| Quantification of the data in medical imaging not only requires signal quantitation and localization but also a rational as well as metric visual presentation. Some aspects of this in two of the major medical imaging techniques: (a) magnetic resonance spectroscopy and imaging (MRSI) and (b) positron emission tomography (PET); are explored.;The inherent signal characteristic---the longitudinal relaxation time constant of the proton nuclear magnetic resonance (NMR) signal from water---is used to discriminate between three brain tissue types: cerebrospinal fluid (CSF), gray matter (GM) and white matter (WM). Signals from the three brain tissue types have different T1 ranges in the relaxogram , the almost continuous probability distribution of the relaxation time constants. By integrating the appropriate regions of the relaxogram, edited relaxographic images are created.;A novel method---PETAMRI---for the intimate combination of two different real-space images (PET and MRI) of the same tissue is developed by utilizing the three-dimensional nature of color. The use of color scales for a rational quantitative presentation of the medical data is explored.;The spatial response functions (SRFs) of the PET and MR images and their effects on image resolution and signal quantitation are presented. A method for correction of MR images for B1 inohomogeneity is also presented and illustrated for a phantom sample. Also, the SRF and quantitative tissue segmentation are utilized to characterize the transverse relaxation time, T2, of brain ethanol (methyl proton) signal using MRSI. There is a significant difference between the T2 values in two tissues (GM and WM). This has further implications in the interaction of ethanol in human brain and its intoxication effect. |
