Spectroscopy is a useful tool for non-invasive medical diagnostic applications. In general, most medical applications of spectroscopy use near-infrared light to assess tissues. For this reason, photon penetration in tissues has been studied with infrared and near-infrared light, but it has not yet been determined for visible-range light. The purpose of this project is to determine the photon penetration depth of visible-range light in tissues and compare it to the near-infrared photon penetration depth. To accomplish this, I designed a micro-motion photon-migration mapping system and experimental protocol to map the 3D migration path of photons during reflectance spectroscopy in tissue phantoms. This photon-migration mapping system measured the 3D path of photons during their travel from a surface emitter, through a tissue phantom, and to a surface detector. The 3D map was then used to determine the most probable penetration depth of photons during reflectance spectroscopy in tissue phantoms. By using visible-range and near-infrared light, penetration depth results were compared at 580nm and 760nm. The photon-migration mapping system was also used to determine the effect of source/detector separation and varying optical properties (scattering coefficient and absorption coefficient) on photon penetration depth. Results from this project will be used in the design of optical probes for an intracellular oximeter which measures muscle oxygenation as an indicator of shock severity.

It is often necessary for children suffering from hydrocephalus to undergo frequent computerized tomography scans (CTs) of the brain in order to facilitate continued treatment. These scans provide physicians and parents visual feedback regarding the extent of the child’s condition, as well as the success of any treatments. Unfortunately, each scan exposes the child to small amounts of radiation, and repeated scans may cause adverse side effects. Magnetic Resonance Imaging (MRI), by contrast, poses no risk of radiation exposure, but is more resource intensive. This report reviews Seattle Children’s Hospital’s experience with the use of rapid acquisition MRI (HASTE), an alternative technique which circumvents both the radiation exposure associated with CT, as well as some of the time requirements and expense of conventional MRI. HASTE MRIs of hydrocephalus patients were reviewed by two evaluators to assess their viability as diagnostic tools. Images were rated on scales of overall image quality, catheter visualization, presence of motion artifact, and ventricular size. The study adds further support to the emerging evidence that HASTE MRI is an adequate substitute for CT scanning in evaluating ventricular size. However, HASTE MRI limits catheter visualization, a phenomenon unexplained by motion artifact or poor image quality. Despite this limitation, HASTE MRI can significantly reduce the need for CT scanning and concomitant radiation exposure to children and should therefore be considered a viable diagnostic alternative.

Diffusion-weighted imaging (DWI), a specialized form of magnetic resonance imaging (MRI), has shown promise for early detection of treatment response in tumors. DWI measures the mobility of water molecules in vivo and is sensitive to tissue characteristics such as cell density, membrane permeability, and microstructure. DWI studies of the breast have shown decreased diffusivity in malignant breast lesions, primarily attributed to the increased cell density associated with breast tumors. Previous studies have found that the DWI apparent diffusion coefficient (ADC) in tumors increases in response to treatment earlier than detectable changes in tumor size or vascularity found through conventional imaging techniques. This increase in ADC is thought to be due to cell death and necrosis, and may be a valuable early indicator of treatment efficacy. In this study we used DWI to measure the ADC of tumors in patients undergoing preoperative chemotherapy. Serial changes in breast tumor ADC were measured by region-of-interest (ROI) analysis, which involves using a computer program to trace the border of the tumor on an MR image of the breasts and to calculate ADC for the pixels within the region. We compared the changes in tumor ADC throughout treatment with the pathological findings of the removed breast tissue to find the relationship between changes in ADC and the efficacy of treatment. A better understanding of the relationship between changes in ADC and treatment response will allow treatment regimens to be tailored for maximum benefit to the individual patient. This would not only allow the patients to be put on regimens that are more likely to treat the disease, it prevents them from going through toxic treatments that have no survival benefit.

Muscular dystrophy is characterized by muscle weakness and loss of muscle tissue over time. It is often observed by muscle biopsy, which is invasive, and by appearance of external symptoms, which is nonspecific. Currently, there is promising development in finding a more effective treatment for the muscular dystrophy, and consequently a growing need to develop noninvasive imaging methods for tracking the damage and regeneration associated with muscular disease, treatment, and therapy. Our overall goal is to develop quantitative magnetic resonance imaging (MRI) methods that correlate with known cellular events in order to 1) derive markers that identify muscle inflammation, damage, and recovery and 2) monitor the time course and progression of these processes. This project explores T2, magnetization transfer, and diffusion weighted imaging methods to identify specific markers for muscle damage, edema, and inflammation during lesion progression and recovery of leg muscle in wild type mice. Components of this project include the design of the physical setup for experimentation, the characterization of image acquisition methods, and the determination of image analysis techniques to quantify the useful information from acquired data. The resulting imaging methods are used in a longitudinal study of three weeks to observe the damage and recovery of mouse muscle after microinjections of barium chloride toxin in the gastrocnemius and tibialis anterior muscles. Thus far, results from several in vivo experiments have shown phase differences in T2, magnetization transfer, diffusion, and volume measurements, with indications of inflammation immediately after toxin injection, peak muscle damage at 3-5 days post-injection, and gradual recovery of muscle to original condition over 3 weeks. The derived imaging markers better reveal the details of the regenerative process in muscle, and can eventually be applicable in muscle research and clinical observation of muscular disease.

