Related Experiment Video
Updated: Jan 8, 2026
07:53
Author Spotlight: Advancing 3D Modeling for Enhanced Diagnosis and Treatment of Pulmonary Nodules in Early-Stage Lung Cancer
Published on: October 13, 2023
2.0K
Enhanced Diagnosis of Lung and Colon Cancer Severity Through Deep Feature Analysis Using DenseNet201 and SVM With
Pragati Patharia1, B Surya Prasad Rao2, Prabira Kumar Sethy3
1Department of ECE, Guru Ghasidas Vishwavidyalaya, Bilaspur, India.
Cancer Reports (Hoboken, N.J.)
|December 23, 2025
Summary
This study introduces a deep learning model combining Fast Super-Resolution Convolutional Neural Network (FSRCNN) and Support Vector Machine (SVM) for accurate lung and colon cancer detection. The novel approach significantly enhances diagnostic accuracy, achieving 98% overall accuracy.
Area of Science:
- Oncology
- Medical Imaging
- Artificial Intelligence
Background:
- Rising lung and colon cancer prevalence necessitates improved diagnostic methods.
- Histopathologist expertise is crucial but limited, impacting patient outcomes.
- Deep learning offers advanced solutions for medical image analysis.
Purpose of the Study:
- To develop a novel deep learning diagnostic method for enhanced lung and colon cancer detection.
- To improve the accuracy and reliability of cancer diagnosis through advanced algorithms.
Main Methods:
- Utilized the LC25000 dataset (25,000 histopathological images).
- Implemented Fast Super-Resolution Convolutional Neural Network (FSRCNN) for image enhancement.
- Employed a Support Vector Machine (SVM) model with DenseNet201 for classification.
Main Results:
- Achieved 98.00% overall accuracy, 98.10% precision, 98% sensitivity, and 99.50% specificity.
- Demonstrated superior performance compared to existing Convolutional Neural Network (CNN) models.
- The FSRCNN and SVM-based DenseNet201 model significantly improved diagnostic reliability.
Conclusions:
- The FSRCNN and SVM-based DenseNet201 model substantially improved lung and colon cancer detection.
- Further validation with diverse datasets and integration of genetic/EHR data are recommended.
- The model shows promise for practical application in cancer diagnosis.
