Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Explainability of Deep Neural Networks for Brain Tumor Detection

S. Park, J. Kim

TL;DR

This study applies explainable AI (XAI) techniques to assess the performance of various models on real-world medical data and finds that CNNs with shallower architectures are more effective for small datasets and can support medical decision-making.

Abstract

Medical image classification is crucial for supporting healthcare professionals in decision-making and training. While Convolutional Neural Networks (CNNs) have traditionally dominated this field, Transformer-based models are gaining attention. In this study, we apply explainable AI (XAI) techniques to assess the performance of various models on real-world medical data and identify areas for improvement. We compare CNN models such as VGG-16, ResNet-50, and EfficientNetV2L with a Transformer model: ViT-Base-16. Our results show that data augmentation has little impact, but hyperparameter tuning and advanced modeling improve performance. CNNs, particularly VGG-16 and ResNet-50, outperform ViT-Base-16 and EfficientNetV2L, likely due to underfitting from limited data. XAI methods like LIME and SHAP further reveal that better-performing models visualize tumors more effectively. These findings suggest that CNNs with shallower architectures are more effective for small datasets and can support medical decision-making.

Explainability of Deep Neural Networks for Brain Tumor Detection

TL;DR

This study applies explainable AI (XAI) techniques to assess the performance of various models on real-world medical data and finds that CNNs with shallower architectures are more effective for small datasets and can support medical decision-making.

Abstract

Medical image classification is crucial for supporting healthcare professionals in decision-making and training. While Convolutional Neural Networks (CNNs) have traditionally dominated this field, Transformer-based models are gaining attention. In this study, we apply explainable AI (XAI) techniques to assess the performance of various models on real-world medical data and identify areas for improvement. We compare CNN models such as VGG-16, ResNet-50, and EfficientNetV2L with a Transformer model: ViT-Base-16. Our results show that data augmentation has little impact, but hyperparameter tuning and advanced modeling improve performance. CNNs, particularly VGG-16 and ResNet-50, outperform ViT-Base-16 and EfficientNetV2L, likely due to underfitting from limited data. XAI methods like LIME and SHAP further reveal that better-performing models visualize tumors more effectively. These findings suggest that CNNs with shallower architectures are more effective for small datasets and can support medical decision-making.

Paper Structure

This paper contains 31 sections, 13 figures, 9 tables.

Table of Contents

  1. Introduction
  2. Related work
  3. Deep learning and XAI in healthcare
  4. Classification Models
  5. VGG-16 architecture
  6. ResNet-50 architecture
  7. EfficientNet architecture
  8. ViT architecture
  9. Explainable AI Models
  10. LIME
  11. SHAP
  12. Grad-CAM
  13. Method
  14. Data
  15. Transfer Learning Experiments Design
  16. ...and 16 more sections

Figures (13)

  • Figure 1: Block diagram of VGG-16 network tammina2019transfer
  • Figure 2: ResNet-50 architectureshatnawi2023deep
  • Figure 3: EfficientNet architecturexu2021forest
  • Figure 4: ViT architecturedosovitskiy2020image
  • Figure 5: Explaining the predictions using LIMEribeiro2016should
  • ...and 8 more figures