Cor STIR

[Barrymagnetic]

How can I reduce the scan time when imaging with a STIR protocol? I'm using a GE scanner by the way.

I've been shown to increase the TR until the max # of slices equals the total # of slices. This will reduce the scan time into an acceptable range usually. The problem arises when the TR approaches 6000 ms or more. What other variables can be adjusted to keep the TR lower than that? It has been suggested to set TE in the middle of the min/max settings. Reduce number of slices? Reduce slice thickness? Adjust ETL? Other ideas?

Thank you for any and all help!

[AndrewBworth]

It all depends on what you are trying to image, the quality of the images you are currently getting, and how much time you need to save. There is no "one size fits all" approach to reducing scan time.

First of all - your TR for a STIR does not necessarily become "too long" at any point. It is just important that it is long enough - where I work a general rule of thumb is at least 3500 ms. Some techs I work with get scared to go over 6000, but that actually does not matter. What does matter though is if there is "dead time" during the TR, that is generally just a waste (although you might be forced to occasionally run a longer TR because of SAR).

There are methods to reduce scan time that can either increase or decrease SNR. A method that increases SNR is to reduce the number of phase points in the matrix. You should do this carefully, as the radiologist has likely specified the resolution as what is needed to identify pathology. Increasing the turbo factor - this supposedly increases noise, yet I have never noticed a noise increase - this can lower scan time. This method does increase SARS. Lastly, the overall resolution can be decreased - sometimes this is actually a very good option, but if reduced so low that details are lost you could be in trouble. I would say these methods wouldn't get used much by a novice technologist.

Methods that reduce scan time and decrease signal to noise are sometimes more acceptable. Use of partial fourier transform, SENSE methods, and GRAPPA/ARC methods can really cut down time. You can also see if a protocol is using more oversampling than is necessary as well.

Personally - I am finding that really the best ways to reduce scan times are "old fashioned" and weren't really taught to me during MRI school. For example - generally the smallest coil available that covers the anatomy you want to image should be used - this minimizes noise, allowing for time saving strategies. Another very good strategy is patient coaching/education/restraints. For example, instructing patients to take shallow breaths on some exams dramatically reduces motion - leading to fewer repeat scans and less reliance on motion control techniques. Scans where the machine observes respiration can sometimes take 10 minutes if the patient breathes erratically and moves about during the scan - patient education before the scan begins can also dramatically slash this time sometimes down to just a few minutes. Lastly, don't be afraid to use a well placed sandbag or two - just make sure you aren't making the patient uncomfortable.

(04-28-2015 10:35 PM)Barrymagnetic Wrote:  How can I reduce the scan time when imaging with a STIR protocol? I'm using a GE scanner by the way.

I've been shown to increase the TR until the max # of slices equals the total # of slices. This will reduce the scan time into an acceptable range usually. The problem arises when the TR approaches 6000 ms or more. What other variables can be adjusted to keep the TR lower than that? It has been suggested to set TE in the middle of the min/max settings. Reduce number of slices? Reduce slice thickness? Adjust ETL? Other ideas?

Thank you for any and all help!