CHAPTER SIX 6 CLINICAL APPLICATIONS

MRI CLINICAL APPLICATIONS

The clinical use of MRI has increased dramatically in the past decade and scanners may now be found in all major UK hospitals and a growing number of regional ones. The installation of a new scanner is a major capital expense but the provision of government assistance and the growing number of private finance initiatives (PFI) has brought the acquisition of an MRI scanner within the reach of most NHS trusts. As technology improves and more research is carried out, the range of clinical applications of MRI has expanded. This chapter presents some clinical investigations that I observed during my MRI training at Ninewells Hospital. 1 Introduction

In clinical MRI, contrast in the image is governed by the fundamental tissue relaxation parameters T1 and T2, and the corresponding proton density. Varying these parameters on a case-by-case basis provides good soft tissue contrast and facilitates reliable diagnosis. In the clinical setting, T1 weighted images generally highlight the anatomy and T2 weighted images generally highlight pathology 1. Contrast allows for the distinction among soft tissue types and it is the result of differences in tissue signal intensities. There are many factors that affect image contrast and these are highlighted in Table 1 below 2. In general, T1 weighted images are characterised by bright fat and dark free water and T2 weighted images by dark fat and bright free water. Air produces no MR signal and will appear dark in both cases. 2 Clinical Investigations

During my training I observed various MRI procedures performed by radiographers. These scans were carried out on both the Siemens Impact (1.0T) and Symphony (1.5T) scanners during normal scanning hours. The following sections describe some examples of MRI examinations I observed during my MRI training along with the appropriate MRI Physics sequences. I have presented case studies covering musculoskeletal MRI (knee), MR angiography (renal arteries), Neuro MRI (brain), Neuro MRI (spinal cord), body MRI (breast) and finally body MRI (liver). These studies used basic T1W, T2W and PDW spin echo and turbo spin echo sequences as well as inversion recovery sequences (STIR and FLAIR) and gradient echo sequences (predominantly FLASH and DESS). Many of these sequences involved the use of contrast agents or fat suppression techniques to improve the contrast and dynamic range of the resulting images. These methods are all described where appropriate, and the reason for use in each case is discussed. 1 Musculoskeletal MRI (Knee)

A 40-year-old male patient presented with severe knee pain and trauma. The first sequence used was a fat suppressed dual-echo turbo spin echo, with a TR 4000ms and two TEs of 22ms for PDW and 90 ms for T2W. The advantage of the dual echo sequence is that proton density weighted images and T2 weighted images can be acquired from the same RF excitation pulse, giving twice the amount of image information for the same scan time. In addition, the use of a turbo factor (in this case 5) reduces the scan time further and allows the data to be acquired in just a few minutes despite the long TR. Fat suppression ensures that fluid signal is clearly visible, and will highlight pathology such as bone bruising. In this study, the extremity (knee) coil was used in transmit and receive mode, with the patient in a supine position and feet first in the Siemens Symphony scanner. Figure 6-1
46

CHAPTER SIX

MRI CLINICAL APPLICATIONS

below shows a coronal fat-suppressed T2W image from the dual echo data. The high signal area to the left is indicative of severe joint effusion, which appears to be associated with extensive damage to the lateral meniscus.

Figure 6- 1: 2D Coronal Fat Suppressed T2W Knee Image from Dual Echo TSE Sequence (TR=4000ms, TE=22/90ms, FA=180, SL=3mm, NoS=19, FoV=179cm, RFoV=7/8, Matrix=225x512, Scan Time=64, AC=2, Turbo=5, Sequence Name=tse5_22rrb130_90rrb130, FATSAT) Figure 6-2 below shows sagittal images acquired using the PD_T2 dual echo turbo spin echo sequence. A large popliteal cyst is evident, which is more clearly visible on the T2W image.

