Introduction

The pulse sequence grassB0 is available on all Lucas scanners to provide maps of magnetic field for use in distortion correction or other applications.  It uses 2DFT GRE acquisitions at two echo times in concatenated fashion. While this takes longer than does a dual echo acquisition, it is more accurate because effects of eddy currents are avoided. For each slice six images are collected: magnitude, real and imaginary for each of the two TEs. 

Setup

Spatial resolution and slice prescription is entered as usual.  Several options should be considered for field mapping in the “Advanced” tab (Fig. 1).  A typical protocol is in site/head/gary-B0map.

*Output Image-numbering format

If 47 is used in CV3, an output series is created as acq_series + 200, with nslc field map images numbered as 1 to nslc, and nslc magnitude images numbered nslc +1 to 2*nslc.

If 49 is used in CV3, an output series is created as acq_series + 100, with nslc magnitude images numbered as 1, 7, 13 … images, and nslc field map images numbered as 2, 8, 14, ….  This is the legacy product and is retained for existing postprocessing pipelines.

Please contact Gary at gary.glover@stanford.edu for additional questions.

Posted 2021_0421

Notes from Gary regarding Physio Monitoring during fMRI scans:

Physio Data Recording Rates

The physio data recording rates for PPG (plethysmograph), RESP and ECG are set to 100, 25 and 1000 Hz, respectively. Changing the CVs has no effect.

Calculation of the Exact Duration of the Physio Data Recording

If the CV 'phys_record_flag' = 1, then the record will start exactly 30.0 sec before the scan button is pushed, and ends within a few msec of the end of the scan. Because an external trigger is often used for starting functional scans, the duration of the physio record is a function of the scan duration + 30 sec, but also the time from when the scan button was pushed to when the scan actually started (up to 24 sec). Thus, counting backwards from the end of the physio record to find the beginning of the record in order is necessary to know the exact duration of the scan. The extra 30 sec of prepended data might be used for generating histograms that can be helpful for initiating peak-finding algorithms.

Maximum Duration of a Physio Record.

The maximum duration of a physio record, including 30 sec of prepeneded data is 25.0 minutes in rev </= 25x.

Monitoring the Physio Data Collection

Monitoring of the physio data collection can be done through the AGP terminal, accessible through mgd_term.

Contact Information

Please contact Gary at gary.glover@stanford.edu, 650-723-7577 or 650-302-6902 for additional questions.

Posted 2017_0301

Notes regarding Gary's SMS research psd ('episms'):

SNR (signal-to-noise ratio)

SMS does not intrinsically cost SNR because the kxy sampling of a given slice is not changed. Rather, multiple slices are gathered simultaneously. However, typically the TR is reduced concomitantly with the SMS factor and that *does* cost SNR because of increased T1 saturation with the shorter TR. The net result is that for an SMS = 6 (what we typically use) and a TR of 350 ms, you gain approximately 25% in CNR over the scan. It is less than the square root (6) because of the T1 effects and temporal autocorrelation (non-independent samples as the TR is shortened).

N Frames

The first N frames are part of the calibration procedure required for SMS, and they are discarded. This needs to be taken into account for a task scan. Here N = Num_calibration scans * SMS_factor (eg., 2*5 = 10).

Reconstruction

The recon is run on an offline computer and is part of a complex set up. A very large raw data file is transferred automatically to the offline computer along with some other smaller files. The nifti files are then automatically sent directly to the Dell PC dicom computer, not the MR scanner.

TR Triggers for the EEG System

The episms research psd supplies the EEG system with TR triggers, needed for post-processing.

Number of slices (TR & SMS factor selection)

The TR depends on the number of slices, the SMS factor and the spatial resolution. For 64x64 resolution, it takes approximately 60 ms to acquire a SMS group of slices. With higher resolution, it takes more time than 60 ms. For example, in 1110 ms, 18 * 3 slices can be acquired. If more or less slices are required, either the TR or the SMS factor must be modified. SMS factors up to about 6 work with the Nova 32 channel head coil - if using SMS factor of 8, some artifacts may appear due to coil geometry factor (g).

