Surface distance metrics

Summary

When comparing multiple image segmentations, performance metrics that assess how closely the surfaces align can be a useful difference measure. This group of surface distance based measures computes the closest distances from all surface points on one segmentation to the points on another surface, and returns performance metrics between the two. This distance can be used alongside other metrics to compare segmented regions against a ground truth.

Surfaces are represented using surface elements with corresponding area, allowing for more consistent approximation of surface measures.

Metrics included

This library computes the following performance metrics for segmentation:

• Average surface distance (see compute_average_surface_distance)
• Hausdorff distance (see compute_robust_hausdorff)
• Surface overlap (see compute_surface_overlap_at_tolerance)
• Surface dice (see compute_surface_dice_at_tolerance)
• Volumetric dice (see compute_dice_coefficient)

Installation

First clone the repo, then install the dependencies and surface-distance package via pip:

$ git clone https://github.com/deepmind/surface-distance.git
$ pip install surface-distance/

Usage

For simple usage examples, see surface_distance_test.py.