A Framework for Lung and Colon Cancer Diagnosis via Lightweight Deep Learning Models and Transformation Methods.

Area of Science:

• Oncology
• Medical Imaging
• Artificial Intelligence

Background:

• Lung and colon cancers are leading causes of mortality and morbidity.
• Early diagnosis and histopathological detection are critical for effective cancer treatment.
• Traditional deep learning models for cancer diagnosis often require significant computational resources.

Purpose of the Study:

• To propose a novel framework for the early detection of lung and colon cancers.
• To develop a computationally efficient deep learning approach for cancer diagnosis.
• To improve the accuracy and speed of histopathological analysis.

Main Methods:

• Utilized multiple lightweight deep learning models (ShuffleNet, MobileNet, SqueezeNet) for feature extraction.
• Applied feature reduction techniques including Principal Component Analysis (PCA) and Fast Walsh-Hadamard Transform (FHWT).
• Employed Discrete Wavelet Transform (DWT) for feature fusion and fed reduced features into machine learning algorithms.

Main Results:

• Achieved a highest accuracy of 99.6% in distinguishing lung and colon cancer variants.
• Demonstrated reduced computational complexity and feature requirements compared to existing methods.
• Showcased the effectiveness of transformation methods for feature reduction and improved data interpretation.

Conclusions:

• The proposed lightweight deep learning framework enables efficient and accurate early detection of lung and colon cancers.
• Feature reduction and fusion techniques enhance diagnostic performance.
• This approach offers a cost-effective and faster alternative for cancer diagnosis in histopathology.