Robust gray-level standardization in brain Magnetic-Resonance images. G. De Nunzio1, R. Cataldo1, A. Carlà1. (1) University of Salento, Dept. of Mathematics and Physics, and INFN, Lecce Purpose: it is known that intensities in MRI do not have a fixed tissue-specific numeric meaning, even within the same MRI protocol, for the same body region, or for images of the same patient obtained on the same scanner in different moments. Consequently many problems can arise in large multi-site clinical studies, making the interpretation of results difficult or confused, or affecting post processing phases such as segmentation and registration. In spite of the fact that the lack of a standard and quantifiable interpretation compromises the precision, accuracy, and efficiency of those applications, few papers have explicitly addressed the problems. In this context, we propose a tiSsue-Based Standardization Technique (SBST) of MR brain images. Methods and materials: the system was developed and tested on a large number of images, belonging to healthy people and to patients with different degrees of neurodegenerative pathology, obtained from public databases and the clinical practice. Both histogram and tissue-specific intensity information were used, performing piecewise linear intensity transformations between images, so sharing the simplicity and robustness of landmark techniques, while remaining fully automated and quite light from the computational point of view. Results: the efficacy in minimizing the risk of “mixing” brain tissues during intensity transformations was assessed, and particular attention was devoted to a thorough examination of the benefits comparing SBST with other approaches available in the literature. Conclusion: the technique proved robust in standardizing tissues, giving similar intensities to similar tissues, even across images coming from different sources.
Robust gray-level standardization in brain Magnetic-Resonance images
DE NUNZIO, Giorgio;CATALDO, Rosella;
2014-01-01
Abstract
Robust gray-level standardization in brain Magnetic-Resonance images. G. De Nunzio1, R. Cataldo1, A. Carlà1. (1) University of Salento, Dept. of Mathematics and Physics, and INFN, Lecce Purpose: it is known that intensities in MRI do not have a fixed tissue-specific numeric meaning, even within the same MRI protocol, for the same body region, or for images of the same patient obtained on the same scanner in different moments. Consequently many problems can arise in large multi-site clinical studies, making the interpretation of results difficult or confused, or affecting post processing phases such as segmentation and registration. In spite of the fact that the lack of a standard and quantifiable interpretation compromises the precision, accuracy, and efficiency of those applications, few papers have explicitly addressed the problems. In this context, we propose a tiSsue-Based Standardization Technique (SBST) of MR brain images. Methods and materials: the system was developed and tested on a large number of images, belonging to healthy people and to patients with different degrees of neurodegenerative pathology, obtained from public databases and the clinical practice. Both histogram and tissue-specific intensity information were used, performing piecewise linear intensity transformations between images, so sharing the simplicity and robustness of landmark techniques, while remaining fully automated and quite light from the computational point of view. Results: the efficacy in minimizing the risk of “mixing” brain tissues during intensity transformations was assessed, and particular attention was devoted to a thorough examination of the benefits comparing SBST with other approaches available in the literature. Conclusion: the technique proved robust in standardizing tissues, giving similar intensities to similar tissues, even across images coming from different sources.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
