Image Processing and Machine Learning-Based Classification and Detection of Liver Tumor

Area of Science:

• Hepatology
• Medical Imaging
• Machine Learning

Background:

• The liver performs vital functions including digestion, detoxification, and nutrient metabolism.
• Failure in cell regeneration and removal of damaged cells can lead to liver tumors, categorized as benign or malignant.
• Malignant liver tumors pose a significant health risk.

Purpose of the Study:

• To present an image processing and machine learning-based technique for liver cancer classification and identification.
• To develop a method for distinguishing between benign and malignant liver tumors using computational approaches.

Main Methods:

• Image preprocessing using fuzzy histogram equalization to reduce noise.
• Image segmentation to isolate regions of interest.
• Classification using Radial Basis Function-Support Vector Machine (RBF-SVM), Artificial Neural Network (ANN), and Random Forest algorithms.

Main Results:

• The proposed technique effectively classifies and identifies liver tumors.
• Different machine learning models (RBF-SVM, ANN, Random Forest) were evaluated for their performance in liver cancer detection.
• Image processing techniques were crucial for preparing data for machine learning models.

Conclusions:

• The developed image processing and machine learning approach shows promise for accurate liver cancer diagnosis.
• This technique can aid in the early and precise identification of liver tumors, potentially improving patient outcomes.
• Further research can explore advanced algorithms and larger datasets to enhance diagnostic accuracy.