Lung Cancer Detection from CT Images: Modified Adaptive Threshold Segmentation with Support Vector Machines and...

Area of Science:

• Medical Imaging
• Artificial Intelligence in Healthcare
• Oncology

Background:

• Lung cancer remains a leading cause of cancer-related mortality worldwide.
• Early detection is crucial for improving patient outcomes and survival rates.
• Computed Tomography (CT) scans are a primary tool for lung cancer screening and diagnosis.

Purpose of the Study:

• To implement an advanced modified threshold segmentation and classification model for early and accurate lung cancer detection.
• To enhance the precision of lung cancer diagnosis from CT images.
• To develop a tool that facilitates rapid and accurate decision-making for clinicians.

Main Methods:

• Modified adaptive threshold segmentation was employed for image segmentation.
• Support Vector Machines (SVM) and Artificial Neural Network (ANN) classifiers were utilized for cancer detection.
• The Lung Image Database Consortium (LIDC) dataset, comprising CT scans, was used for model validation.

Main Results:

• The proposed model demonstrated high accuracy in lung cancer detection.
• The Artificial Neural Network (ANN) classifier achieved a 96.3% detection rate.
• The Support Vector Machines (SVM) classifier achieved a 97% detection rate, showcasing world-record accuracy.

Conclusions:

• The developed model shows significant potential for early lung cancer detection, improving treatment efficacy.
• Accurate and early diagnosis facilitates timely intervention when lung tumors are most treatable.
• This method provides valuable information, supporting clinicians in making swift and precise diagnostic decisions.