Magnetic Resonance - Technology Information Portal Welcome to MRI Technology
Info
  Sheets

Out-
      side
 

MRI is trending to low field magnets :
reduced costs will lead to this change 
AI will close the gap to high field 
only in remote areas 
is only temporary 
never 




 
MRI Forum
'Image'
SEARCH FORUM FOR   
 
Result: Searchterm 'Image' found in 129 messages
Result Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 [14] 15 16 17 18 19 20 21 22 23 24 25 26 
More Results: Database  (454)  News Service  (255)  Resources  (73)  
Forum Overview
 bottom
Aaliah Thomas

Mon. 23 May.11,
07:41

[Reply (1 of 3) to:
'LCD Display Monitors'
started by: 'robert popilock'
on Mon. 23 May.11]


 
  Category: 
Safety

 
LCD Display Monitors
Hey! The role of an LCD display is only to draw a picture of the signals emitted from different body tissues. Hence the size of the display doesn’t put any effect as such. But the bigger the display is, the clearer will be the image. So it’s better to go for 32” LCD display with full HD resolution for clear picture. Yes, IEC JTC1 is developing a governance standard aimed at helping organizations govern information security, specifically the standard of deliverables.
 
 

Winners never quit and quitters never win -Anon.
 View the whole thread
Steven Ford

Thu. 17 Mar.11,
03:16

[Reply (2 of 5) to:
'Building 3d Volumes from MRI DICOM'
started by: 'Robert Patten'
on Thu. 3 Mar.11]


 
  Category: 
General

 
Building 3d Volumes from MRI DICOM
In almost all cases, MRI images have a slice thickness far greater than their in-plane resolution, making multiplanar reconstruction meaningless. Also, most MRI images have a gap between the slices, which also renders multiplanar meaningless and impossible (unless you're happy with black bars where the slice gaps exist).

You can look at your images and the slice thickness is on the graphics. the slice gap is usually not displayed, but if you look at the slice location displayed on adjacent slices, and compare to the slice thickness, you can easily compute the slice gap.

The in-place resolution is (approx) the field of view divided by the acquisition matrix, both of which are also printed on the image somewhere.

By the way, if you have the dicom (.dcm) files, there is lots of data that's 'hidden' that you can access with a full featured dicom file editor.
 View the whole thread
Steven Ford

Mon. 7 Mar.11,
16:07

[Reply (6 of 8) to:
'6-1.5T MAGNETS, DIFFERING GRADIENTS'
started by: 'Elise Gough'
on Wed. 23 Feb.11]


 
  Category: 
Applications and Examinations

 
6-1.5T MAGNETS, DIFFERING GRADIENTS
We maintain a lot of magnets. The leading cause of image quality problems is applications related. Nobody can possibly know all the nuances of pulse sequences by various vendors, software levels, etc. I've heard many times that a certain machine is no good, when in fact the sequences are inefficiently set up. Look there first.

Usually the techs welcome good training. If they don't want to be trained, then you have a problem, but it sounds like this issue is caused by other factors.

If there are differences in baseline quality between the machines, then compensate for that by other means. Signal starvation is easily remedied; keep the quality as consistent as you can and let time be the variable, if it comes down to that.




 View the whole thread
Elise Gough

Mon. 7 Mar.11,
15:56

[Reply (5 of 8) to:
'6-1.5T MAGNETS, DIFFERING GRADIENTS'
started by: 'Elise Gough'
on Wed. 23 Feb.11]


 
  Category: 
Applications and Examinations

 
6-1.5T MAGNETS, DIFFERING GRADIENTS
Thanks so much for the input. NOBODY in my group seems to have a handle on why images vary in quality from 8 different magnets. I'm being pressed to take position of imaging specialist and thru my own research we have gradient strengths from 50 slew rate up to 120. Radiologists complain neuro stinks on the 50 slew rate 1.5T magnet and it is the most pounded on magnet in our group doing hospital inpatients. So many variables including tech expertise.
 
 

Elise Gough RT(R)(CT)(MR)
 View the whole thread
Elise Gough

Fri. 4 Mar.11,
13:56

[Reply (2 of 8) to:
'6-1.5T MAGNETS, DIFFERING GRADIENTS'
started by: 'Elise Gough'
on Wed. 23 Feb.11]


 
  Category: 
Applications and Examinations

 
6-1.5T MAGNETS, DIFFERING GRADIENTS
Thank you for your response. Our group is strict about FOV, thickness/gap parameters being the same across all magnets. One magnet has slew rate of 77, FRFSE T2, TE 85,TR 3500 Classic Fat supressed images which are pristine, vs magnets of 120 slew rate same parameters look signal starved. So gradient strength makes a speed difference, not an image quality difference?
 
 

Elise Gough RT(R)(CT)(MR)
 View the whole thread

Result Pages : 1 2 3 4 5 6 7 8 9 10 11 12 13 [14] 15 16 17 18 19 20 21 22 23 24 25 26 
 top
 
Share This Page
FacebookTwitterLinkedIn

MR-TIP    
Community   
User
Pass
Forgot your UserID/Password ?    



Look
      Ups






MR-TIP.com uses cookies! By browsing MR-TIP.com, you agree to our use of cookies.

Magnetic Resonance - Technology Information Portal
Member of SoftWays' Medical Imaging Group - MR-TIP • Radiology-TIP • Medical-Ultrasound-Imaging • 
Copyright © 2003 - 2024 SoftWays. All rights reserved. [ 1 December 2024]
Terms of Use | Privacy Policy | Advertising
 [last update: 2024-02-26 03:41:00]