Format

Send to

Choose Destination
Sci Rep. 2017 Nov 6;7(1):14588. doi: 10.1038/s41598-017-15092-3.

Interactive phenotyping of large-scale histology imaging data with HistomicsML.

Author information

1
Department of Biomedical Informatics, Emory University School of Medicine, Atlanta, USA.
2
Department of Neurology, Emory University School of Medicine, Atlanta, USA.
3
Emory University School of Medicine, Atlanta, USA.
4
Department of Pathology & Laboratory Medicine, Emory University School of Medicine, Atlanta, USA.
5
Winship Cancer Institute, Emory University, Atlanta, USA.
6
Department of Biomedical Informatics, Emory University School of Medicine, Atlanta, USA. lee.cooper@emory.edu.
7
Winship Cancer Institute, Emory University, Atlanta, USA. lee.cooper@emory.edu.
8
Department of Biomedical Engineering, Georgia Institute of Technology/Emory University School of Medicine, Atlanta, GA, USA. lee.cooper@emory.edu.

Abstract

Whole-slide imaging of histologic sections captures tissue microenvironments and cytologic details in expansive high-resolution images. These images can be mined to extract quantitative features that describe tissues, yielding measurements for hundreds of millions of histologic objects. A central challenge in utilizing this data is enabling investigators to train and evaluate classification rules for identifying objects related to processes like angiogenesis or immune response. In this paper we describe HistomicsML, an interactive machine-learning system for digital pathology imaging datasets. This framework uses active learning to direct user feedback, making classifier training efficient and scalable in datasets containing 108+ histologic objects. We demonstrate how this system can be used to phenotype microvascular structures in gliomas to predict survival, and to explore the molecular pathways associated with these phenotypes. Our approach enables researchers to unlock phenotypic information from digital pathology datasets to investigate prognostic image biomarkers and genotype-phenotype associations.

Supplemental Content

Full text links

Icon for Nature Publishing Group Icon for PubMed Central
Loading ...
Support Center