Journal of Medical Physics and Applied Sciences

Automated Liver Surface Texture Analysis with Irregularity Index for Improved Classification of Hepatic Disorder

Background: The liver plays a crucial role in various physiological and pathological processes, making accurate analysis of its surface texture essential for diagnostic and therapeutic purposes. Automated methods, particularly those employing machine learning techniques have shown promise in this domain. Purpose: To enhance the classification accuracy of liver surface texture analysis by introducing a novel feature, the irregularity index. This paper investigates the potential of this approach to provide more reliable and accurate classification results in comparison to existing methods.

Materials and methods: To achieve our research goals, we collected a dataset of liver surface textures and developed a methodology that incorporates SVM as a classification tool. We introduced the irregularity index as a new feature to capture irregularities in liver texture patterns. This feature, in combination with SVM, was employed for the automated analysis of liver surface textures.

Results: Our experimental results indicate that the inclusion of the irregularity index significantly improves the classification accuracy of liver surface textures when compared to traditional methods. The SVM-based approach, coupled with the irregularity index, achieved an impressive level of accuracy, demonstrating the potential for more precise classification of liver surface textures.

Conclusions: In conclusion, the integration of the irregularity index with support vector machine-based analysis represents a promising advancement in the field of automated liver surface texture analysis. This innovative approach has the potential to enhance the accuracy and reliability of liver disease diagnosis and prognosis., ensuring optimal radiation protection for patients, staff, and the public.

Author(s): Huasi Hu*, Saadat Ali, Ahmer Saleem and Mingfei Yan

Abstract | PDF

Share this