Contact Information

Please contact Gary at gary.glover@stanford.edu, 650-723-7577 or 650-302-6902 for additional questions.

Posted 2017_0103, 2017_0301

Elimination of SRbox from 3T2 & 3T3

We will be eliminating the PST SRbox from 3T2 and 3T3 scheduled for 2017_0119 as we have made the decision to no longer support the device.

If you are using the SRbox either for triggering the scanner or collecting responses, you must modify your Eprime scripts. 

I recommend making copies of relevant scripts and modifying them according to the "simple instructions" attached here (PDF).

It shouldn't take more than a few minutes to make the modifications.

A copy of the NewScanTrig.es is also available on the Eprime computer as an example of how to trigger the scanner.

With the SRbox eliminated, scan triggering will occur by writing to a serial port, and button responses will be collected by a USBbox.

Thus, the same response pads and computer mice will continue to work, and no reaction time-sensitive experiments are at risk.

At this time we will not be changing the Eprime revision, and 1.2 and 2.0 will both remain operable.

The mock scanner (0T1) will remain with a USBbox for button response input.

Contact Information

Please contact Gary at gary.glover@stanford.edu, 650-723-7577 or 650-302-6902 for additional questions.

Posted 2017_0118

Notes regarding Spiral research psds (eg., sprlio) at 3T2 & 3T3:

Post Processing

The MR scanner creates nifti files from spiral scans called 'Pfile.nii'. It first does the sprlioadd to combine in-out images, if you used the spiral-in/out sequence. If you collected physio data, it will also apply 'retroicor' and then 'rvhrcor' for denoising, and generates a second nifti file called 'Pfile.den.nii'. The nifti files will also be transferred to the Dell PC dicom computers with the other files.

High Order Shim

The popup note that the GE product software on the MR scanner posts "I will null your HO shim because you changed the slice prescription", continues to be followed by Gary's popup note "Ha! Just kidding". You do not need to worry about the little check box that asks if you want to preserve the hoshim as it some times fails when the box is checked. Gary has installed a fix that will maintain the High Order shim you have achieved after successfully running the hoshim scan.

Contact Information

Please contact Gary at gary.glover@stanford.edu, 650-723-7577 or 650-302-6902 for additional questions.

Posted 2017_0103

Notes regarding Eprime use at 3T2 & 3T3:

Connections & Triggers

Eprime scripts on the eprime computer usually starts the MR scanner through the SRBox, which is a serial port device.

Eprime can send task markers to the EEG interface if desired. The parallel port switch should be set to EEG to do so. Typically bits 0-4 are used as event markers in your eprime script. For example, bit 0 can be used for a normal tone, bit 1 for an odd ball trial. The MR scanner communicates with the EEG interface with a 10 MHz clock and scan TR triggers in the background.

The MR scanner sends physio data including PPG (finger photoplethysmograph) signals to the rtviz computer in the background. The rtviz computer running the physmon program derives the PPG triggers and sends them to the EEG interface in the background for a marker used later by the OBS algorithm for denoising of BCG (balistic cardiogram).

Contact Information

Please contact Gary at gary.glover@stanford.edu, 650-723-7577 or 650-302-6902 for additional questions.

Posted 2017_0103

With the increasing use of fMRI-EEG, Gary has set up a workstation for EEG post processing at 3T3.

This Mac computer (not the iMac) with a large screen is located in the computer room adjacent to the 3T3 scanner control room; it contains a waveform tools menu.

Please bring your EEG data from the Mac laptop used for acquiring the EEG data and upload to this Mac for post-processing.

Please create a folder with your name or your PI's name within the "Data" folder and upload there.

There is a folder on the Desktop called "Manuals" which contains all the Netstation manuals and a document called "EEG Postprocessing" that is also available at http://med.stanford.edu/lucasmri/current-users.html.

The "EEG Postprocessing" document is updated to include information regarding OBS denoising.

Please contact Gary at 650-723-7577 or 650-302-6902 for additional questions.

Thanks!