A geometry-first tutorial for time-resolved morphological analysis with PyPETANA

TL;DR

This paper introduces PyPETANA, an open-source Python framework with a tutorial for geometry-first, time-resolved analysis of evolving morphologies from image data, emphasizing reproducibility and multiscale boundary analysis.

Contribution

It provides a reproducible tutorial and framework for geometry-based, time-resolved morphological analysis directly from images, supporting multiscale boundary analysis and benchmarking.

Findings

Demonstrates applicability to tumor morphologies

Shows analysis of boundary evolution in cancer growth

Provides a reproducible computational workflow

Abstract

We present a step-by-step, reproducible tutorial for PyPETANA, an open-source Python framework for geometry-first, time-resolved quantification of evolving morphology from image data. Starting from time-lapse video input, the tutorial demonstrates how to extract binary masks, compute time-resolved geometric observables including area, perimeter, circularity, and effective fractal dimensions, and analyze their temporal evolution. The workflow emphasizes direct reconstruction of morphology from images without assuming microscopic growth mechanisms. In addition to compactness-sensitive geometric descriptors, the framework supports multiscale boundary analysis through supersampled box-counting methods applied to filled morphologies and finite-width boundary bands. The benchmark suite further demonstrates applicability to invasive tumor morphologies and multiscale boundary evolution in time-resolved cancer-growth interfaces. This tutorial accompanies the computational workflow underlying arXiv:2602.05958 and provides a reproducible foundation for geometry-based analysis of evolving non-equilibrium morphologies.