysses G. J., University of Michigan Michigan Medicine, Ann Arbor, Michigan, United States
  • Jain, Sanjay, Washington University in St Louis, St Louis, Missouri, United States
  • Tomaszewski, John E., University at Buffalo, Buffalo, New York, United States
  • El-Achkar, Tarek M., Indiana University, Bloomington, Indiana, United States
  • Eadon, Michael T., Indiana University, Bloomington, Indiana, United States
  • Sarder, Pinaki, University at Buffalo, Buffalo, New York, United States

Group or Team Name

  • Kidney Precision Medicine Project
Background

Glomerular histomorphology establishes kidney disease diagnosis and prognosis. Spatial transcriptomics facilitates spatial resolution of molecular signatures superimposed upon histology. We trained a machine learning (ML) method to automatically segment glomeruli and quantify pixel level image features to align with spatial transcriptomics (ST) performed on the same section.

Methods

Brightfield histology images of 17 kidney disease biopsies and 3 reference kidney sections were used for training a deep learning model for glomerular segmentation. We used 3 additional reference images for testing. Unsupervised clustering of ML features was performed across all glomeruli using Seurat. Transcriptomic signatures of 10X Visium ST spots were deconvoluted with Seurat 3.2.3 transfer scores using the kidney precision medicine project (KPMP) snRNA-seq atlas as reference. Association between cell type composition and ML classification was assessed.

Results

The ML glomerular segmentation achieved sensitivity/specificity/precision 0.88/1.0/0.97. The glomeruli were classified into 3 clusters; with cluster one statistically associated with increases in degenerative podocyte; and a second cluster with a glomerular capillary signature (p <0.033).

Conclusion

Our pipeline will serve as a framework to map molecular to histomorphologic data over large areas captured from large sample libraries, and will aide in the development of kidney precision medicine.

Funding

  • NIDDK Support