ECM-CSD: An Efficient Classification Model for Cancer Stage Diagnosis in CT Lung Images Using FCM and SVM Techniques

Area of Science:

• Medical Imaging
• Computer-Aided Diagnosis
• Oncology

Background:

• Lung cancer diagnosis is challenging due to complex cell structures and superimposed cells.
• Early detection and accurate staging are crucial for improving patient survival rates.
• Image processing techniques offer potential for enhancing diagnostic accuracy and efficiency.

Purpose of the Study:

• To propose an Efficient Classification Model for Cancer Stage Diagnosis (ECM-CSD) for lung cancer.
• To develop a region-based Fuzzy C-Means Clustering (FCM) technique for lung cancer segmentation.
• To implement a Support Vector Machine (SVM) based classification for diagnosing lung cancer stages.

Main Methods:

• Utilized Computed Tomography (CT) lung images from the LIDC-IDRI dataset.
• Applied Gaussian and Gabor filters for image pre-processing (smoothing and enhancement).
• Employed FCM for lung nodule segmentation and SVM for cancer stage classification.

Main Results:

• The proposed ECM-CSD model demonstrated high accuracy in segmenting lung nodules and classifying cancer stages.
• Comparative experiments showed improved performance with increased accuracy and reduced error rates.
• The model effectively aids in clinical practice for lung cancer diagnosis.

Conclusions:

• The developed image processing model significantly enhances the accuracy and efficiency of lung cancer diagnosis.
• FCM and SVM integration provides a robust approach for segmentation and staging.
• This technique holds promise for improving lung cancer patient outcomes through earlier and more precise diagnosis.