Lung Cancer Classification and Prediction Using Machine Learning and Image Processing.

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Oncology

Background:

• Lung cancer detection remains a significant challenge for medical professionals.
• Early detection is crucial for effective lung cancer treatment.
• Current methods for identifying cancerous lung regions often involve complex image processing techniques.

Purpose of the Study:

• To develop and evaluate an accurate classification and prediction system for lung cancer.
• To leverage machine learning and image processing for enhanced lung cancer diagnosis.
• To compare the efficacy of different machine learning models in lung cancer prediction.

Main Methods:

• Utilized a dataset of 83 CT scans from 70 patients for experimental investigation.
• Applied geometric mean filtering for image preprocessing to enhance quality.
• Employed K-means clustering for image segmentation to identify relevant lung regions.
• Implemented and compared Artificial Neural Network (ANN), K-Nearest Neighbors (KNN), and Random Forest (RF) for machine learning classification.

Main Results:

• The geometric mean filter effectively improved the quality of CT scan images.
• Image segmentation using K-means successfully isolated regions of interest.
• The Artificial Neural Network (ANN) model achieved the highest accuracy in lung cancer classification and prediction.
• ANN outperformed KNN and RF in the experimental evaluation.

Conclusions:

• Machine learning combined with image processing offers a promising approach for accurate lung cancer detection.
• The ANN model shows significant potential for clinical application in early lung cancer diagnosis.
• Further research can explore larger datasets and advanced algorithms for improved lung cancer prediction.