2021 MSU Virtual Biomedical Imaging Symposium

Research Summaries

Nanoparticle-based contrast agents for intraductal imaging and preventive treatment of breast cancer, Lorenzo Sempere
SempereLab@MSU is interested in developing image-guided local, less invasive treatments for primary prevention of breast cancer.  We previously determined that Tantalum oxide nanoparticles (synthesized by ShapiroLab@MSU) performs better than FDA-approved iodine-containing ISOVUE® for visualizing the initial filling of the ductal tree. In this project, we are conducting a more systematic and longitudinal assessment of Tantalum oxide and commercially available nanoparticle-based blood pool contrast agents. Based on the testing criteria (e.g., initial retention, local clearance, systemic toxicity) and interim data analysis (e.g., microCT imaging, histology), Tantalum oxide performs better than the other contrast agents for ductal tree imaging.  Faster local clearance of Tantalum oxide in the ablative solution compared to PBS control suggests an active process mediated by macrophages and/or other cells types during tissue repair and remodeling.

Sensitivity-enhanced wireless detectors for   imaging   and physiological sensing, Chunqi Qian
Electroencephalography (EEG) is widely used to monitor brain activities in real-time, but its limited capability for source localization has precluded the unambiguous explanation of its signal dynamics. We will develop a wirelessly powered dual modal detector that can correlate EEG signals with functional magnetic resonance imaging, thus significantly improving the diagnostic accuracy of EEG. Besides brain monitoring, correlation between electrophysiology and MR imaging will be useful to improve the diagnostic accuracy of electrocardiography and electromyography.

Fusing Imaging and Clinical Information for Improved   Automatic Breast Cancer Detection, Gregory Holste
In this work, we explore methods to fuse information from patient imaging and non-imaging clinical information to improve automatic breast cancer detection from MRI. We propose three multimodal fusion architectures, each of which outperform a unimodal (MR image-only or clinical feature-only) model by AUC. Our results suggest that medical AI researchers should seriously consider what non-imaging features are available to them when building a system that learns from medical images.

Green Lanmodulin-based Reporter: An Intensiometric   Rare Earth Element Biological Sensor, Harvey Lee
Our lab and collaborators have created a strain of methylotrophic bacteria that has the potential to mine Rare Earth Elements (REE’s) from the environment with increased storage capacity. To evaluate potentially REE-rich environments, we also developed from this strain an intensiometric biological sensor that reports the detection of REE’s through a jump in fluorescence levels. Here we discuss the use of this protein as a versatile contrast agent for biomedical imaging applications as well as its use in biomedical waste treatment.

Detecting brain drain: a PET imaging approach to   monitoring brain waste clearance, Kylie Smith
Waste clearance from the brain’s (g)lymphatic system is reduced in ageing and disease and may contribute to neurotoxic protein accumulations common to several neurodegenerative disorders. Still, we have yet to establish a sensitive, quantitative approach to assessing brain waste clearance non-invasively. To address this need, we are developing a translatable PET imaging protocol for tracking brain clearance using 18F-labeled radiotracers. In the future, we plan to compare PET and MR imaging metrics to determine whether either (or both) provide meaningful information about (g)lymphatic anatomy.

Ultrasound Modulated Bioluminescence Tomography, Tianyu Yang
Our project is about the mathematical model of Ultrasound Modulated Bioluminescence Tomography (UMBLT). Specifically, we derived two reconstruction algorithms to recover an isotropic source and proved their convergence under sufficient conditions.

Diurnal rhythms in skeletal muscle oxidative capacity   and microvascular function, Jill Slade
Diurnal rhythms are well established in vascular physiology and have recently gained more attention in skeletal muscle metabolism.

Examining the diurnal rhythm of in-vivo skeletal muscle oxidative capacity and peripheral microvascular function in young healthy physically active adults is initially proposed.  It is possible that the diurnal variation provides a functional range reflecting important physiological flexibility and otherwise could provide insight for optimal function important for the timing of interventions or data collection. 

 Shape Analyses of Brain Regional Structures, David Zhu
Zernike transformation can be applied to systematically characterize shape differences in brain regional structures. It potentially can be used to identify shapes related to fast enlargement of brain lesions such as T2 FLAIR white-matter hyperintensities and tumors, and atrophy in regional brain structures such as hippocampus in Alzheimer’s disease.

Simultaneous PET/MRI with orthogonal contrast agents, Christiane Mallett
We are characterizing tumors using preclinical simultaneous PET (positron emission tomography) and MRI (magnetic resonance imaging). Mice were induced with prostate cancer or breast cancer tumors, and then co-injected with 18F-FDG (a glucose analog that has uptake related to tumor metabolism) and Gd-DTPA (an MRI contrast agent with uptake related to tumor vascularization) for dynamic PET/MRI imaging. We will fit the PET data to a Patlak model to measure glucose uptake and flux, and compare those measures in regions of interest with high and low Gd-DTPA enhancement in the dynamic MRI.

Tracking embryo movement in the mouse uterus in real time using ex-vivo multiphoton microscopyManoj Madhavan
Embryo movement in the uterus to find an ideal site of attachment is crucial for successful pregnancy. While it is known that both uterine and embryonic components are important for embryo movement, there is still a critical need to develop a real time imaging system where different factors affecting embryo movement can be phenotypically screened. Hence, we are currently developing an ex vivo multiphoton imaging system to track the movement of embryos in the uterus and to identify maternal and embryonic factors that aid in embryo movement. Embryo sized beads will be transferred into a pre-pubertal uterus, maintained ex vivo in a dish, and the movement of beads will be recorded using a multiphoton imaging system with a resonant scanner. Pre-pubertal mice have thinner muscle layers around the epithelium, which allows light to traverse the tissue allowing us to image and track the movement of the beads in the uterine lumen. The beads will be coated with different biological agents, to screen for molecular signals emitted by embryo that aid in movement. To understand the role of uterine contractions in embryo movement, the uterus will be treated with different agents that induce contractions and relaxations. A better understanding of these maternal-fetal components will open doors for their regulation and to modulate pre-implantation embryo movement in the clinic, for women with recurrent pregnancy loss due to inappropriate site of attachment, or for women needing artificial reproductive technologies including in vitro fertilization (IVF) and embryo transfer.