The biological correlates of macroscopic breast tumour structure measured using fractal analysis in patients undergoing neoadjuvant chemotherapy.

Breast cancers are evolving, multi-scale systems that are characterized by varied complex spatial structures. In this study, we measured the structural characteristics of 33 breast tumours in patients who were to receive neoadjuvant chemotherapy using dynamic contrast enhanced MRI and fractal geometry. The results showed a significant association between fractal measurements and tumour characteristics. The fractal dimension was associated with receptor status (ER and PR) and the fractal fit was associated with response to chemotherapy, measured using a validated pathological response scale, tumour grade and size. This study describes structure measures that may be a consequence of known prognostic factors during the initial and/or maturation phase of tumour growth. These results suggest that measuring tumour structure in this way can predict an individual's response to neoadjuvant therapy and may identify those who will benefit least from neoadjuvant chemotherapy, allowing alternative treatment options to be selected in those patients.

Imaging vascular function for early stage clinical trials using dynamic contrast-enhanced magnetic resonance imaging.

Many therapeutic approaches to cancer affect the tumour vasculature, either indirectly or as a direct target. Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) has become an important means of investigating this action, both pre-clinically and in early stage clinical trials. For such trials, it is essential that the measurement process (i.e. image acquisition and analysis) can be performed effectively and with consistency among contributing centres. As the technique continues to develop in order to provide potential improvements in sensitivity and physiological relevance, there is considerable scope for between-centre variation in techniques. A workshop was convened by the Imaging Committee of the Experimental Cancer Medicine Centres (ECMC) to review the current status of DCE-MRI and to provide recommendations on how the technique can best be used for early stage trials. This review and the consequent recommendations are summarised here. Key Points • Tumour vascular function is key to tumour development and treatment • Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) can assess tumour vascular function • Thus DCE-MRI with pharmacokinetic models can assess novel treatments • Many recent developments are advancing the accuracy of and information from DCE-MRI • Establishing common methodology across multiple centres is challenging and requires accepted guidelines.

Blood oxygen level-dependent (BOLD) MRI: A novel technique for the assessment of myocardial ischemia as identified by nuclear imaging SPECT.

BACKGROUND: The different levels of deoxyhemoglobin in the ischemic myocardium, induced by stressors such as dipyridamole, can be detected by blood oxygen level-dependent (BOLD) MRI and may be used to diagnose myocardial ischemia. The aim of this study was to assess the signal change in the myocardium on BOLD MRI as well as wall thickening between rest and dipyridamole stress images in ischemic and non-ischemic myocardium as identified on SPECT imaging. METHODS: Twelve patients with stress-induced myocardial ischemia on SPECT underwent rest and dipyridamole stress MRI using a double breath-hold, T2()-weighted, ECG-gated sequence to produce BOLD contrast images as well as cine-MRI for wall thickening assessment in 10 of the 12 patients. Signal change on BOLD MRI and wall thickening were compared between rest and stress images in ischemic and non-ischemic myocardial segments as identified on SPECT. In each patient, two MRI slices containing 16 segments per slice were analysed. RESULTS: In total, there were 384 segments for BOLD analysis and 320 for wall thickening. For BOLD signal 137 segments correlated to segments with reversible ischemia on SPECT and 247 to normal segments, while for wall thickening 112 segments correlated to segments with reversible ischemia and 208 to normal segments. The average BOLD MRI signal intensity change was -13.8 (+/-16.3)% in the ischemic segments compared to -10.3 (+/-14.7)% in the non-ischemic segments (p=0.05). The average wall thickening was 6.4 (+/-3.4) mm in the ischemic segments compared to 8.7 (+/-3.8) mm in the non-ischemic segments (p<0.0001). CONCLUSION: Stress-induced ischemic myocardium has a different signal change and wall thickening than non-ischemic myocardium and may be differentiated on BOLD MRI. Larger studies are needed to define a threshold for detection and to determine the sensitivity and specificity of this technique.

Blood oxygen level dependent (BOLD) MRI: A novel technique for the detection of myocardial ischemia.

