Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Improving Medical Waste Classification with Hybrid Capsule Networks

Bennet van den Broek, Javad Pourmostafa Roshan Sharami

TL;DR

This study targets the environmental and public health risks of improper medical waste disposal by improving classification through a hybrid architecture that combines a pretrained DenseNet121 with capsule networks to preserve spatial relationships. Using a large, diverse, multi-source dataset, the authors compare a DenseNet baseline, a frozen DenseNet with a capsule classifier, and an unfrozen DenseNet with a capsule classifier. Results show the unfrozen DenseNet + Capsule configuration achieving the highest $F1$-score around $0.918$, indicating that capsule networks can enhance robustness to transformations in medical waste images when the feature extractor is fine-tuned. The work underscores the importance of dataset diversity for generalization and discusses practical considerations for real-world deployment, while outlining future directions such as expanded augmentation and benchmarked comparisons to advance medical waste management systems.

Abstract

The improper disposal and mismanagement of medical waste pose severe environmental and public health risks, contributing to greenhouse gas emissions and the spread of infectious diseases. Efficient and accurate medical waste classification is crucial for mitigating these risks. We explore the integration of capsule networks with a pretrained DenseNet model to improve medical waste classification. To the best of our knowledge, capsule networks have not yet been applied to this task, making this study the first to assess their effectiveness. A diverse dataset of medical waste images collected from multiple public sources, is used to evaluate three model configurations: (1) a pretrained DenseNet model as a baseline, (2) a pretrained DenseNet with frozen layers combined with a capsule network, and (3) a pretrained DenseNet with unfrozen layers combined with a capsule network. Experimental results demonstrate that incorporating capsule networks improves classification performance, with F1 scores increasing from 0.89 (baseline) to 0.92 (hybrid model with unfrozen layers). This highlights the potential of capsule networks to address the spatial limitations of traditional convolutional models and improve classification robustness. While the capsule-enhanced model demonstrated improved classification performance, direct comparisons with prior studies were challenging due to differences in dataset size and diversity. Previous studies relied on smaller, domain-specific datasets, which inherently yielded higher accuracy. In contrast, our study employs a significantly larger and more diverse dataset, leading to better generalization but introducing additional classification challenges. This highlights the trade-off between dataset complexity and model performance.

Improving Medical Waste Classification with Hybrid Capsule Networks

TL;DR

This study targets the environmental and public health risks of improper medical waste disposal by improving classification through a hybrid architecture that combines a pretrained DenseNet121 with capsule networks to preserve spatial relationships. Using a large, diverse, multi-source dataset, the authors compare a DenseNet baseline, a frozen DenseNet with a capsule classifier, and an unfrozen DenseNet with a capsule classifier. Results show the unfrozen DenseNet + Capsule configuration achieving the highest -score around , indicating that capsule networks can enhance robustness to transformations in medical waste images when the feature extractor is fine-tuned. The work underscores the importance of dataset diversity for generalization and discusses practical considerations for real-world deployment, while outlining future directions such as expanded augmentation and benchmarked comparisons to advance medical waste management systems.

Abstract

The improper disposal and mismanagement of medical waste pose severe environmental and public health risks, contributing to greenhouse gas emissions and the spread of infectious diseases. Efficient and accurate medical waste classification is crucial for mitigating these risks. We explore the integration of capsule networks with a pretrained DenseNet model to improve medical waste classification. To the best of our knowledge, capsule networks have not yet been applied to this task, making this study the first to assess their effectiveness. A diverse dataset of medical waste images collected from multiple public sources, is used to evaluate three model configurations: (1) a pretrained DenseNet model as a baseline, (2) a pretrained DenseNet with frozen layers combined with a capsule network, and (3) a pretrained DenseNet with unfrozen layers combined with a capsule network. Experimental results demonstrate that incorporating capsule networks improves classification performance, with F1 scores increasing from 0.89 (baseline) to 0.92 (hybrid model with unfrozen layers). This highlights the potential of capsule networks to address the spatial limitations of traditional convolutional models and improve classification robustness. While the capsule-enhanced model demonstrated improved classification performance, direct comparisons with prior studies were challenging due to differences in dataset size and diversity. Previous studies relied on smaller, domain-specific datasets, which inherently yielded higher accuracy. In contrast, our study employs a significantly larger and more diverse dataset, leading to better generalization but introducing additional classification challenges. This highlights the trade-off between dataset complexity and model performance.

Paper Structure

This paper contains 32 sections, 9 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Context
  3. Research Questions
  4. Main Findings
  5. Related Work
  6. Using Deep Learning Models for Medical Waste Classification
  7. Using Capsule Networks for General Waste Classification
  8. Using Capsule Networks for Medical Waste Classification
  9. Data Collection and Processing
  10. Dataset Creation
  11. Data Augmentation
  12. Data Normalization and Formatting
  13. Models
  14. Baseline––DenseNet121
  15. Capsule Networks
  16. ...and 17 more sections

Figures (9)

  • Figure 1: Class distribution before data augmentation
  • Figure 2: Class distribution after data augmentation
  • Figure 3: Methodological pipeline detailing feature extraction and classification steps.
  • Figure 4: Confusion matrix of the baseline model (DenseNet121).
  • Figure 5: Confusion matrix of the frozen DenseNet121 with the capsule network.
  • ...and 4 more figures