Do You Need a 3-Tesla MR System?

For facilities weighing whether to purchase an MR system with a 1.5-tesla (T) or 3 T magnet, the short answer is: If you're serving a neurosurgery program, the capabilities of 3 T systems should be seriously considered. For other applications, while it improves image quality and diagnostic confidence, there is little evidence that 3 T is necessary or that it contributes to improved patient outcomes.

In MR imaging, the most important property affecting image quality is the field strength. An MR image is made from the signals emitted by protons in the patient. The signal is very small, so the number of signal-emitting protons needs to be maximized. As the applied magnetic field increases, the number of protons available for imaging also increases. The net result is that doubling the field strength should double the signal-to-noise ratio and therefore the image quality—though only in theory, as we'll discuss.

The first commercial 3 T MR systems became available around 2000 and were designed predominantly for head imaging. Since then, the use of 3 T has expanded. All the major manufacturers now sell full-body systems that are equipped for the full range of MR studies. (Head-only systems are no longer available, except for research at even higher field strengths.) The number of 3 T systems sold now accounts for 40% of new MR installations. While 1.5 T remains the most common field strength, the number of units and market share have declined since 2000, despite the higher cost of a 3 T system (about $800,000, or 50%, more than a 1.5 T system) and the higher construction costs. In other words, a number of facilities believe that the greater expense of 3 T is justified.

Since the introduction of 3 T MR for routine clinical use, many studies have compared 3 T systems with systems that have lower field strengths. Most of these studies have found improved image quality in the 3 T systems (though it is rarely doubled compared to 1.5 T), allowing more lesions to be detected with increased confidence. But the clinical value of these advances is difficult to gauge, since...