BACKGROUND: Blood oxygen level dependent (BOLD) T2* MRI detects signal variance within the myocardium based on changes in the deoxyhaemoglobin level following pharmacological stress, and it has the potential to identify areas of myocardial ischemia. The aim of the present study was to assess the utility of BOLD T2* MRI in the detection of myocardial ischemia in patients with an existing diagnosis of coronary artery disease. METHOD: Twenty-one patients with established three-vessel coronary artery disease on coronary angiography underwent rest and dipyridamole stress MRI using a double breath-hold T2* weighted ECG gated sequence. Analysis was performed on multiple short-axis slices of the heart, projected as a bull's eye. The myocardium was divided into three coronary territories, yielding 63 territories in total. A signal change between rest and stress of more than +/-4% was significant, implying a change in deoxyhaemoglobin concentration. A signal decrease or no changes denote the presence of ischemia, while a signal increase indicates no ischemia. RESULTS: All images were of sufficient quality for signal intensity analysis. In 12/63 territories (19%), a significant signal increase following stress was detected. A significant signal decrease was detected in 34/63 territories (54%), and in 17/63 territories (27%) there was a non-significant change. The presence of a perfusion defect was identified, therefore, in 51/63 (81%), based on the signal difference between rest and stress. CONCLUSION: Changes in myocardial oxygen level appear to be detectable by BOLD T2* MRI without using contrast media. Further, larger comparative studies are required to evaluate the diagnostic and prognostic impact of this technique and to compare it to the gold standard methods for the detection of myocardial ischemia.

The use of the Levenberg-Marquardt curve-fitting algorithm in pharmacokinetic modelling of DCE-MRI data.

The use of curve-fitting and compartmental modelling for calculating physiological parameters from measured data has increased in popularity in recent years. Finding the 'best fit' of a model to data involves the minimization of a merit function. An example of a merit function is the sum of the squares of the differences between the data points and the model estimated points. This is facilitated by curve-fitting algorithms. Two curve-fitting methods, Levenberg-Marquardt and MINPACK-1, are investigated with respect to the search start points that they require and the accuracy of the returned fits. We have simulated one million dynamic contrast enhanced MRI curves using a range of parameters and investigated the use of single and multiple search starting points. We found that both algorithms, when used with a single starting point, return unreliable fits. When multiple start points are used, we found that both algorithms returned reliable parameters. However the MINPACK-1 method generally outperformed the Levenberg-Marquardt method. We conclude that the use of a single starting point when fitting compartmental modelling data such as this produces unsafe results and we recommend the use of multiple start points in order to find the global minima.

A bandwidth indicator for magnetic resonance imagers.

Signal-to-noise ratios in magnetic resonance imaging are crucial in determining image quality, and dependent on a number of factors, one being the signal bandwidth per pixel. Not all manufacturers clearly state the bandwidth per pixel used for all sequences. A small battery-powered portable device is described which produces bright sharp lines on the magnetic resonance image at 10 kHz intervals in the frequency encoding direction. The bandwidth per pixel can then easily be calculated using electronic distance callipers, provided the image matrix and field of view are known. The device is expected to be especially of value when acceptance testing on poorly documented imaging systems.

Estimating achievable signal-to-noise ratios of MRI transmit-receive coils from radiofrequency power measurements: applications in quality control.

The inverse relationship between the radiofrequency (RF) power needed to transmit a 90 degree RF pulse, and the signal-to-noise ratio (SNR) available from a transmit-receive RF coil is well known. The theory is restated and a formula given for the signal-to-noise ratio from water, achievable from a single-shot MRI experiment, in terms of the net forward RF power needed for a rectangular 90 degree RF pulse of known shape and duration. The result is normalized to a signal bandwidth of 1 Hz and a sample mass of 1 g. The RF power information needed is available on most commercial scanners, as it is used to calculate specific absorption rates for RF tissue heating. The achievable SNR figure will normally be larger that that actually observed, mainly because of receiver noise, but also because of inaccuracies in setting RF pulse angles, and relaxation effects. Phantom experiments were performed on the transmit-receive RF head coil of a commercial MRI system at 0.95 T using a projection method. The measured SNR agreed with that expected from the formula for achievable SNR once a correction was made for the noise figure of the receiving chain. Comparisons of measured SNR figures with those calculated from RF power measurements are expected to be of value in acceptance testing and quality control.

