Preliminary assessment of dispersion versus absorption analysis of high spectral and spatial resolution magnetic resonance images in the diagnosis of breast cancer

摘要: Water resonance lineshapes observed in breast lesions imaged with high spectral and spatial resolution (HiSS) magnetic imaging have been shown to contain diagnostically useful non-Lorentzian components. The purpose of this work is update a previous method lesion diagnosis by including phase-corrected absorption dispersion spectra. This includes information about the shape complex water resonance, which could improve performance computer-aided classification scheme. characteristics spectra are characterized comparing plot real versus imaginary components spectrum that perfect Lorentzian spectrum, "dispersion absorption" (DISPA) analysis technique. Distortion indicates underlying physiologic changes, be correlated malignancy. These distortions each voxel quantified summing deviations DISPA radius from an ideal over all Fourier components, yielding "total radial difference" (TRD). We limited our those voxels largest TRD. number considered was dependent on size. TRD used classify 15 malignant 8 benign ([Formula: see text] after elimination). Lesion discrimination evaluated for both average variance within lesion. Area under receiver operating characteristic curve (ROC AUC) assess voxel- lesion-based methods task distinguishing between benign. In lesions, yielded AUC 0.89 (95% confidence interval [0.84, 0.91]). difference 0.90 [0.71, 1.00]) 0.84 [0.61, 0.99]). applied spectroscopic HiSS data order identify quantify displaying characteristics. scheme based obtained acquisitions may outperform similar classifier single off-peak component analysis, as it uses details entire instead magnitude at location.