Friday, November 17, 2023, 8:00 AM – 5:45 PM EST
Columbia University in the City of New York
9th floor, Jerome L. Greene Science Center
3227 Broadway, New York, NY 10027
This year's meeting is sponsored by MISSIONS - The Merck Imaging Network
Registration is now open: Sign Up
Downloadable Speaker List: PDF
Opening
8:00 AM - Check-in & Breakfast
9th Floor, Jerome L. Green Science Center
8:45 AM - Welcome & Agenda
Christoph Juchem, Ph.D., Columbia University
Session I
9:00 AM - Fulfilling a pressing need for standardization in MRS research: MRS-BIDS and NIfTI-MRS
Mark Mikkelsen, Ph.D., Weill Cornell Medicine
In vivo MRS is an increasingly utilized biomedical imaging modality for the noninvasive detection of molecules in the human body. However, it lags behind other MR techniques in terms of standardization of methodological approaches. Namely, there is enduring heterogeneity in data acquisition, analysis, reporting, and sharing methods. In this talk, I will present recent initiatives undertaken by the MRS community to move toward standardizing our science. I will briefly introduce the Brain Imaging Data Structure (BIDS), a standard for organizing neuroimaging data, and then speak on extending this standard to include MRS data and metadata, denoted MRS-BIDS. This extension incorporates the new NIfTI-MRS data format, which I will also describe. These initiatives are just two that will aid in the further standardization of MRS research.
9:30 AM - Deep Learning-based MRS Reconstruction and Denoising with Artificial Fourier Transform Network (AFT-Net)
Yanting Yang, M.S., Columbia University
A complex-valued deep learning framework - artificial Fourier transform network (AFT-Net), which directly processes the complex-value raw data in the sensor domain, is proposed to reconstruct spectra from FIDs and denoise the spectra in parallel. We trained the model on the simulated MEGA-PRESS dataset and the in-vivo Big GABA dataset. An evaluation of different acceleration rates was performed on the in-vivo dataset. AFT-Net demonstrated the ability to reconstruct the data with high fidelity and significantly accelerate acquisition. The proposed AFT-Net is an efficient and accurate approach for MRS spectra reconstruction from raw data.
10:00 AM - MRS of brain glutamate metabolism as a translational biomarker for drug development
Corey Miller, Ph.D., Merck & Co.
Merck is currently pursuing novel symptomatic therapies for Alzheimer’s Disease aimed at improving the associated cognitive decline. One mechanism that shows promise is the pharmacological increase of excitatory glutamatergic neurotransmission. However, while biomarkers are becoming an increasingly important part of the drug development process, a paucity of objective biomarkers exist for this area, as well as for most neurological disorders. In this work we present MRS measurements of glutamate metabolism across multiple species, using 13C-glucose as a tracer, as a proxy for glutamatergic neurotransmission to guide the development of novel compounds and to inform the design of clinical efficacy trials.
10:30 AM - Coffee Break
Session II
11:00 AM - Chemical shift displacement error: prospective and retrospective corrections for 3T PRESS MRS data acquired on GE scanners
Diana Rotaru, Ph.D., Columbia University
Chemical shift displacement error (CSDE) is a well-known problem in the 1H-MRS field, yet still underestimated. Retrospective and prospective CSDE corrections are proposed for data previously acquired and to be acquired on 3T General Electric MR scanners. Extreme CSDE was observed on GE MRS data due to use of standard PRESS GE product sequence and default parameters (i.e. the 2.7 ppm center frequency for the water-unsuppressed acquisition and for the OVS-VSS saturation bands). Results show improved concentration estimates for the major metabolites found in the human brain.
11:30 AM - synMARSS, a novel platform for simulating in vivo synthetic spectra
Karl Landheer, Ph.D., Regeneron Pharmaceuticals
synMARSS will be discussed, which is a novel flexible platform that can simulate in vivo synthetic spectra for any MRS sequence and tissue type. synMARSS was specifically designed for applications such as the generation of high-quality training data for machine learning, to optimize sequence parameters or to investigate the effect of modeling assumptions. synMARSS was used to investigate the effects on quantification of the typical assumption of treating14N heteronuclear coupling as weak homonuclear coupling.
12:00 PM - Panel Discussion, Artificial Intelligence (AI) for MR Spectroscopy
Moderator: Christoph Juchem, Ph.D., Columbia University
Panelist: Jia Guo, Ph.D., Columbia University
Panelist: Maggie Fung, GE Healthcare
Panelist: Sachin Jambawalikar, Ph.D., Columbia University
Panelist: Lawrence Kegeles, M.D., Ph.D., New York State Psychiatric Institute
12:30 PM - Lunch Break
Session III
1:30 PM - The Past, Present, and Future of Functional Magnetic Resonance Spectroscopy (fMRS)
Alex Lin, Ph.D., Harvard Medical School
While early MR spectroscopy studies utilize a functional aspect of tissue metabolism, most MRS data has been collected in the resting state. Recent advances in hardware and software have renewed interest in collecting MRS in response to stimuli, providing dynamic measures of metabolism with greater insight into cognitive and clinical physiology. This presentation will touch upon the early fMRS studies, describe current studies utilizing ultrahigh field MR, and examine the future developments needed to overcome today’s challenges, as discussed at the first fMRS symposium recently held at Bangor University.
