The current trend in magnetic resonance imaging (MRI) is towards higher external magnetic field strengths (B0) to take advantage of increased sensitivity and signal to noise ratio (SNR). Unfortunately, as (B0) increases the effectiveness (relaxivity) of clinical gadolinium (Gd3+)-based contrast agents (CAs) administered to enhance image contrast is significantly reduced. Excellent soft tissue contrast can be generated with current agents despite their non-optimum relaxivities but necessitates large doses. The limits of detection of a CA at high B0 fields can be lowered by recovering the lost relaxivity and is a pre-requisite to the goal of molecular imaging in which CAs are bound to biomarkers of pathology that exist at very low concentrations. Traditional methods for increasing the detectability of CAs have focused on optimizing critical parameters identified from the Solomon-Bloembergen-Morgan (SBM) theory that affect relaxivity. Gains in relaxivity with these methods to date have been modest and are far from the theoretical maximum possible. Although researchers continue to investigate novel complexes that provide improved relaxivities, any such complex would require a lengthy and costly approval process with the U.S. Food and Drug Administration (FDA). Therefore, a method that affords improved relaxivities of current clinically approved CAs, particularly at high B0 fields, that could be adopted into clinical practice rapidly, is of great interest. Spin locking is a nuclear magnetic resonance (NMR) technique that was introduced for imaging in 1985, but has received very little attention in combination with Gd3+-based CAs. The technique employs a low power long duration radiofrequency (RF) pulse (B1) parallel to the net magnetization in the x,y-plane. This locks the magnetization into lower precessional frequencies around an "effective" field (Beff) that is reduced with respect to B0 but maintains the high field advantages required for imaging. When considered in the rotating frame, longitudinal relaxation of the magnetization against Beff exhibits shorter time constants (T1p) expected at these lower precessional frequencies. This leads to higher relaxivities, which has implications for increasing CA detectability. The experiments described herein show that rotating frame longitudinal relaxivities (r1p) for current clinical Gd3+-based CAs are essentially independent of the strength of the spin lock pulse (yB1) as predicted by theory. This result is important because it allows the value of yB1 to be neglected when comparing r1p of Gd3+-based CAs across several B0 fields. The magnetic field dependence of r1p for all clinical agents showed that relaxivity, lost by moving to higher fields, could be "recovered" and that r1p was sensitive to the rotational correlation time constant (TR) of the agent. Using high molecular weight Nanoassembled capsules (NACs) containing a Gd3+-based CA to probe this finding further, we were able to generate relaxivities at high field up to an order of magnitude greater than clinical agents at current imaging fields. These are beyond anything previously reported, or likely to be, with current techniques. Finally, we demonstrated that by spin locking Mn2+ agents, relaxivities at high field increased by a factor of ~ 30 than without spin locking, due to their larger dependence on scalar coupling. These findings show the potential of spin locking to increase detection limits dramatically at high field and are an exciting development towards the goal of molecular imaging.

The detection of disease and abnormal pathology by magnetic resonance imaging (MRI) has been aided significantly by the use of gadolinium (Gd3+)-based contrast agents (CAs) over the past three decades. MRI and MRI CAs play a critical role in diagnosing tumors and diseases of the central nervous system. The agents used clinically have been shown to safely increase MRI contrast despite the toxicity of Gd3+, owing to the high kinetic and thermodynamic stability of these chelates. However, current CAs enhance contrast at a small fraction of what is theoretically possible. This leads to the necessity of introducing high CA doses in practice in order to afford sufficient contrast. Lanthanide (Ln3+) chelates based on 1,4,7,10–tetraazacyclododecane–1,4,7,10–tetraacetate (DOTA) have been shown to be particularly stable and effective. Chelates of DOTA exist in two interconverting coordination geometries which have varying water exchange rates. Researchers have envisioned a way to increase the per-dose efficacy both by control of the Gd3+–inner–sphere water exchange rate and via binding specificity (i.e. to tumors). The efficacy gains using these approaches have thus far been modest. A thorough structural characterization of europium (Eu3+) chelates of a DOTA-derivative which cannot undergo conformational exchange was carried out. These studies show that a single enantiomer of the ligand (S)–2–(4-nitrobenzyl)–1,4,7,10–tetraazacyclododecane–1,4,7,10–tetra(α–methyl)acetate (NB–DOTMA) can yield chelates which are both diastereoisomeric (previously reported) and regioisomeric (not previously speculated). Molecular mechanics simulations generated from the characterization data indicate that the nitrobenzyl (NB) substituent is oriented in different directions for the two possible regioisomers. The NB group can be chemically converted to confer macromolecular binding capability, and the orientation of the NB substituent may have a significant impact on the binding and/or relaxation behavior of a prototypical CA. The nuclear magnetic resonance (NMR) spectra of Eu–NB–DOTMA at various temperatures were compared. Unexpectedly, the chelates showed time-averaged structures which differ with a change in water exchange rate — the faster the rate, the greater the deviation from the expected structure. Consideration of the structures of Ln3+ chelates without accounting for their dynamic behavior does not yield an accurate value for the time-averaged hydration state. These observations suggest the "optimal" water exchange rate calculated using Solomon-Bloembergen-Morgan (SBM) theory may not lead to the highest-efficacy CAs. Binding and relaxometric studies of macromolecule-targeting derivatives of Gd–NB–DOTMA both by the author and in another lab showed that the coordination isomer with the slower water exchange rate should lead to more effective contrast, in direct opposition to the prevailing view of water exchange in the MRI community. Preliminary data do not indicate that regioisomeric chelates have significantly different relaxivity or macromolecular binding behavior. The ratios between regioisomeric Ln–NB–DOTMA chelates formed were shown to be dependent on the concentration and pH of the chelation reaction, but conditions were not found which led to the production of a single regioisomer. Attempts to carry out an efficient synthesis of a Ln3+ chelate with no potential for regioisomerism were unsuccessful.

Indiana University-Purdue University Indianapolis (IUPUI) / Adaptive optics-optical coherence tomography (AO-OCT) is a non-invasive method of imaging the human retina in vivo. It can be used to visualize microscopic structures, making it incredibly useful for the early detection and diagnosis of retinal disease. The research group at Indiana University has a novel multi-camera AO-OCT system capable of 1 MHz acquisition rates. Until this point, a method has not existed to process data from such a novel system quickly and accurately enough on a CPU, a GPU, or one that can scale to multiple GPUs automatically in an efficient manner. This is a barrier to using a MHz AO-OCT system in a clinical environment. A novel approach to processing AO-OCT data from the unique multi-camera optics system is tested on multiple graphics processing units (GPUs) in parallel with one, two, and four camera combinations. The design and results demonstrate a scalable, reusable, extensible method of computing AO-OCT output. This approach can either achieve real time results with an AO-OCT system capable of 1 MHz acquisition rates or be scaled to a higher accuracy mode with a fast Fourier transform of 16,384 complex values.