by Anthony Lasala
Magnetic Resonance Imaging is one of the most useful imaging modalities available in healthcare. A complex relationship between a high strength magnet, amplifiers, radiofrequency pulses, gradients, and a person’s hydrogen protons all form diagnostic imaging of internal structures. The catalyst in this multifaceted relationship is signal. In order to obtain a diagnostic image, or in fact any image, is to acquire signal. The signal is then put through rigorous equations in order to provide the technologist with an onscreen image. Different tissues offer different signal; thus, one is able to differentiate between bodily structures based on contrasting grayscales. This is incredibly important when trying to view and ultimately diagnose pathology within the human body.
There are two aspects within the magnetic field used to create MR imaging: signal and noise. Therefore, signal-to-noise ratio, or SNR, is an extremely crucial parameter. SNR can be defined as the ratio of the amplitude of signal received by the coil to the amplitude of the noise. More simply, SNR is the amount of signal divided by the amount of noise. The higher this ratio, the better the chance of coming out with diagnostic images. Why diagnostic images? The answer is in the question. A radiologist needs the images to be just that, diagnostic, in order to make a diagnosis of an observed pathology. MRI is an imaging modality with many parameters. These parameters can both positively and negatively affect relative signal-to-noise ratio. For example, a spin echo sequence, as opposed to a gradient echo sequence, automatically offers increased SNR. There are ways, however, to alter SNR during every type of sequence. Parameters such as number of signal averages, slice thickness, and field of view can all be increased in order to accordingly increase signal-to-noise. On the other hand, parameters such as bandwidth, TE and the matrix can all be decreased in order to increase SNR. It is important for a technologist to understand how to adjust certain parameters to yield certain outcomes. Similarly, it is crucial to realize that MRI deals with parameter relationships and increasing SNR may cause another parameter to unwantedly increase or decrease.
Sacrificing signal-to-noise ratio can result in a bevy of issues for a technologist. Most importantly, the radiologist will not be able to read the images. A lot is riding on an accurate readout of these images for the patient. A false diagnosis could cause a patient to be subjected to unnecessary or faulty treatment. Moreover, he or she may not receive any treatment. This is considered a failure in the healthcare setting. Patients who do not receive proper and timely treatment will only see his or her condition deteriorate, perhaps even to the point of demise. Thankfully, this is a rarity. Poor SNR usually denotes that the scanning sequence must be performed again. Although this may not seem like a big deal, it is. In MRI, time is of the essence. When a four or five minute sequence comes out with poor signal, it is a major setback. Patients often find themselves uncomfortable and agitated while being scanned, which is very easy to understand. For this reason, it is important to attain good signal for all scanning sequences in order to avoid having to repeat sequences.
While signal-to-noise ratio may be a particularly important parameter, it is not the most important. Signal does, though, play a large role in the most important factor in magnetic resonance imaging – contrast-to-noise ratio. Contrast-to-noise ratio, or CNR, can be simply stated as the difference in the SNR between two adjacent areas. Why is this relevant? In order to differentiate tissues, the most important aspect of MRI, there must be good contrast-to-noise ratio. This is how pathology can be visualized, as opposed to healthy tissue. At times, CNR is not always ideal and it may be tough to truly distinguish pathology from normal tissue. When this is the case, there are certain methods that can be employed. The most common method, if the patient is deemed healthy enough, is to administer a contrast agent. This agent is usually gadolinium-based and alters the relaxation times of tissues as to help segregate different tissue types. Another method would be to utilize different types of sequences. For example, T2 weighted sequences, magnetization transfer (MT) sequences, short TI inversion recover (STIR) sequences, and fluid attenuated inversion recovery (FLAIR) sequences can all be employed to help boost CNR. If these methods do not prove effective, a technologist can apply chemical/spectral presaturation to certain tissues.
Good CNR is critical in order for the radiologists to make a diagnosis, even more critical than SNR. Contrast-to-noise ratio is the main way a radiologist discerns whether a tissue is healthy or unhealthy. A false diagnosis in this case is perilous. The radiologist must be given the best chance at coming to a precise conclusion. Similar to SNR, when a technologist realizes he or she has images with poor CNR, the scan must be repeated. This is time consuming and frustrating for the patient, but it is crucial for that patient’s health. Every patient is different and brings with them a different scanning scenario. Therefore, it is vital for a technologist to understand how to both effectively and swiftly acquire images with good contrast-to-noise ratio.
Signal-to-noise and contrast-to-noise are both tremendously prevalent ratios within the magnetic resonance imaging world. It is in the hands of the technologist to acquire the best possible ratio in both. Time, money, and, more importantly, the health of the patient hang in the balance. The idea of SNR and CNR is an easy one to understand. However, the process of achieving proper ratios is not so simple. A technologist must pride themselves in his or her work and submerge themselves into the vast knowledge of SNR and CNR.
McRobbie, D. W., Moore, E. A., Graves, M. J., & Prince, M. R. (2007). Mri: From picture to proton. (Second ed.). Cambridge, UK: Cambridge University Press.
Westbrook, C. (2011). Handbook of mri technique . (Third ed.). Cambridge, UK: Wiley-Blackwell.