2:00 PM - MC-ECLIPSE for B0 shimming and arbitrary ROI shaped OVS for applications in human brain 1H-MRSI
Chathura Kumaragamage, Ph.D., Yale University School of Medicine
The utility of ECLIPSE (a pulsed second-order gradient insert with Z2 and X2Y2 fields) was previously demonstrated, that allows unparalleled extracranial lipid suppression for applications in human brain 1H-MRSI. ECLIPSE offers excellent axial slice coverage for elliptical brain shapes, however, is less effective for asymmetrical head shapes. We recently constructed a 54-channel multi-coil array in combination with ECLIPSE (MC-ECLIPSE) to achieve improved axial slice coverage for OVS, in addition to B0 shimming with the MC array. Here we demonstrate >92% axial slice coverage over challenging head shapes, in addition to improved B0 shimming relative to SH2 shims for 1H-MRSI.
2:30 PM - First-in-human Evaluation of [1-13C] Pyruvate in D2O for Hyperpolarized MRI Promises More Accurate Cancer Diagnosis
Guannan Zhang, Ph.D., Memorial Sloan Kettering Cancer Center
Dissolution dynamic nuclear polarization is a technique that dramatically improves the signal to noise ratio (SNR) of magnetic resonance imaging (MRI), making it possible to image metabolism in patients by administering hyperpolarized [1-13C] pyruvate for more accurate cancer diagnosis. However, the observable time window and the achievable image SNR are primarily governed by the relaxation time (T1) of the nucleus within the molecules. To address this limitation, we increased T1 of hyperpolarized [1-13C] pyruvate by using D2O as solvent. Here, we demonstrate the safety and feasibility of the method in healthy volunteers and its application to a pancreatic cancer patient.
3:00 PM - NMR-based Metabolomics
Leo Cheng, Ph.D., Harvard Medical School
To account for holistic and real-time metabolic activities, NMR-based metabolomics is considered to be the most accurate reflection of overall clinical states. Accurate measurements human metabolomics can assist clinical evaluations, from disease detection and diagnosis to patient prognostication and treatment assessment. To facilitate clinical implementations, we developed high-resolution magic angle spinning (HRMAS) NMR methodology, for better characterization of cellular metabolomics for human diseases, and to guide the development of in vivo NMR-based metabolomic imaging. The HRMAS method also allows us to establish disease metabolomic biomarkers measurable from mere microliters of human blood and follow real-time metabolic activities with live cells.
3:30 PM - Coffee Break
Session IV
4:00 PM - Integrating Structural, Functional, and Biochemical Brain Imaging Data with MRShiny Brain - An Interactive Web Application
Jessica Archibald, Ph.D., Weill Cornell Medicine
The utilization of structural, functional, and biochemical data from the human brain has grown in addressing inquiries related to neurodegenerative and neuropsychiatric conditions. However, the normal variability within these measures has not been systematically reported. In this study, a live database comprising these outcome measures in a healthy population was established to potentially serve as a comparative reference for neurodegenerative and neuropsychiatric research. This study provides a comprehensive overview of the methodology and data acquisition process, aiming to facilitate potential future utilization of the collected data.
4:30 PM - Latest Spectroscopy Features in PV360
Saaussan Madi, Ph.D., Bruker BioSpin
Bruker products enable researchers to take advantage of the most recent hardware and software upgrades for preclinical spectroscopic applications. We introduced MEGA PRESS preparation to the PRESS method. MEGA PRESS is typically used for Gamma Aminobutyric acid GABA spectroscopy applications. We also introduced a new “Spectroscopy” visualization tool for spectroscopy and spectroscopic imaging data.
5:00 PM - Spectral registration for retrospective B0 drift correction of rosette MRSI data at 3 Tesla
Jamie Near, Ph.D., University of Toronto
Correction of B0 field drift is a routine step in single-voxel magnetic resonance spectroscopy (MRS) data preprocessing. However, drift correction in MR spectroscopic imaging (MRSI) data is more challenging due to phase encoding steps between repetitions. Here, we implemented a rapid-high-resolution MRSI sequence using a rosette k-space trajectory which samples the centre of k-space in every TR. Using the central k-space FIDs, we employed spectral registration to estimate and correct B0 field drifts across TRs. The magnitude of observed field drifts was substantial, and we show that correction using the proposed method yields meaningful improvements in linewidth and signal-to-noise ratio.