Session Y38: Invited Session: Quantitative Biomedical Imaging

Quantitative Magnetic Resonance Imaging and Phantom Development

Magnetic Resonance Imaging (MRI) uses strong magnetic fields and radiofrequency pulses to produce images of proton locations and properties. Image contrast reflects relative density of excited water protons, differences in relaxation times of water protons due to surrounding structure, and the sequence of RF pulses used to excite the water protons. MRI can be used to quantitatively measure longitudinal (T1) and transverse (T2) spin relaxation times, measure tissue volumes, track motion of water molecules (flow/diffusion), measure temperature, assess susceptibility differences, create maps of tissue electrical properties, etc. This talk will focus on quantitative measurement of relaxation times, diffusion and electrical properties. Diffusion MRI varies the homogeneous magnetic field using an initial gradient, followed by a refocusing gradient with the same magnitude with opposite direction: protons begin to precess at different rates, depending on the applied gradient, and will disperse. The refocusing gradient cannot refocus spins that have moved between gradient pulses, and the apparent proton diffusion can be calculated from the signal attenuation. Typically, gradient pulses are applied in three orthogonal directions to calculate a bulk diffusion coefficient. Tissue electrical properties can be mapped by measuring the complex RF transmit and receive fields (B1$+$, B1-). New methods estimate local electrical conductivity from \textit{in vivo} B1$+$ phase measurements based on the homogeneous Helmholtz equation. Quantitative relaxation measurements, diffusion and electrical properties can distinguish healthy tissue from malignant tumor from benign tumor or identify the time of a particular event, e.g. a stroke. In this talk, I will describe how the NIST system, diffusion, and breast phantoms help validate these important measurements.   

Quantitative Ultrasound: Transition from the Laboratory to the Clinic

There is a long history of development and testing of quantitative methods in medical ultrasound. From the initial attempts to scan breasts with ultrasound in the early 1950's, there was a simultaneous attempt to classify tissue as benign or malignant based on the appearance of the echo signal on an oscilloscope. Since that time, there has been substantial improvement in the ultrasound systems used, the models to describe wave propagation in random media, the methods of signal detection theory, and the combination of those models and methods into parameter estimation techniques. One particularly useful measure in ultrasonics is the acoustic differential scattering cross section per unit volume in the special case of the 180$^{\circ}$ (as occurs in pulse-echo ultrasound imaging) which is known as the backscatter coefficient. The backscatter coefficient, and parameters derived from it, can be used to objectively measure quantities that are used clinically to subjectively describe ultrasound images. For example, the ``echogenicity'' (relative ultrasound image brightness) of the renal cortex is commonly compared to that of the liver. Investigating the possibility of liver disease, it is assumed the renal cortex echogenicity is normal. Investigating the kidney, it is assumed the liver echogenicity is normal. Objective measures of backscatter remove these assumptions. There is a 30-year history of accurate estimates of acoustic backscatter coefficients with laboratory systems. Twenty years ago that ability was extended to clinical imaging systems with array transducers. Recent studies involving multiple laboratories and a variety of clinical imaging systems has demonstrated system-independent estimates of acoustic backscatter coefficients in well-characterized media (agreement within about 1.5dB over about a 1-decade frequency range). Advancements that made this possible, transition of this and similar capabilities into medical practice and the prospects for quantitative image-based biomarkers will be discussed.   

High Resolution Peripheral Quantitative Computed Tomography for Assessment of Bone Quality

The study of bone quality is motivated by the high morbidity, mortality, and societal cost of skeletal fractures. Over 10 million people are diagnosed with osteoporosis in the US alone, suffering 1.5 million osteoporotic fractures and costing the health care system over {\$}17 billion annually. Accurate assessment of fracture risk is necessary to ensure that pharmacological and other interventions are appropriately administered. Currently, areal bone mineral density (aBMD) based on 2D dual-energy X-ray absorptiometry (DXA) is used to determine osteoporotic status and predict fracture risk. Though aBMD is a significant predictor of fracture risk, it does not completely explain bone strength or fracture incidence. The major limitation of aBMD is the lack of 3D information, which is necessary to distinguish between cortical and trabecular bone and to quantify bone geometry and microarchitecture. High resolution peripheral quantitative computed tomography (HR-pQCT) enables in vivo assessment of volumetric BMD within specific bone compartments as well as quantification of geometric and microarchitectural measures of bone quality. HR-pQCT studies have documented that trabecular bone microstructure alterations are associated with fracture risk independent of aBMD.... Cortical bone microstructure -- specifically porosity -- is a major determinant of strength, stiffness, and fracture toughness of cortical tissue and may further explain the aBMD-independent effect of age on bone fragility and fracture risk. The application of finite element analysis (FEA) to HR-pQCT data permits estimation of patient-specific bone strength, shown to be associated with fracture incidence independent of aBMD. This talk will describe the HR-pQCT scanner, established metrics of bone quality derived from HR-pQCT data, and novel analyses of bone quality currently in development. Cross-sectional and longitudinal HR-pQCT studies investigating the impact of aging, disease, injury, gender, race, and therapeutics on bone quality will be discussed.   

