Fracture Morphology Classification: Local Multiclass Modeling for Multilabel Complexity

TL;DR

This paper introduces a local multiclass modeling approach for classifying fracture morphology, improving accuracy in multilabel diagnosis tasks by reformulating the problem and leveraging public datasets.

Contribution

It presents a novel method to extract fracture morphology by converting a global multilabel task into a local multiclass problem, enhancing classification performance.

Findings

Improved average F1 score by 7.89%

Effective use of public datasets for fracture classification

Performance drops with imperfect fracture detectors

Abstract

Between $15\,\%$ and $45\,\%$ of children experience a fracture during their growth years, making accurate diagnosis essential. Fracture morphology, alongside location and fragment angle, is a key diagnostic feature. In this work, we propose a method to extract fracture morphology by assigning automatically global AO codes to corresponding fracture bounding boxes. This approach enables the use of public datasets and reformulates the global multilabel task into a local multiclass one, improving the average F1 score by $7.89\,\%$. However, performance declines when using imperfect fracture detectors, highlighting challenges for real-world deployment. Our code is available on GitHub.