Improving proton MR spectroscopy of brain tissue for noninvasive diagnostics

Abstract

Purpose: To examine preprocessing methods affecting the potential use of Magnetic Resonance Spectroscopy (MRS) as a noninvasive modality for detection and characterization of brain lesions and for directing therapy progress. Materials and Methods:Two reference point re-calibration with linear interpolation (to compensate for magnetic field nonhomogeneity), weighting of spectra (to emphasize consistent peaks and depress chemical noise), and modeling based on chemical shift locations of 97 biomarkers were investigated. Results for 139 categorized scans were assessed by comparing Leave-One-Out (LOO) cross-validation and external validation.Results: For distinction of nine brain tissue categories, use of re-calibration, variance weighting, and biomarker modeling improved LOO classification of MRS spectra from 31% to 95%. External validation of the two best nine-category models on 47 unknown samples gave 96% or 100% accuracy, respectively, compared with pathological diagnosis. Conclusion:Preprocessing of MRS spectra can significantly improve their diagnostic utility for automated consultation of pattern recognition models. Use of several techniques in combination greatly increases available proton MRS information content. Accurate assignment of unknowns among nine tissue classes represents a significant improvement, for a much more demanding task, than has been previously reported.

A Review of Applications of Metabolomics in Cancer

“…MRSI detection of choline has been reported to be 100% sensitive for detection of malignant versus benign breast tumors and could be used to reduce the need of unnecessary breast biopsies [138]. The ability to use single voxel MRSI data to classify nine different types of brain cancer has been successful [141]. Alusta and coworkers used four steps to process the MRSI data before building a classification modeling that included data normalization, re-calibration of the spectra to specific peaks, weighting of the data, and re-normalization of the MRSI spectral data.…”

“…Alusta and coworkers used four steps to process the MRSI data before building a classification modeling that included data normalization, re-calibration of the spectra to specific peaks, weighting of the data, and re-normalization of the MRSI spectral data. The four step data processing increased the accuracy of predicting nine different brain cancer categories from 31% to 95% [141]. …”

A quantum mechanics-based approach for optimization of metabolite basis-sets. Application to quantitation of HRMAS-NMR signals

“…Warping methods were recently given in [8], see references herein. Also, in in vivo MR Spectroscopy, chemical shift recalibration based on linear interpolation was proposed [9]. In the present work, we propose a more accurate method based on quantum mechanical (QM)-simulations, thus respecting the correct fingerprints of metabolites [10,11].…”

A Novel mCAD for pediatric metabolic brain diseases incorporating DW imaging and MR spectroscopy

“…1 H MRS is the commonest spectroscopy technique allowing the concentration of different metabolites containing hydrogen nuclei to be measured and be used to identify metabolic changes (Filippi et al, ; Alusta et al, ). The diffusion of water molecules can also be investigated with MR through the use of DWI technique (Robertson & Glasier, ; Galanaud et al, ).…”