A cine-SPAMM sequence for NMR imaging of pulsatile cerebrospinal fluid flow.

The technical difficulties in designing a SPAMM sequence to image pulsatile cerebrospinal fluid (CSF) flow at different phases of the cardiac cycle are described. The criteria used to select the most appropriate order of binomial SPAMM sequence are outlined. Data collection times required to view both cephalad and caudad flow for all R-R intervals were considered. The flip angles of the RF pulses required to produce images with equal CSF intensity throughout the cardiac cycle were investigated in detail.

The effects of renal variation upon measurements of perfusion and leakage volume in breast tumours.

Dynamic contrast enhanced MRI (DCE-MRI) and pharmacokinetic models have been used to measure tumour permeability (K(trans)) and leakage volume (ve) in numerous studies. The construction of pharmacokinetic models describing such tissue properties relies on defining the blood plasma concentration of contrast agent with respect to time (Cp(t)). When direct measurement is not possible a bi-exponential decay has been applied using data from healthy volunteers. This work investigates, by simulation, the magnitude of errors resulting from this definition with respect to normal variation in renal function and for cases with renal impairment. Errors up to 23% in ve and 28% in K(trans) were found for the normal simulations, and 67% in ve and 61% in K(trans) for the impaired simulations. If this bi-exponential curve is used as an input function to the generalized kinetic model and used in oncology, estimates of tissue permeability and leakage volume will possess large errors due to variation in Cp(t) curves between subjects.

The use of SPAMM to assess spatial distortion due to static field inhomogeneity in dental MRI.

In planning placement of dental implants using MRI, a SPAMM (spatial modulation of magnetization) magnetization preparation sequence was incorporated into a spin-echo imaging sequence. A phantom was imaged with a ferromagnetic object attached. Spatial distortion due to deviations in Larmor frequency was detected by a deviation of SPAMM lines. Both SPAMM line deviation and interline spacing were found to agree with a deltaB0 map generated from phase images. Imaging of a volunteer with and without typically used metallic implants positioned in a template showed SPAMM line deviations to correlate with expected deviations in vivo. SPAMM lines showed possible distortion due to chemical shift in the bone marrow and the presence of titanium implants to be insignificant. SPAMM may thus be used to provide a qualitative estimate of the accuracy of the MRI image when planning dental implants.

Estimating patient dielectric losses in NMR imagers.

Dielectric losses in the patient may impair radiofrequency receiver coil sensitivity, and transmitter coil efficiency, in nuclear magnetic resonance (NMR) imagers. The frequency dependence of this loss mechanism is derived. Patient losses in a solenoidal head coil used for imaging heads were simulated by a cylindrical saline phantom. The frequency dependence of the loss introduced by the phantom can indicate whether dielectric losses in the patient will be significant compared to eddy current losses. The detuning caused by the phantom is used to calculate an upper limit for the distributed stray capacitance between coil and patient. Given the approximate conductivity of the patient, an upper limit for the dielectric loss can be estimated. Some methods of reducing patient dielectric losses are suggested.

Estimating spatial resolution of in vivo MR images using spatial modulation of magnetization.

Spatial Modulation of Magnetization is shown to provide a means of estimating perceived spatial resolution directly in vivo. On the first magnetic resonance system tested, resolution in conventional spin echo images was found to be stability limited in the phase encoding direction and voxel limited (via the Nyquist sampling theorem) in the frequency encoding direction both in vitro and in vivo. As the voxel size approaches half the stripe separation, fringes of resolved and unresolved stripes are formed across the image. This phenomenon is explained and described mathematically. On a second magnetic resonance scanner, resolution in the phase encoding direction of fast spin echo images with centrically ordered phase encoding is shown to be voxel limited in substances with long T2, with poorer resolution in substances with short T2. Resolution in fast spin echo images with linearly ordered phase encoding was shown to be voxel limited in the phase encoding direction.