As Type II diabetes affects more and more individuals, there has been a push to identify pathophysiological causes of the disease. A recent finding suggests that the insulin resistance that is associated with the disease (leading to high blood glucose levels) may be an indirect response to high fat levels. Essentially, it has been proposed that oxidative phosphorylation pathways become uncoupled (i.e. mitochondria produce a lower than normal amount of ATP per O2 molecule) in order to protect against reactive oxygen species (ROS) generated as a result of high fat metabolism. Based off of this theory, this study aims to test the hypothesis of a correlation between intramyocellular lipid (IMCL) levels and mitochondrial uncoupling. This project will non-invasively quantify, through Magnetic Resonance Spectroscopy (MRS), the concentration of IMCL in comparable volumes of the First Dorsal Interossei (FDI) muscle of the hand (a very well coupled muscle across all subjects) and the Tibialis Anterior (TA) muscle of the leg (a variably coupled muscle between subjects—data has already been collected as part of another experiment). The hypothesis is that well-coupled FDI, in terms of oxidative efficiency, will have low IMCL content while the variably coupled TA will have IMCL content that is inversely related to coupling. In order to produce MRS spectra of the two muscles, magnetic resonance coils that were specifically designed for each muscle group will be used. Upon completion of this project, we will have a better understanding of how IMCL is related to mitochondrial coupling and perhaps devise an IMCL marker for mitochondrial dysfunction and Type II diabetes. This may provide an additional tool for the diagnosis and early treatment of the disease.

Stem cells have regenerative functions, and possess the ability to migrate to and repair damaged or diseased tissue. Previously our group demonstrated that magnetic resonance imaging (MRI) can be used to track migration of pre-labeled neural stem cells (NSCs) to gliomas (brain tumor cells) in vivo, by labeling stem cells with Feridex. At the present, I am validating the ability of other stem cell types to migrate to gliomas using MRI tracking. The first are the same strain of NSCs that overexpress the ferritin gene, a natural MRI contrast agent. The second are the emerging class of induced pluripotent stem cells (iPSCs). Using a cell migration assay, I validated migration of both stem cell types in vitro. Now I am examining stem cell migration using mouse models. Before the cell transplantation, the iPS cells are labeled with the MRI agent Feridex using via simple incubation methods, and NSCs are transfected with the ferritin gene. Then labeled stem cells will be transplanted into mice through stereotactic injection at the contralateral side of the brain tumor. Nine days after transplantation MR imaging will be performed to monitor the possible migration of iPSC and NSC to gliomas, followed by histological confirmation. Successful migration of stem cells to the tumor sites should demonstrate a correlation between hypointensity on MR images and histological analysis at the tumor sites due to the labeling agents.

We are continuing research to implement a display and analysis tool to be used with a calibration phantom for medical imaging scanners. An effective method of diagnosing caner is to use dual-mode positron emission tomography and x-ray computed tomography (PET/CT) scans. Inherent variations between scans can lead to data that does not always indicate whether the cancer therapy is having the desired effect. We propose to use a calibration phantom, a solid plastic/epoxy container with spheres of radioactive germanium, as a reference control in cross-calibrating different scanners. Our hypothesis is we can utilize the contrast of CT images to determine consistent and accurate estimation of key parameters in PET images. First, a parser was created to read in the DICOM files used for medical image data transfer. The second phase is to find the regions of interest (ROIs) encompassing the spheres within each 3-dimensional image. Our hypothesis is to use the Otsu threshold to take advantage of the high contrast in the PET image to localize each sphere. These are then used to find more precise ROIs in the CT image, which shows better anatomical structure. The average and maximum radioactivity levels of each sphere can be found within each ROI on the PET image, which will be the data that will serve as the basis for the cross-calibration. Manual analyses conducted by our lab indicate that the variability is indirectly proportional to the size of the phantom’s spheres, and we expect our automated analysis tool to corroborate these results. We believe the automated analysis will be robust and efficient, enabling analysis of large data sets while removing any manual error in the analysis. Our results can be used to allow for a more reliable comparison of PET/CT images, and a more accurate diagnosis of cancer treatment effectiveness.

Corneal blindness, the second most common form of treatable blindness worldwide is when the transparent tissue covering the anterior of the eye becomes opacified. Currently corneal transplant surgery presents a viable treatment for some patients, but is limited by the availability of human donor corneas. An alternative is the artificial cornea. However, the artificial corneas currently available on market show poor cellular integration with host cells which result in extensive post-operative complications. In order for the artificial cornea to be effective, the implanted synthetic biomaterial such as a polyurethane polymer must integrate well with host cells to minimize post-operative infection and inflammation. The polymer-cell integration process involves a combination of dynamic processes like cell attachment and migration, therefore identifying the structural and dynamic characteristics of the polymer-cell integration process is essential in designing a polymer with highly biocompatible properties. Current imaging techniques like Scanning Electron Microscopy (SEM) may provide structural details but cannot provide direct observation of cellular dynamics.  Multiphoton microscopy however is not limited by the aforementioned restriction. In this project, multiphoton microscopy was used as a tool to probe the dynamic interactions between corneal epithelial cells and polyurethane. To accomplish this goal, corneal epithelial cells were cultured then fluorescently labeled  and seeded onto polyurethane to evaluate the extent of cell proliferation, migration and division during imaging.