Using optimized, asymmetric radiofrequency (RF) pulses for slice selection, the authors demonstrate that stimulated echo acquisition mode (STEAM) localization with ultra‐short echo time (1 ms) is possible. Water suppression was designed to minimize sensitivity to B₁ inhomogeneity using a combination of 7 variable power RF pulses with optimized relaxation delays (VAPOR). Residual water signal was well below the level of most observable metabolites. Contamination by the signals arising from outside the volume of interest was minimized by outer volume saturation using a series of hyperbolic secant RF pulses, resulting in a sharp volume definition. In conjunction with FASTMAP shimming (Gruetter Magn Reson Med 1993;29:804–811), the short echo time of 1 msec resulted in highly resolved in vivo ¹H nuclear magnetic resonance spectra. In rat brain the water linewidths of 11–13 Hz and metabolite singlet linewidths of 8–10 Hz were measured in 65 μl volumes. Very broad intense signals (Δν_1/2 > 1 kHz), as expected from membranes, for example, were not observed, suggesting that their proton T₂ are well below 1 msec. The entire chemical shift range of ¹H spectrum was observable, including resolved resonances from alanine, aspartate, choline group, creatine, GABA, glucose, glutamate, glutamine, myo‐inositol, lactate, N‐acetylaspartate, N‐acetylaspartylglutamate, phosphocreatine, and taurine. At 9.4 T, peaks close to the water were observed, including the H‐1 of α‐D‐glucose at 5.23 ppm and a tentative H‐1 resonance of glycogen at 5.35 ppm. Magn Reson Med 41:649–656, 1999.

Clinical studies consistently report structural impairments (i.e.: ventricular enlargement, decreased volume of anterior cingulate cortex or hippocampus) and functional abnormalities including changes in regional cerebral blood flow in individuals suffering from schizophrenia, which can be evaluated by magnetic resonance imaging (MRI) techniques. The aim of this study was to assess cerebral blood perfusion in several schizophrenia-related brain regions using Arterial Spin Labelling MRI (ASL MRI, 9.4 T Bruker BioSpec 94/30USR scanner) in rats. In this study, prenatal exposure to methylazoxymethanol acetate (MAM, 22 mg/kg) at gestational day (GD) 17 and the perinatal treatment with Δ-9-tetrahydrocannabinol (THC, 5 mg/kg) from GD15 to postnatal day 9 elicited behavioral deficits consistent with schizophrenia-like phenotype, which is in agreement with the neurodevelopmental hypothesis of schizophrenia. In MAM exposed rats a significant enlargement of lateral ventricles and perfusion changes (i.e.: increased blood perfusion in the circle of Willis and sensorimotor cortex and decreased perfusion in hippocampus) were detected. On the other hand, the THC perinatally exposed rats did not show differences in the cerebral blood perfusion in any region of interest. These results suggest that although both pre/perinatal insults showed some of the schizophrenia-like deficits, these are not strictly related to distinct hemodynamic features.

While the recommended analysis method for magnetic resonance spectroscopy data is linear combination model (LCM) fitting, the supervised deep learning (DL) approach for quantification of MR spectroscopy (MRS) and MR spectroscopic imaging (MRSI) data recently showed encouraging results; however, supervised learning requires ground truth fitted spectra, which is not practical. Moreover, this work investigates the feasibility and efficiency of the LCM-based self-supervised DL method for the analysis of MRS data. We present a novel DL-based method for the quantification of relative metabolite concentrations, using quantum-mechanics simulated metabolite responses and neural networks. We trained, validated, and evaluated the proposed networks with simulated and publicly accessible in-vivo human brain MRS data and compared the performance with traditional methods. A novel adaptive macromolecule fitting algorithm is included. We investigated the performance of the proposed methods in a Monte Carlo (MC) study. The validation using low-SNR simulated data demonstrated that the proposed methods could perform quantification comparably to other methods. The applicability of the proposed method for the quantification of in-vivo MRS data was demonstrated. Our proposed networks have the potential to reduce computation time significantly. The proposed model-constrained deep neural networks trained in a self-supervised manner can offer fast and efficient quantification of MRS and MRSI data. Our proposed method has the potential to facilitate clinical practice by enabling faster processing of large datasets such as high-resolution MRSI datasets, which may have thousands of spectra.