It has been recognized for over a decade that the nuclear magnetic resonance (NMR) relaxation data (the spin-lattice relaxation time, T1, and the spin-spin relaxation time T2) of the composite proton signal from tissues can be used to distinguish malignant from benign tissue. The use of proton T1 and T2 values for the detection of malignancy has, in the past year, begun to gain clinical relevancy, since NMR imaging instruments capable of making such measurements are beginning clinical trials. However, the use of proton relaxation data for detection of malignancy has neither the firm scientific foundation nor the quality of data which is required in order to differentiate between malignant and benign tissue with a level of confidence that is required in order to base clinical decisions on such NMR data. In order to improve this situation our overall goal in this research program is to test the hypothesis that by examination of the proton NMR spectrum and the relaxation data of each of its resolvable resonances resulting from the various classes of protons (water, aliphatic, aromatic, etc.) which contribute to the composite proton signal, a substantially improved basis for differentiation between malignant and benign tissue can be achieved. Such an improved basis will make possible diagnosis of malignancy by NMR imaging with high levels of confidence. Additionally, the information derived from these studies will be the basis for establishing a rational mechanism for the empirical findings. Preliminary studies in our laboratory have proven that this approach can be successful. We have shown in these studies that T1 values of certain individual resolvable proton resonances show a much greater difference between normal and malignant tissue than does the composite T1 of all proton resonances grouped together and that T1 values of other resonances (e.g. water) show little or no differences between normal and malignant tissue.
