Logo image
Back
Applications of Topological Data Analysis in Oncology
Journal article   Open access   Peer reviewed

Applications of Topological Data Analysis in Oncology

Anuraag Bukkuri, Noemi Andor and Isabel K Darcy
Frontiers in artificial intelligence, Vol.4, pp.659037-659037
04/13/2021
DOI: 10.3389/frai.2021.659037
url
https://doi.org/10.3389/frai.2021.659037View
Published (Version of record) Open Access

Abstract

The emergence of the information age in the last few decades brought with it an explosion of biomedical data. But with great power comes great responsibility: there is now a pressing need for new data analysis algorithms to be developed to make sense of the data and transform this information into knowledge which can be directly translated into the clinic. Topological data analysis (TDA) provides a promising path forward: using tools from the mathematical field of algebraic topology, TDA provides a framework to extract insights into the often high-dimensional, incomplete, and noisy nature of biomedical data. Nowhere is this more evident than in the field of oncology, where patient-specific data is routinely presented to clinicians in a variety of forms, from imaging to single cell genomic sequencing. In this review, we focus on applications involving persistent homology, one of the main tools of TDA. We describe some recent successes of TDA in oncology, specifically in predicting treatment responses and prognosis, tumor segmentation and computer-aided diagnosis, disease classification, and cellular architecture determination. We also provide suggestions on avenues for future research including utilizing TDA to analyze cancer time-series data such as gene expression changes during pathogenesis, investigation of the relation between angiogenic vessel structure and treatment efficacy from imaging data, and experimental confirmation that geometric and topological connectivity implies functional connectivity in the context of cancer.
 Artificial Intelligence clonal evolution imaging oncology persistent homology single cell analysis topological data analysis tumor heterogeneity

Details

Metrics

26 Record Views
41 Times Cited - Web of Science
Logo image