Unsupervised machine learning for identifying important visual features through bag-of-words using histopathology data from chronic kidney disease.

Journal: Scientific reports
Published Date: Mar 22, 2022

Abstract

Pathologists use visual classification to assess patient kidney biopsy samples when diagnosing the underlying cause of kidney disease. However, the assessment is qualitative, or semi-quantitative at best, and reproducibility is challenging. To discover previously unknown features which predict patient outcomes and overcome substantial interobserver variability, we developed an unsupervised bag-of-words model. Our study applied to the C-PROBE cohort of patients with chronic kidney disease (CKD). 107,471 histopathology images were obtained from 161 biopsy cores and identified important morphological features in biopsy tissue that are highly predictive of the presence of CKD both at the time of biopsy and in one year. To evaluate the performance of our model, we estimated the AUC and its 95% confidence interval. We show that this method is reliable and reproducible and can achieve 0.93 AUC at predicting glomerular filtration rate at the time of biopsy as well as predicting a loss of function at one year. Additionally, with this method, we ranked the identified morphological features according to their importance as diagnostic markers for chronic kidney disease. In this study, we have demonstrated the feasibility of using an unsupervised machine learning method without human input in order to predict the level of kidney function in CKD. The results from our study indicate that the visual dictionary, or visual image pattern, obtained from unsupervised machine learning can predict outcomes using machine-derived values that correspond to both known and unknown clinically relevant features.

Authors

  • Joonsang Lee
    Department of Computational Medicine and Bioinformatics, University of Michigan, Ann Arbor, MI, USA.
  • Elisa Warner
    Department of Computational Medicine and Bioinformatics, University of Michigan, Ann Arbor, MI, USA.
  • Salma Shaikhouni
    Department of Internal Medicine, Nephrology, University of Michigan, Ann Arbor, MI, USA.
  • Markus Bitzer
    Department of Internal Medicine, Nephrology, University of Michigan, Ann Arbor, MI, USA.
  • Matthias Kretzler
    Division of Nephrology, University of Michigan, Ann Arbor, Michigan, USA. Electronic address: kretzler@umich.edu.
  • Debbie Gipson
    Department of Pediatrics, Pediatric Nephrology, University of Michigan, Ann Arbor, MI, USA.
  • Subramaniam Pennathur
    Department of Internal Medicine, Nephrology, University of Michigan, Ann Arbor, MI, USA.
  • Keith Bellovich
    Department of Internal Medicine, Nephrology, St. Clair Nephrology Research, Detroit, MI, USA.
  • Zeenat Bhat
    Department of Internal Medicine, Nephrology, Wayne State University, Detroit, MI, USA.
  • Crystal Gadegbeku
    Department of Internal Medicine, Nephrology, Cleveland Clinic, Cleveland, OH, USA.
  • Susan Massengill
    Department of Pediatrics, Pediatric Nephrology, Levine Children's Hospital, Charlotte, NC, USA.
  • Kalyani Perumal
    Department of Internal Medicine, Nephrology, Department of JH Stroger Hospital, Chicago, IL, USA.
  • Jharna Saha
    Department of Pathology, University of Michigan, Ann Arbor, MI, USA.
  • Yingbao Yang
    Department of Pathology, University of Michigan, Ann Arbor, MI, USA.
  • Jinghui Luo
    Department of Pathology, University of Michigan, Ann Arbor, MI, USA.
  • Xin Zhang
    First Department of Infectious Diseases, The First Affiliated Hospital of China Medical University, Shenyang, China.
  • Laura Mariani
    Department of Internal Medicine, Nephrology, University of Michigan, Ann Arbor, MI, USA.
  • Jeffrey B Hodgin
    Department of Pathology, University of Michigan, Ann Arbor, Michigan, USA.
  • Arvind Rao
    Department of Bioinformatics and Computational Biology, The University of Texas MD Anderson Cancer Center, Houston, Texas.

Keywords

External Resources

Popular Topics
Recent Journals