The influence of stimulated echoes on contrast in fast spin-echo imaging.

Tissue contrast obtained using fast spin-echo (FSE) and conventional spin-echo (SE) sequences is not identical and a number of mechanisms are thought to contribute to these contrast differences. The effect of stimulated echoes has previously been apparently ruled out as a contributing mechanism. Signal-to-noise ratios of single-slice matched FSE and conventional SE sequences were compared in aqueous solutions of CuSO4, Cr2(SO4)3 and MnSO4 with various T1 and T2 values. Enhancement of the FSE signal was observed in short T2 solutions and the effect was greater in samples where the T1 to T2 ratio was high. Reducing the refocusing pulse flip angle to increase the contribution from stimulated echoes also resulted in slightly increased enhancement. Experimental results were verified by computer simulations. Our results show that stimulated echoes do contribute to the contrast differences between FSE and conventional SE images and may be significant in the imaging of brain hemorrhage.

Choice of contrast enhancement index for dynamic magnetic resonance mammography.

Indices are often used in dynamic MRI of the breast to quantitate signal enhancement within suspicious lesions. Two indices are commonly used: one calculates the difference in pre- and postcontrast signal intensity, normalised to a base-line signal intensity such as that of fat (which does not enhance) whilst the other calculates the ratio of pre- to postcontrast signal intensity. The results of a computational simulation are presented which demonstrate the superiority of the normalised signal difference index, based on the criterion that the best index is that which is least influenced by initial tissue T1. This hypothesis was tested by comparing the two indices in a group of patients with clinical or mammographic suspicion of recurrent breast carcinoma. Of 37 patient examinations using Gadolinium enhanced MRI of the breast, 11 patients showed 13 lesions with some degree of enhancement, which were subsequently diagnosed histologically as either benign or malignant. The normalised signal difference index showed no overlap between the benign and malignant groups, whereas some overlap was observed with the signal ratio index. The clinical findings are therefore consistent with the results of the computational simulation.

Determination of normal regional left ventricular function from cine-MR images using a semi-automated edge detection method.

A semi-automated edge detection method for the delineation of the endo- and epicardial borders of the left ventricle from cine MR images has been developed. The feasibility of this was demonstrated by processing end diastolic and end systolic ECG-gated images of four short axis images in 10 healthy subjects. The first derivative method combined with a 2D weighted polynomial fitting procedure was used to determine the endo- and epicardial borders, which then allowed determination of the wall motion, wall thickening, and ejection fraction, of the left ventricle. The results show that the end-systolic radial wall motion varies from (32+/-8)% to (76+/-12)%, and wall thickening from (0.60+/-0.46) cm to (1.26+/-0.50) cm. An average ejection fraction of (69+/-6)% was found which agrees well with literature values. The method described, for the delineation of the borders, reduces considerably the long and tedious operator time inherent in manual measurement and greatly increases the reproducibility of the measurements.

The measurement of fetal liver T(*)(2) in utero before and after maternal oxygen breathing: progress towards a non-invasive measurement of fetal oxygenation and placental function.

Utero-placental insufficiency is thought to be a major cause of growth retardation in utero and an important risk factor in the perinatal period. The purpose of this study was to investigate whether MRI could detect changes of fetal oxygenation, based on the blood oxygenation level dependence (BOLD) of the MRI tissue signal. Nine third trimester women (34-38 weeks) with normal pregnancies underwent abdominal MRI examinations. Following localization of the fetal liver using T(2)-weighted single-shot HASTE scans, up to 7 breath-held transaxial single-slice gradient-echo image sets were obtained through the fetal liver. The mother then commenced oxygen breathing with the imaging procedure repeated after 20 minutes of O(2) breathing. For each image set, T(*)(2) values are calculated using linear regression of log (signal) versus TE for a region of interest within the fetal liver selected by the attending radiologist. Fetal liver T(*)(2) values were calculated before and after O(2) breathing for each multi-echo image acquisition set. A signed rank test was used to test for a significant change in fetal liver T(*)(2) between the pre-O(2) and post-O(2) image sets. A significant increase in T*(2) (alpha < 0.05) was seen in 5 of the 9 fetal livers, a smaller increase (of borderline statistical significance, alpha = 0.057) in 2 livers, and no significant change (alpha > 0.05) in 2 livers. Our study indicates that T(*)(2) measurement of the fetal liver may detect alteration in fetal oxygen level following maternal oxygenation using the BOLD effect. This technique may potentially be applied to the identification and understanding of placental dysfunction in intra-uterine growth retardation.