Figure 6- 2: 2D Sagittal Knee Images Obtained Using (L) PDW and (R) T2W Dual Echo TSE Sequence (TR=3000ms, TE=14/85ms, FA=180, SL=4mm, NoS=19, FoV=179cm, RFoV=7/8, Matrix=225x512, Scan Time=433, AC=2, Turbo=5, Sequence Name=tse5_14b130_85b130, no FATSAT) The cyst is probably fluid filled, and is contrasted very well on the T2W data relative to the
47

CHAPTER SIX

MRI CLINICAL APPLICATIONS

surrounding muscle which is more hypointense. The cyst is also evident on the PDW image, and this is most likely because the content of the cyst has a larger proton density than the surrounding muscle tissue. A dual echo steady state (DESS) sequence was also used in addition to the dual echo turbo spin echo sequence. The DESS sequence provides rather complex image contrast, which is roughly equivalent to T2 weighting. The DESS sequence is effectively a combined FISP and PSIF sequence, where the FID signal from the FISP and the echo signal from the PSIF are added together. This improves the signal to noise ratio of the final image. The DESS sequence (TR 27ms, TE 9 ms, FA 40o) again clearly highlights the lateral soft tissue effusion in figure 6-3 (left) and the popliteal cyst figure 6-3 (right) below.

Figure 6- 3: 3D DESS Sequence showing (L) Lateral Soft Tissue Effusion and (R) Popliteal Cyst (TR=27ms, TE=9ms, FA=40, SL=1.5mm, NoS=64, FoV=149cm, RFoV=7/8, Matrix=192x512, Scan Time=534, AC=1, Sequence Name=de3d_9b130, no FATSAT) Finally, an inversion recovery (STIR) sequence was implemented to search in more detail for marrow oedema as shown by figure 6-4 below. The image does not appear to show marrow oedema but again clearly highlights the soft tissue swelling on the lateral side of the knee.

48

CHAPTER SIX

MRI CLINICAL APPLICATIONS

Figure 6- 4: STIR Sequence. No Marrow Oedema but Clear Swelling on Lateral Side of Knee (TR=6091ms, TE=30ms, TI=150ms, FA=90, SL=4mm, NoS=23, FoV=179cm, RFoV=7/8, Matrix=378x512, Scan Time=535, AC=1, Turbo=7, Sequence Name=tirm7_30b130, STIR FATSAT) This confirms that the soft tissue swelling is almost certainly fluid filled, since fat suppression does not appear to affect the area. This could be further confirmed by running a FLAIR sequence to suppress the fluid, although this was not implemented in this particular study. The Radiologists report highlighted the several features derived from the MR images of the above examination. Knee effusion and popliteal cyst is present, deficiency of the anterior cruciate ligament, tears to both horns of the lateral meniscus, and small parameniscal cyst and finally a vertical tear to the posterior horn of the medial meniscus 2 MR Angiography (Renal Arteries)

There are three main methods of obtaining images of the human vasculature using MRI. These MR angiographic techniques are known as time-of-flight (TOF) MRA, phase contrast (PC) MRA and contrast-enhanced (CE) MRA. 1 Time-Of-Flight MRA

TOF MRA utilises the principle of in-flow enhancement as illustrated in chapter 9, slide 6 (artefacts lecture). This method uses a T1W gradient echo sequence where the initial RF pulse saturates the blood within the slice. At the time the data is acquired however, the original blood within the slice would have left the slice and been replaced by fresh blood that did not experience the initial RF pulse. Consequently the new blood protons would appear to have fully recovered to longitudinal equilibrium, relative to the surrounding stationary tissue protons within the slice. TOF MRA is still used routinely for imaging the cranial arteries around the Circle of Willis. 2 Phase Contrast MRA

Phase contrast (PC) methods use a different technique to generate vascular contrast, based on phase shifts of the flowing blood. Following the initial 90 pulse, bipolar gradients are applied separately along the three axes to induce phase shifts to moving protons. Protons in stationary tissues acquire no net phase change as a result of the bipolar gradient pulses, but flowing protons within vessels accumulate phase as they move. For the next
49

CHAPTER SIX

MRI CLINICAL APPLICATIONS

RF pulse, the polarity of the bipolar gradient is inverted, and a subtraction technique can eliminate the stationary background signal, to leave the moving blood clearly visible. An important variable in PC MRA is the velocity-encoding (VENC) factor, which can be controlled to highlight arteries or veins. Higher VENC factors (e.g. 60-80 cm/sec) will selectively image the arteries, whereas slower VENC values (e.g. 20 cm/sec) will highlight the venous system. The phase images from the PC MRA sequence can be displayed to show direction of flow and to quantify these flow velocities within a vessel 2. 3 Contrast Enhanced MRA

Contrast enhanced MRA is most widely used method of MRA in clinical MRI. The use of a gadolinium based contrast agent significantly reduces the T1 of the blood as the bolus of contrast agent passes through the vessels and a fast T1 weighted gradient echo sequence provides well defined, hyperintense vessels, provided that the images are acquired at the appropriate time. The technique works by subtraction of post-contrast images from a pre-contrast baseline image. This will be described in more detail using the following case of renal CE-MRA as an example. 4 CE-MRA Examination

A female patient aged 74 was referred for a CE-MRA examination on the basis of hypertension and suspected cardiovascular disease. The examination consisted of four phases, namely the planning phase, the bolus-timing phase, the CE-MRA phase and the data analysis phase. The body phased array RF coil was selected for the examination. The coil was positioned on the patient, and an appropriate vein in the arm of the patient was identified for contrast injection. The patient was positioned supine and head first into the Siemens Symphony scanner. The first phase of the study was implemented to ensure that the appropriate slice orientation for CE-MRA was selected. Scout images were obtained (with the patient in breath hold) using the TruFisp sequence (TR 6ms, TE 3ms, 4mm slice thicknesses). An example of a sagittal TruFisp image is shown in figure 6-5 below.

Figure 6- 5: 2D Sagittal TruFisp image of Abdomen (TR=6ms, TE=3ms, FA=70, SL=4mm, NoS=11, FoV=458cm, RFoV=8/8, Matrix=256x256, Scan Time=8 to 17, AC=1, Sequence Name=trufi_3b560, No FATSAT) From this dataset and the other orientations acquired, the correct imaging plane (running parallel and through the descending aorta) for coronal oblique CE-MRA was identified. The appropriate positioning of the coronal oblique slices is shown in dotted line in figure 6-5
50

CHAPTER SIX above.

MRI CLINICAL APPLICATIONS

Once the appropriate planning had been completed, a bolus timing experiment was undertaken. This was achieved by delivering a 2ml test bolus of gadolinium contrast agent (Prohance, Bracco Inc), and then immediately acquiring back-to-back axial 3D volume interpolated breathold examination (VIBE) datasets with a temporal resolution of just under seven seconds (see figure 6-6 A-C below).

A B

C Figure 6- 6: (A C) 3D (VIBE) Datasets With Temporal Resolution approx. 7s (TR=4.2ms, TE=1.9ms, FA=40, SL=3.1mm, NoS=32, FoV=498cm, RFoV=8/8, Matrix=256x256, Scan Time=6.8, AC=1, Sequence Name=fl3db_80m_2b488, FATSAT) The approximate time taken for the contrast agent to reach the renal arteries was calculated using this method. The advantage of using this method is that it is also possible to obtain an assessment of the perfusion of the contrast agent through the renal arteries and into the renal cortex 3.
51

CHAPTER SIX

MRI CLINICAL APPLICATIONS

After the delay time for CE-MRA had been calculated from the VIBE timing experiments, the third phase, CE-MRA was implemented. The radiographers acquired a baseline coronal oblique 3D dataset (with slices less than 1.5mm thick) using a FLASH sequence with fat suppression prior to contrast agent delivery. The contrast agent was then delivered as an 18ml bolus with 20ml saline flush into the vein in the arm, and a delay time equivalent to that calculated from the VIBE timing experiment was implemented, prior to the next 3D coronal oblique (arterial phase) FLASH sequence. Finally, a third coronal oblique (venous phase) FLASH sequence was acquired in order to highlight the kidneys more clearly for the purposes of kidney length and volume measurements (see chapter 8). An example of a pre-contrast FLASH image, together with arterial phase post-contrast image and subtracted dataset can be observed in figures 6-7 A C below.

A B

C Figure 6- 7: (A-C) Examples of a Pre-Contrast Coronal 3D FLASH image, Arterial Phase Post-Contrast and Subtracted Dataset (TR=4.6ms, TE=1.8ms, FA=25, SL=1.4mm, NoS=56, FoV=419cm, RFoV=8/8, Matrix=512x512, Scan Time=up to 18, AC=1, Sequence
52

CHAPTER SIX Name=fl3d_itn_2b390, FATSAT)

MRI CLINICAL APPLICATIONS

The data analysis phase consisted of generating a set of maximum intensity projection (MIP) images. This was achieved by subtracting the arterial phase data from the pre-contrast data, and using the MIP software on the subtracted data. The MIP algorithm provides a series of angular projections through the subtracted 3D data, where the maximal signal intensities are thresholded and highlighted. This allows the visualisation of the renal arteries more clearly through multiple planes. A diagnosis of renal artery stenoses was then made by a radiologist who classified the thinning of the arteries as minimal (0-30% occlusion), moderate (30-60% occlusion) or severe (70-100% occlusion). The highlighted example in figure 6-8 below shows a patient with tortuous vessels supplying both kidneys. However, the lumen of the right renal artery appears normal, whilst the left renal artery lumen is severely thinned and may require intervention by angioplasty or stent placement.

Figure 6- 8: Outcome of CE-MRA Shows Tortuous Vessels Supplying Both Kidneys 3 Neuro MRI (Brain)

A 61-year-old male was referred for an MRI examination with a history of complex visual and vocal problems. The patients head was positioned in a transmit-and-receive quadrature RF head coil, and the coil was subsequently centred into the magnet (Siemens Symphony scanner). The first sequence used was the dual echo turbo spin echo sequence, where PDW and T2W axial slices of the entire brain were acquired as shown in figure 6-9 below.

53

CHAPTER SIX

MRI CLINICAL APPLICATIONS

Figure 6- 9: 2D Dual Echo TSE Sequence Showing (L) PDW and (R) T2W Axial Slices of Brain (TR=3000ms, TE=119/17ms, FA=180, SL=5mm, NoS=20, FoV=250cm, RFoV=6/8, Matrix=200x256, Scan Time=4 3, AC=2, Sequence Name=tse5_17b130_119b130, No FATSAT) There is some evidence of proton density change on the left side of the brain, but the lesion areas are much more clearly visible on the T2 weighted image (figure 6-9 right). A standard T1W multi-slice spin echo sequence was acquired in the sagittal orientation shown in figure 6-10 below, but the lesions were quite difficult to spot on this image.

Figure 6- 10: 2D T1W Multi-Slice Spin Echo Sequence Showing Sagittal Brain (TR=525ms, TE=15ms, FA=90, SL=5mm, NoS=19, FoV=250cm, RFoV=6/8, Matrix=192x256, Scan Time=3 25, AC=1, Sequence Name=se15b130, No FATSAT) Finally, an axial turbo FLAIR sequence was implemented with TR 9000ms, TE 110ms and TI 2500ms. These data highlight the lesion areas as hyperintense relative to the rest of the
54

CHAPTER SIX brain tissue as seen in figure 6-11 below.

MRI CLINICAL APPLICATIONS

Figure 6- 11: 2D FLAIR Sequence Showing Lesions Hyperintense Relative to Normal Brain (TR=9000ms, TE=110ms, TI=2500, FA=180, SL=5mm, NoS=19, FoV=250cm, RFoV=6/8, Matrix=176x256, Scan Time=2 33, AC=1, Sequence Name=tirm11_ild_100b195, No FATSAT) The appearance of hyperintense lesions on T2W and FLAIR images may be suggestive of an internal bleed, together with the presence of fluid and blood by-products (such as haemosiderin). If one compares the T2W turbo spin echo and the FLAIR images, small central lesions can be seen that are hyperintense on the T2W image and hypointense on the FLAIR image. These lesions are possibly fluid filled - with a T1 equivalent to CSF (since the chosen TI in the FLAIR acquisition should eliminate the signal from CSF). The radiologists report highlighted the following features derived from the MR images of the above examination. There was evidence of previous infarction within the left posterior cerebral artery territory, (this includes the visual cortex). Also there appears to be ischemic change within both middle cerebral arterial territories 4 Neuro MRI (Spinal Cord)

In this example, a 20 year-old female was referred for MRI for a brain and spinal scan for suspected MS plaques. The cervical spine was initially scanned with the patient supine and head first in the scanner (Siemens Symphony). T1W spin echo and T2W turbo spin echo sequences were both used in the sagittal orientation, with the neck coil, and appropriately selected spine coil elements. The resulting images showed little information on the T1W data as shown in figure 6-12 (left) below, but areas of hyperintense signal in the spinal cord on the T2W data in figure 6-12 (right).

55

CHAPTER SIX

MRI CLINICAL APPLICATIONS

Figure 6- 12: TSE Sequences Showing Sagittal Spine: (L) 2D T1W SE (TR=501ms, TE=15ms, FA=90, SL=3mm, NoS=13, FoV=280cm, RFoV=6/8, Matrix=224x512, Scan Time=5 39, AC=3, Sequence Name=se_15b130, No FATSAT) and (R) T2W TSE (TR=5000ms, TE=112ms, FA=180, SL=3mm, NoS=14, FoV=279cm, RFoV=6/8, Matrix=279x279, Scan Time=4 50, AC=3, Turbo=15, Sequence Name=tse15_112b130, No FATSAT) An axial MEDIC sequence was then implemented in the region of the signal change on the c-spine, derived from figure 6-12. A MEDIC sequence (multiple echo data image combination) was used instead of a T2 weighted sequence in this case because the MEDIC sequence contains flow compensation gradients that remove CSF flow artefacts. The sequence is basically a gradient echo sequence with multiple echo acquisitions resulting from reversed read out gradients. The multiple echoes are then superimposed onto the first image with the shortest echo time, which results in a predominantly T2 weighted final image. In this case, an axial slice at the level of one of the high-signal lesions can be seen in figure 6-13 below.

Figure 6- 13: (L) Axial MEDIC Sequence Shows Predominantly T2W Image (TR=775ms, TE=27ms, FA=30, SL=3mm, NoS=15, FoV=229cm, RFoV=6/8, Matrix=192x256, Scan Time=4 59, AC=2, Sequence Name=me2d_6u_27b195, No FATSAT) and (R) 2D Sagittal FLAIR sequence eliminates CSF signal (TR=9000ms, TE=110ms, TI=2500, FA=180, SL=3mm, NoS=13, FoV=300cm, RFoV=6/8, Matrix=352x512, Scan Time=4 39, AC=1,
56

CHAPTER SIX

MRI CLINICAL APPLICATIONS

Turbo=11, Sequence Name=tirm11_ild_110b195, WATER SAT) Again, this shows an area of high signal within the spinal cord. Finally a sagittal FLAIR sequence was run, with inversion time TI = 2500 ms to eliminate the CSF signal. It can be seen in this case that the lesion is hyperintense relative to the rest of the spinal cord, which would be consistent with the suspected demyelinating disease that is suspected. The radiologists report highlighted the following features derived from the MR images of the above examination. A small plaque is visible at C1 and a larger one at C3. Axial images confirm the position of the plaques and demyelination is suspected. 5 Body MRI (Breast)

To demonstrate how MRI is used for breast imaging, data obtained from a female patient (aged 41) who was scanned for suspected recurrence of metastatic disease in the left breast was observed. The patient was positioned prone with both breasts enclosed by the breast RF coil, and imaging was performed on the 1T Siemens Impact scanner. A coronal high-resolution 3D T1W gradient echo sequence (FLASH) was used to acquire thin slices through the breasts, and a T2W spin echo sequence with fat saturation was also implemented. Neither the high resolution T1W images nor the fat-suppressed T2W images in figure 6-14 were able to provide any direct indication of recurring metastases, although both sets of images did indicate that the normal anatomy of the left breast was somewhat disrupted (it is interesting to compare the structural differences in the circled region of the left breast with a similar symmetrical area on the right breast).

Figure 6- 14: (Top) Coronal 3D T1W GE Sequence (FLASH) (TR=14.2ms, TE=6.7ms, FA=35, SL=2.5mm, NoS=64, FoV=339cm, RFoV=4/8, Matrix=192x512, Scan Time=2 56, AC=1, Sequence Name=fl3d_7b150n3, No FATSAT) and (Bottom) Coronal 3D T2W SE Sequence with FATSAT (TR=9365ms, TE=90ms, FA=180, SL=2mm, NoS=13, FoV=339cm, RFoV=4/8, Matrix=98x256, Scan Time=4 31, AC=1, Sequence Name=tse7_90rrb130, FATSAT) In order to further characterise this suspicious region of the left breast, a dynamic contrast enhanced T1W study was performed. The T1W study consisted of 3D T1W coronal gradient echo FLASH images through the breasts. The sequence parameters were TR 14ms, TE 7ms, with 2.5mm thick slices being acquired over a 340cm field of view. The imaging time was just over 1min 30 seconds for each sequence. Two pre-contrast datasets
57

CHAPTER SIX

MRI CLINICAL APPLICATIONS

were acquired (figures 6-15 a and b), and then 20ml of contrast agent (Prohance, Bracco Inc) followed by a saline flush was delivered into a vein in the arm. Immediately following contrast agent injection, a series of five post-contrast FLASH datasets were acquired (one after another figures 6-15 c-f) each with a scan time of just over 1min 30 seconds.

Figure 6- 15: Dynamic Contrast Enhanced 3D T1W Study (TR=14ms, TE=7ms, FA=35, SL=2.5mm, NoS=64, FoV=340cm, RFoV=4/8, Matrix=98x256, Scan Time=1 29, AC=1, Sequence Name=fl3d_7b150, No FATSAT) Regions of interest were then placed over the suspicious region (circled) where the contrast agent was taken up (see example subtracted dataset in figure 6-16 below), along with an area of suspected healthy breast tissue.

Figure 6- 16: Subtracted Dataset This analysis was performed for each dataset that was acquired. Plots of signal intensity versus scan number (= time) yielded contrast uptake curves (figure 6-17) for suspected lesion area and healthy breast tissue.

58

CHAPTER SIX

MRI CLINICAL APPLICATIONS

Figure 6- 17: Contrast Uptake Curves for Suspected Lesion Area and Healthy Breast Tissue The rapid uptake of contrast agent in the suspicious lesion, followed by gradual washout over the period of the study suggested that the lesion was malignant. Had the contrast uptake been slower and steadier over the period of this study, with no washout period then this would have indicated the presence of a benign lesion. The radiologists report confirmed that the lesion was indicative of metastatic disease on the basis of these MRI findings. 6 Body MRI (Liver)

Liver imaging was observed on two patients, who both had suspected liver lesions. The patients were referred for MRI in the hope that it would be possible to identify the nature and extent of these lesions. Imaging was performed with each patient supine on the Siemens Symphony scanner, and the body array coil was used in combination with selected elements from the spine coil. All sequences chosen provided axial breath hold images, so scan times were restricted to 20 seconds or less. The specific sequence parameters are shown below: 1. T1W GE (2D FLASH) - TR=102ms, TE=4.5ms, FA=70, SL=6mm, NoS=13, FoV=398cm, RFoV=6/8, Matrix=512x512, Scan Time=20, AC=1, Sequence Name=fl2d_4b260, No FATSAT). 2. T2W 2D HASTE - TR=1100ms, TE=120ms, FA=150, SL=6mm, NoS=12, FoV=398cm, RFoV=6/8, Matrix=512x512, Scan Time=15, AC=1, Sequence Name=haste_60b488, FATSAT). 3. T2W 2D STIR - TR=4311ms, TE=66ms, FA=160, SL=6mm, NoS=12, FoV=398cm, RFoV=6/8, Matrix=512x512, Scan Time=21, AC=1, Turbo=33, Sequence Name=tirm33_78b325, STIR FATSAT).
59

CHAPTER SIX

MRI CLINICAL APPLICATIONS

4. T1W 2D GE QFS (Quick FATSAT) - TR=148ms, TE=2.3ms, FA=70, SL=6mm, NoS=22, FoV=369cm, RFoV=6/8, Matrix=512x512, Scan Time=20, AC=1, Sequence Name=fl2d_qfs_2b488, FATSAT). A T1W gradient echo sequence was run for each patient (see figures 6-19 and 6-20 images 1a and 1b), and the lesion appeared as hypointense relative to the rest of the liver (indicating that the lesion probably has a longer T1 than the liver in each case).

Figure 6- 18: FLASH, HASTE, FATSAT and STIR Images of Suspected Liver Lesion A T2W HASTE sequence (HASTE = Half Acquisition Turbo Spin Echo) and T2W STIR both highlighted the lesions as hyperintense for both patients (see figures 6-19 and 6-20 images 2a, 2b, 3a, 3b).

Figure 6- 19: FLASH, HASTE, FATSAT and STIR Images of Suspected Liver Lesion
60

CHAPTER SIX

MRI CLINICAL APPLICATIONS

This indicates that each lesion probably has a longer T2 than the liver in both cases. A HASTE sequence is basically a fast version of a Turbo Spin Echo sequence, where just over half of k-space is acquired in the same way as a Turbo Spin Echo sequence, and the other half of k-space is synthesised (based on the first half of the acquisition). This ensures that HASTE is ideal for heavily T2 weighted breath hold examinations, and is ideal for cases such as body imaging. Finally a fast T1W gradient echo FLASH sequence was implemented before and after administration of a bolus of contrast agent (figures 6-19, 6-20 and 6-21 images 4a, 4b, 5a, 5b).

Figure 6- 20: Post Gadolinium T1W Enhancement The pre-contrast images provided essentially the same data as (images 1a and 1b), but the post-contrast data showed different enhancement patterns. The lesion for the first patient demonstrated rim enhancement, whilst the lesion for the second patient demonstrated complete enhancement. These differences in enhancement patterns provide additional confidence by which it is possible to diagnose the nature of the lesion. Post-contrast rim enhancement is often indicative of a metastatic lesion, whilst complete enhancement is often indicative of a haemangioma. The radiologists final report confirmed these suspicions on the basis of these MRI investigations. 3 Conclusion on Clinical Applications

The time I spent shadowing radiographers in the MRI suite at Ninewells Hospitals was immensely beneficial to me. It enabled me to better understand the various sequences used and why they were used in particular cases. This chapter has presented a range of case studies I was able to observe, covering musculoskeletal MRI (knee), MR angiography (renal arteries), Neuro MRI (brain), Neuro MRI (spinal cord), body MRI (breast) and body MRI (liver). The cases presented here used basic T1W, T2W and PDW spin echo and turbo spin echo sequences as well as inversion recovery sequences (STIR and FLAIR) and gradient echo sequences (predominantly FLASH and DESS). Some of these sequences also involved the use of contrast agents or fat suppression techniques and these methods were all described and justified. 4 [1] References Buxton R. B. Introduction to Functional Magnetic Resonance Imaging: Cambridge
61

CHAPTER SIX

MRI CLINICAL APPLICATIONS

University Press, 2002: 126-128. [2] Houston J. G., Gandy S. J., Allen A., Dick J. B., Belch J. J., Stonebridge P. A. Spiral laminar flow in the abdominal aorta: A predictor of renal impairment deterioration in patients with renal artery stenosis? Nephrology Dialysis Transplantation (2004 in press). [3] Sudarshan T., Gandy S., Armoogum K., Allan L., Sheppard D., Houston G. Measurement of Renal Cortical Perfusion in Patients with Renovascular Disease Using MRI. Nephrol. Dial. Transplant. 2003; 18 (Suppl 4): W324. [4] McIntyre C. Magnetic Resonance Imaging Portfolio, 2001: 30-31. [5] Kim J. H., Kim M. J., Park S. I. et al. MR cholangiography in symptomatic gallstones: Diagnostic accuracy according to clinical risk group. Radiology 2002; 224:410-416. [6] Van Hoe L., Gryspeerdt S., Vanbeckevoort D. et al. Normal Vaterian sphincter complex: evaluation of morphology and contractility with dynamic single-shot MR cholangiopancreatography. AJR Am J Roentgenol 1998; 170:1497-1500.

62