An Aggregation of Aggregation Methods in Computational Pathology

TL;DR

This paper reviews various aggregation methods used in computational pathology for analyzing whole-slide images, categorizes them, compares their effectiveness on a specific prediction task, and discusses future research directions.

Contribution

It provides a comprehensive categorization and comparison of aggregation methods in computational pathology, highlighting their principles, advantages, and limitations.

Findings

Multiple instance learning is a common aggregation approach.

Comparison of methods on a specific WSI prediction task.

Recommendations for future aggregation method development.

Abstract

Image analysis and machine learning algorithms operating on multi-gigapixel whole-slide images (WSIs) often process a large number of tiles (sub-images) and require aggregating predictions from the tiles in order to predict WSI-level labels. In this paper, we present a review of existing literature on various types of aggregation methods with a view to help guide future research in the area of computational pathology (CPath). We propose a general CPath workflow with three pathways that consider multiple levels and types of data and the nature of computation to analyse WSIs for predictive modelling. We categorize aggregation methods according to the context and representation of the data, features of computational modules and CPath use cases. We compare and contrast different methods based on the principle of multiple instance learning, perhaps the most commonly used aggregation method, covering a wide range of CPath literature. To provide a fair comparison, we consider a specific WSI-level prediction task and compare various aggregation methods for that task. Finally, we conclude with a list of objectives and desirable attributes of aggregation methods in general, pros and cons of the various approaches, some recommendations and possible future directions.