Ensemble learning driven Kolmogorov-Arnold Networks-based Lung Cancer classification
2024

Lung Cancer Detection Using Advanced Imaging Techniques

Sample size: 20894 publication 10 minutes Evidence: high

Author Information

Author(s): Sait Abdul Rahaman Wahab, AlBalawi Eid, Nagaraj Ramprasad

Primary Institution: King Faisal University

Hypothesis

Can ensemble learning and feature fusion improve lung cancer detection from PET/CT images?

Conclusion

The proposed lung cancer detection model achieved an accuracy of 99.0%, demonstrating its effectiveness in clinical practice.

Supporting Evidence

  • The model achieved an accuracy of 99.0% with a minimal loss of 0.07.
  • Five-fold cross-validation was used to evaluate the model.
  • The proposed model requires limited resources for classification.
  • Integrating LIME provided interpretability of the model's predictions.
  • The study highlights the importance of ensemble learning in medical image analysis.

Takeaway

This study created a smart computer program that helps doctors find lung cancer early by looking at special pictures of the lungs.

Methodology

The study used PET/CT images and enhanced MobileNet V3 and LeViT models for feature extraction, followed by classification using Kolmogorov-Arnold Networks.

Potential Biases

Potential dataset bias due to variability in medical imaging data and patient demographics.

Limitations

The model's performance may vary with different datasets and is limited by the quality of PET/CT images.

Participant Demographics

The dataset included various lung cancer types with a total of 20894 images.

Statistical Information

P-Value

0.0001

Confidence Interval

[95.4–96.7]

Statistical Significance

p<0.05

Digital Object Identifier (DOI)

10.1371/journal.pone.0313386

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication