Understanding MRI from an orthopedic perspective

Chief medical editor Douglas W. Jackson, MD, speaks with radiologist William Bradley, MD, about the finer points of MRI films and terms.

Issue: October 2004

More orthopedists are reading the actual MRI films obtained on their patients. They compare their reading to the radiologist’s reading. It is important to look at the actual studies and not just the report to make sure it is an adequate study, to confirm the pathology and to be certain you agree with the written report.

To read the MRI, it is important to understand the imaging of specific tissues to be certain that the study demonstrates all of the possible information contributing to an accurate diagnosis. To speak with colleagues and our patients, it is helpful if we understand the appropriate terminology.

This interview gave me the opportunity to talk to a very knowledgeable friend and professor of radiology, William Bradley, MD.

Bill and I have spent many early morning conferences reviewing MRIs on my patients, looking at interesting findings from his files and working on research projects. I asked him to give us his usual succinct answers addressed to orthopedists to the following four questions. It is a pleasure to share them with our readers.

Douglas W. Jackson, MD
Chief Medical Editor

William Bradley, MD
Professor and Chairman
University of California at San Diego Medical Center

Douglas W. Jackson, MD: What is a T1-weighted image and what does it show?

William Bradley, MD: All tissues can be described by three fundamental MR parameters: T1, T2 and proton density. MR scans that bring out T1 contrast are defined as T1-weighted.

The primary basis for different types of contrast in MRI is the choice of two scan parameters: TR (repletion time) and TE (echo delay time). The MR technologist must set these before the patient is scanned. A T1-weighted image has a short TR and a short TE. In practice, the TR is about 500 msec and the TE is about 15 msec.

On T1-weighted images, tissues with short T1 times (like subcutaneous fat or fatty bone marrow) appear bright; tissues with long T1 times (like fluid) appear dark. Solids (like cortical bone) also appear dark. If “fat saturation” is used, fat will appear dark on a T1-weighted image.

T1-weighted images are generally considered to show the best anatomy, although they are not that sensitive to pathology. They have the best signal-to-noise per-unit time of scanning (see figure below).

This T1-weighted image has bright subcutaneous fat, yellow marrow, dark muscle and black meniscus and tendons, including the pes anserinus (arrow).	This proton density-weighted coronal knee image shows a medial meniscus tear (long arrow). Subcutaneous fat and fatty marrow are bright, although not as bright as on a T1WI. Joint effusion is relatively bright (short arrows), although not as bright as on a T2WI.
COURTESY OF WILLIAM BRADLEY

Jackson: What is a T2-weighted image and what does it show?

Bradley: T2-weighted images have a long TR (more than 2000 msec) and a long TE (more than 80 msec). Tissues with short T2s appear dark; those with long T2s are bright. Since fluid has a long T2, joint effusions and muscle or bone marrow edema appear bright. Thus T2-weighted images are the most sensitive to pathology.

Cortical bone has a very short T2; thus, it appears dark on T2-weighted images. Since the time to acquire the image is proportional to TR, these images tend to take more time than T1-weighted images. They also tend to have less signal-to-noise per-unit time of acquisition.

Over the past decade, a technique called fast spin echo has become the one most commonly used for T2-weighted images. On such images, fat appears bright. In order to see subtle bone marrow edema, fatty bone marrow must be suppressed. This can be accomplished on a high field magnet (1.5 Tesla or higher) using a fat saturation pulse (“fat sat”).

Another way to suppress fat is to use a technique called short T1 inversion recovery (STIR), which can be used at any field strength. On such images, the background signal tends to be dark, allowing the greatest contrast with any edematous process, highlighting pathology.

Fat saturated T2-weighted image shows kissing bone marrow contusions (arrows) in the lateral compartment associated with an ACL tear.

Jackson: What is a proton density-weighted image? How is it different from the other types of images?

Bradley: A proton density-weighted image is halfway between a T1- and a T2-weighted image. It has a long TR (more than 2000 msec) and a short TE (about 15 msec). Like a T2-weighted image, it is good for anatomy but it is more sensitive to pathology, although not as sensitive as a T2-weighted image. It has better signal-to-noise per-unit scan time than a T2-weighted image, which translates to higher spatial resolution or thinner slices.

In the early days of MRI using conventional spin echo, proton density-weighted images were acquired automatically with a T2-weighted image. These were known as the “long TR, double echo” sequence. On modern scanners using fast spin echo, the first echo — proton density — is no longer free and must be separately programmed, if wanted.

Jackson: What does signal-to-noise ratio mean to the clinician?

Bradley: Signal to noise is the currency of MRI. High signal to noise can be traded for faster scans, higher spatial resolution or thinner slices. Faster scans are the best way to prevent patient motion and motion artifact. High spatial resolution allows better depiction of anatomy. Thinner slices minimize the averaging of normal tissue with pathology (so-called “partial volume averaging”).

Signal to noise is increased on higher field MR scanners. For the past 20 years, “high field” has been 1.5 Tesla (30,000 times stronger than the earth’s magnetic field). Over the next decade, it is clear that the new standard for “high field” will be 3 Tesla. This will afford even higher resolution, faster, thinner slice MR imaging.

Another way to increase signal to noise is to use phased array coils. These are the radiofrequency coils that are put around the body or body part right before MR scanning. Phased array coils also enable a technique called “SENSE” (SENSitivity Encoding) or “parallel imaging” to be performed. This allows even faster scanning than would otherwise be possible.

For more information:

Hendrick RE. Image contrast and noise. In: Stark DD, Bradley WG, eds. Magnetic Resonance Imaging. 3rd ed. St Louis, Mo.: Elsevier;1999:43-69.