Patient Examination

Dynamic contrast enhanced MR can be satisfactorily performed on the majority of currently available clinical scanners. Because of the requirements for high temporal resolution there will be restrictions on spatial resolution and spatial coverage dependent on gradient performance. However, it should be stressed that useful clinical information can be obtained in many applications using even a single slice of dynamic data. The majority of currently available clinical scanners will comfortably allow mul-tislice acquisitions to be performed with adequate temporal resolution for most analysis techniques.

Contrast administration is performed through a peripheral vein. A large antecubital vein is commonly employed and the injections can be given through a small cannula which should be inserted and secured in place prior to the investigation. The injection technique is of considerable importance. Most dynamic imaging methodologies use a bolus injection of contrast and it is important that this be administered in a consistent manner. Most centres now use an auto mated pressure injection system to ensure reproducibility. Protocols for contrast administration vary depending on the technique in use. Typically however, a single dose of contrast (0.1 mmol/Kg) of a standard gadolinium chelate will be administered at a rate in the region of 4 ml/s. Some centres prefer to vary the injection rate so that the overall period of contrast administration is kept constant rather than having a constant injection rate of different volumes in different patients. It is important that the contrast bolus remain coherent in its passage through the body and in order to achieve this a chaser injection of normal saline is given immediately after the contrast. The chaser injection must be given at the same flow rate as the contrast and must be of adequate volume to empty the draining veins, typically 20-30 ml, so that the contrast passes into the systemic circulation as a coherent bolus. The venous injection should be placed into the right arm if possible since variations in venous anatomy can lead to significant jugular

Cp 25

20 40 60 80 100 120

Fig. 1.3. Showing the change in contrast concentration over time within the middle cerebral artery following a bolus injection of contrast media into a peripheral vein. The two graphs represent different patients and show clear differences in the spread of the bolus despite the use of the same injection technique and contrast dose reflux on the left side which can impair the coherence of the contrast bolus (Fig. 1.3).

The imaging of the dynamic sequence is usually performed following initial anatomical and localisation scans. Localisation is of prime importance particularly if hardware restrictions limit the number of slices or the spatial resolution of the images that can be obtained. There are a number of important considerations concerning slice localisation. Clearly it is essential that the pathology is included in the slice but also, for most analysis techniques it is important to include an appropriate large blood vessel. This allows the measurement of the contrast concentration changes in the plasma over time, which is commonly referred to as the arterial (AIF; Fig. 1.3) or vascular input function (VIF). This is used in many analysis techniques to represent the contrast changes occurring in the blood vessels within the tumour and to allow calculation of the contrast concentration gradient between blood and the tumour extravascu-lar extracellular space. A number of other technical complications will be encountered depending on the choice of acquisition sequence and anatomical location. Problems associated with respiratory and other physiological motion and with the presence of inflow artefacts distorting the dynamic contrast signal in blood vessels must be considered. These will be discussed in more detail in Chap. 5. If it is intended to use a simple subjective or semi-quantitative analysis of enhancement curve then the dynamic image series provides adequate data for this approach. If it is intended to use a pharmacokinetic analysis then it is necessary to calculate contrast concentration in each image in the dynamic series. Unfortunately the relationship between contrast concentration and signal intensity is non linear (see Chaps. 5 and 6) and will be affected by the underlying native T1 of the tissues. For pharmacokinetic analyses it is therefore necessary to add additional imaging sequences to the investigation before the dynamic run is performed. These sequences are designed to allow calculation of quantitative T1 maps to enable subsequent calculation of contrast concentration. The approaches taken for T1 mapping and the imaging methods employed are also described in detail in Chap. 6 (Figs. 1.4, 1.5).

The dynamic sequence is performed following preliminary anatomical imaging and T1 mapping. The choice of image acquisition technique, injection rate and temporal resolution will be entirely dependent on the analysis method to be employed. This in turn will be chosen to optimise the amount of biologically relevant information that can be extracted from the data which will depend on the organ system

Fig. 1.4. A series of dynamic MR images (top) showing contrast enhancement and passage of the contrast agent into the interstitial tissues using a 3D T1-weighted gradient echo acquisition. Illustrated images are spaced approximately 10 s apart. The lower row of images shows the calculated concentration of contrast agent which is derived from the images in the top row and which can be used as the basis for pharmacokinetic analysis of enhancement patterns

Fig. 1.4. A series of dynamic MR images (top) showing contrast enhancement and passage of the contrast agent into the interstitial tissues using a 3D T1-weighted gradient echo acquisition. Illustrated images are spaced approximately 10 s apart. The lower row of images shows the calculated concentration of contrast agent which is derived from the images in the top row and which can be used as the basis for pharmacokinetic analysis of enhancement patterns

Fig. 1.5. Parametric images from a dynamic susceptibility contrast enhanced study in a patient with a grade four glioma. The left-hand image shows a map of the baseline relaxivity (R10) with areas of significant prolongation of T1 (blue) in the location of peri-tumoral oedema. The centre image shows a map of Ktrans calculated using a simple two compartment model. Note the presence of elevated values in the region of normal cerebral arteries. The right-hand image shows the distribution of ve ( the size of the extravascular extracellular space)

Fig. 1.5. Parametric images from a dynamic susceptibility contrast enhanced study in a patient with a grade four glioma. The left-hand image shows a map of the baseline relaxivity (R10) with areas of significant prolongation of T1 (blue) in the location of peri-tumoral oedema. The centre image shows a map of Ktrans calculated using a simple two compartment model. Note the presence of elevated values in the region of normal cerebral arteries. The right-hand image shows the distribution of ve ( the size of the extravascular extracellular space)

and the pathology being studied. At one extreme, measurements of blood flow in the brain require a temporal resolution in the region of 2 s or less in order to adequately demonstrate the first passage of the contrast bolus through the cerebral vasculature (Barbier 2001) and lower temporal resolutions will introduce progressively greater errors into the calculated parameters. Some Tl-weighted dynamic techniques demand accurate measurement of the arterial input function and a temporal resolution of at least 5 s is necessary for these techniques. Where these high temporal resolutions cannot be achieved satisfactorily then the analysis method can be compromised to some extent and the use of surrogate or averaged arterial input functions may still allow meaningful pharmacokinetic analysis to be performed (Tofts 1997; Tofts et al. 1999). As an extreme example, the temporal resolution may be reduced to a period of minutes as in the case of multi-phasic imaging of the liver where the dynamic sequence consists only of one pre-contrast image and 3-4 post-contrast images performed over a period of several minutes (Bartolozzi et al. 1999). Clearly pharmacokinetic analysis of this data is impossible but nonetheless it provides extremely valuable clinical diagnostic data for showing variations in temporal enhancement patterns.

Because of the complexity of these issues the design of dynamic contrast enhanced MRI protocols can seem daunting and is indeed complex. The design process must begin with clear identification of the biological parameters available from dynamic contrast enhanced MRI which are likely to be of clinical or research value. This will in turn govern the choice of analysis techniques, which will define the characteristics necessary within the dataset and the design of the image acquisition protocol. It is important that this design process is followed appropriately and it must be stressed that there is no recipe for dynamic contrast enhanced MRI that works in all clinical cases or all applications.

In order to try and provide an understanding of these mechanisms and the problems associated with them this book will provide detailed overviews of the potential acquisition techniques and analysis methods that are currently in use or being explored by groups working with dynamic contrast enhanced MRI in oncological applications. The early chapters will deal with the technical aspects of acquisition and analysis and the later chapters will focus on the clinical benefits of dynamic contrast enhanced MRI and will review in detail the methods that have been applied in specific clinical areas.

0 0

Post a comment