Statistical Parameter-based Automatic Liver Tumor Segmentation from Abdominal CT Scans: A Potiential Radiomic Signature

• Published in: Procedia computer science Vol. 93; pp. 446 - 452
• Main Authors: Kumar, Y. Rakesh, Muthukrishnan, N. Moorthy, Mahajan, Abhishek, Priyanka, P., Padmavathi, G., Nethra, M., Sneha, R., Thakur, Meenakshi H.
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 2016
• Subjects: Biomarker; CT scan; Morphological operations; Radiomics; Region growing; Statistical parameters; Thresholding; Biomarker; Region growing; Thresholding; Statistical parameters; CT scan; Morphological operations; Radiomics
• ISSN: 1877-0509; 1877-0509
• DOI: 10.1016/j.procs.2016.07.232

Liver imaging using abdominal CT images has been widely studied in the recent years and is a challenging task. Processing CT image includes the automatic diagnosis of liver and lesions part. Because of the high intensity similarity between liver tissues and nearby organs of liver it is difficult to segment liver and tumor. Segmentation of extracted region as an imaging biomarker forms an essential component of “Radiomics”. This paper presents automatic liver tumor segmentation from abdominal CT scan images. A statistical parameter-based approach is used to distinguish liver tumor tissue from other abdominal organs. The existing segmentation methods such as region growing and intensity based thresholding methods are investigated. First, the CT images are preprocessed by filtering to remove noise from the image. Then the statistical mean-based thresholding is applied to extract the tumor. After applying median filtering, isodata threshold is used to turn the image into binary with tumors as black spots on white background. Finally postprocessing as filtering techniques like mean filter and median filter and morphological operations are performed to remove residues. This paper highlights liver tumor segmentation analysis which has the potential as a imaging biomarker for “Personalised cancer imaging”.