Categorization & Recognition of Lung Tumor Using Machine Learning Representations.

Area of Science:

• Medical Imaging
• Oncology
• Machine Learning

Background:

• Lung cancer is a prevalent global disease where early detection is challenging due to rapid tumor cell growth.
• Timely identification of anomalies in medical images is critical for effective lung cancer diagnosis and treatment planning.
• Understanding genetic and environmental factors is essential for developing novel lung tumor detection strategies.

Purpose of the Study:

• To develop and evaluate a novel framework for accurate lung cancer detection and diagnosis.
• To investigate the efficacy of Gray Level Co-occurrence Matrix (GLCM) features in identifying lung tumors.
• To enhance early detection rates, reduce the need for invasive surgery, and improve patient survival.

Main Methods:

• Image preprocessing and feature extraction using GLCM.
• Utilizing a probability framework to identify tumor location within lung images.
• Performance analysis comparing GLCM features with histogram features.

Main Results:

• GLCM features demonstrated high accuracy in predicting lung tumors.
• The proposed method effectively identifies tumor positions, offering promising diagnostic outcomes.
• GLCM features proved more accurate than histogram features for lung tumor prediction, despite longer processing times.

Conclusions:

• Early and accurate detection of lung cancer significantly improves treatment options and patient survival rates.
• The GLCM-based approach provides a reliable method for lung cancer diagnosis.
• Further research into machine learning systems for lung cancer detection holds significant potential.