| Popis: |
Two-dimensional COSY spectroscopy has been developed as an adjunct for the detection and staging of malignant disease. However the application of 2D spectroscopy clinically has been hampered by difficulties in quantification resulting from variabilities in the cross peak volumes measured in 2D COSY spectra. The aim of this thesis is to optimise data acquisition parameters and post acquisitional data processing parameters for two dimensional (2D) MR spectroscopy for use with biological specimens. In this thesis, the factors affecting cross peak volumes in twodimensional COSY spectra of human thyroid biopsy tissues are systematically measured and catalogued. Model systems of cultured human colorectal cells and standard samples of amino acids are used to demonstrate that similar dependencies in cross peak volume profiles can be obtained across a wide range of biological specimens. Computer simulation of 2D COSY spectra of different spin systems, is used to model the dependence of cross peak volumes on the T2, J coupling and of the sample components. The effects of spin system and on the short and long range couplings are described. This information is used to design a new method of 2D data acquisition in which the time of data acquisition can be reduced by up to 37.5%. A step—shaped or weighted acquisition scheme is used to mould the data set to the desired final data shape in t1, by modulating the number of scans collected in each free induction decay. Compensating window functions (OPERA house - Optimised Processing to Enhance time—Reduced Acquisition) are designed to minimise artefacts arising from the stepped acquisition. This scheme is successfully used to reduce acquisition time, in magnitude-mode and phase—sensitive double quantum filtered lH-lH COSY spectra of amino acid standards and cultured human tumour cells. Finally, a small study of thyroid biopsies is performed where the potential clinical usefulness of these methods are demonstrated by showing that the collection of additional information is possible with increased resolution in a reduced experimental time. |