Quantitative Magnetic Resonance Thermometry and Its Use with MR-Guided Focused Ultrasound

Focused ultrasound (FUS) uses a large area array, typically outside the body, that is geometrically or electronically focused to a point deep in the body. Such focusing provides amplification of the ultrasound intensity, thereby allowing heating of tissue to the point of coagulation at the focus, without damage to the intervening tissue. Guidance of FUS treatments deep in the body can be done quantitatively with magnetic resonance (MR) thermometry, termed MRgFUS. The physics behind MR thermometry lie in the changes in hydrogen bonding with temperature. As tissue temperature rises, hydrogen bonds break, allowing the return of the electron cloud to shield water protons, reducing the magnetic field seen by the protons, and the resonant frequency. The change in resonant frequency is -0.01 ppm per degree C and is the same for all aqueous tissues. The result of the shift in proton resonant frequency is seen in the phase of gradient echo images. Subtraction of the phase of images acquired before and during heating allows the removal of background phase from other sources, yielding quantitative temperature maps. Temperature standard deviations less than 1 degree C are readily achievable and thermal dose maps are easily calculated. Thermal dose is found from a conversion of the whole temperature-time curve to an equivalent number of minutes at 43 degrees C. A thermal dose of 240 minutes is often taken as the threshold for tissue damage. MR thermometry is complicated by the motion of the target tissue and/or motion of other organs such as occurs during respiration. More sophisticated algorithms than the simple baseline subtraction take advantage of the facts that motion can be repetitive (in the case of respiratory motion) and/or the fact that the focal region in MRgFUS is small, allowing for extraction of the heat from the phase profile without subtraction of a background phase.   

Development of Traceable Phantoms for Improved Image Quantification in Positron Emission Tomography

Clinical trials for new drugs increasingly rely on imaging data to monitor patient response to the therapy being studied. In the case of radiopharmaceutical applications, imaging data are also used to estimate organ and tumor doses in order to arrive at the optimal dosage for safe and effective treatment. Positron Emission Tomography (PET) is one of the most commonly used imaging modalities for these types of applications. In large, multicenter trials it is crucial to minimize as much as possible the variability that arises due to use of different types of scanners and other instrumentation so that the biological response can be more readily evaluated. This can be achieved by ensuring that all the instruments are calibrated to a common standard and that their performance is continuously monitored throughout the trial. Maintaining links to a single standard also enables the comparability of data acquired on a heterogeneous collection of instruments in different clinical settings. As the standards laboratory for the United States, the National Institute of Standards and Technology (NIST) has been developing a suite of phantoms having traceable activity content to enable scanner calibration and performance testing. The configurations range from small solid cylindrical sources having volumes from 1 mL to 23 mL to large cylinders having a total volume of 9 L. The phantoms are constructed with $^{\mathrm{68}}$Ge as a long-lived substitute for the more clinically useful radionuclide $^{\mathrm{18}}$F. The contained activity values are traceable to the national standard for $^{\mathrm{68}}$Ge and are also linked to the standard for $^{\mathrm{18}}$F through a careful series of comparisons. The techniques that have been developed are being applied to a variety of new phantom configurations using different radionuclides. Image-based additive manufacturing techniques are also being investigated to create fillable phantoms having irregular shapes which can better mimic actual organs and tumors while still maintaining traceability back to primary standards for radioactivity. This talk will describe the methods used to construct, calibrate, and characterize the phantoms, focusing on the preservation of the traceability link to the primary standards of the radionuclides used. The on-going development of specialized traceable phantoms for specific organ dosimetry applications and imaging physics studies will also be discussed.