Magnetic resonance imaging screening in women at genetic risk of breast cancer: imaging and analysis protocol for the UK multicentre study. UK MRI Breast Screening Study Advisory Group.

The imaging and analysis protocol of the UK multicentre study of magnetic resonance imaging (MRI) as a method of screening for breast cancer in women at genetic risk is described. The study will compare the sensitivity and specificity of contrast-enhanced MRI with two-view x-ray mammography. Approximately 500 women below the age of 50 at high genetic risk of breast cancer will be recruited per year for three years, with annual MRI and x-ray mammography continuing for up to 5 years. A symptomatic cohort will be measured in the first year to ensure consistent reporting between centres. The MRI examination comprises a high-sensitivity three-dimensional contrast-enhanced assessment, followed by a high-specificity contrast-enhanced study in equivocal cases. Multiparametric analysis will encompass morphological assessment, the kinetics of contrast agent uptake and determination of quantitative pharmacokinetic parameters. Retrospective analysis will identify the most specific indicators of malignancy. Sensitivity and specificity, together with diagnostic performance, diagnostic impact and therapeutic impact will be assessed with reference to pathology, follow-up and changes in diagnostic certainty and therapeutic decisions. Mammography, lesion localisation, pathology and cytology will be performed in accordance with the UK NHS Breast Screening Programme quality assurance standards. Similar standards of quality assurance will be applied for MR measurements and evaluation.

MRI developments in perspective.

Following Paul Lauterbur's seminal 1973 paper in Nature, considerable work was needed to overcome a number of physical, engineering and technical problems before the new technique of magnetic resonance imaging (MRI) could be applied clinically. Much of that pioneering work was done in the UK. Since the first head and whole-body images were obtained in the late 1970s, MRI has become a widely used clinical imaging modality capable of yielding tomographic images of excellent spatial resolution and tissue contrast. This review outlines the historical development of MRI in the context both of the technical problems which had to be overcome, and of the clinical uses of MRI. Current areas of research, such as the use of MRI to map brain function, the measurement of physiological parameters such as tissue perfusion, and the use of open-access real-time MRI to guide interventional procedures, are briefly discussed.

Fat suppressed magnetic resonance imaging at 0.5 T using binomial radiofrequency pulses.

Fat suppressed MR imaging can be achieved by selectively saturating lipid protons, just before applying a conventional spin-echo rf pulse sequence. The difference in the Larmor frequency between fat and water protons is only 3.5 ppm, so that the frequency response of the suppressing pulse, or pulses, has to be carefully designed. The choice of suitable binomial and modified trains of hard rf pulses was investigated by computer simulation, phantom testing and T1 weighted spin-echo imaging of a normal volunteer's optic nerve at 0.5 T.

Cine magnetic resonance imaging of pulsatile cerebrospinal fluid flow using CSPAMM.

Spatial modulation of magnetization (SPAMM) is a well established imaging technique which superimposes a tagging pattern on conventional magnetic resonance (MR) images, allowing movement to be visualized. A modification to the SPAMM technique, called complementary spatial modulation of magnetization (CSPAMM), which improves the contrast of the tagging pattern is explained. The application of CSPAMM to the visualization of pulsatile cerebrospinal fluid flow (CSF) using an 8 frame cardiac-gated cine sequence is described. Various combinations of binomial pulses, up to the fifth order, were investigated to see which produces the optimum tagging pattern in the CSPAMM images. The flip angles of the imaging RF pulses were studied to see which would give equal maximum CSF signal intensity in all the cine images. The optimized cine CSPAMM technique was compared in vivo with SPAMM and CSF motion was found to be more easily visualized in the CSPAMM images.