Radiation therapy (RT) is a cornerstone in cancer treatment (used in 50% of cases), yet challenges persist because damage to normal tissue through direct impact of radiation or bystander effects is inevitable. Injury of macrovessels by RT manifests as obstructive disease, which is akin to atherosclerotic disease. Historically observed in coronary arteries of patients treated for breast cancer and lymphoma, it also affects patients receiving contemporary therapy for lung and chest cancers. Moreover, radiation at various sites can lead to peripheral vascular disease. An aspect of radiation-induced injury that has received little attention is microvascular injury, which typically results from damage to the endothelium and is considered the primary driver of RT-induced toxicity in the skin, kidney, and brain. This review delves into the clinical manifestations of RT-induced vascular disease, signaling pathways, cellular targets affected by radiation injury, and preclinical models of RT-induced vascular injury. The goal is to inspire the development of innovative strategies to prevent RT-related cardiovascular disease.

Retinal image registration is essential for monitoring eye diseases and planning treatments, yet it remains challenging due to large deformations, minimal overlap, and varying image quality. To address these challenges, we propose RetinaRegNet, a multi-stage image registration model with zero-shot generalizability across multiple retinal imaging modalities. RetinaRegNet begins by extracting image features using a pretrained latent diffusion model. Feature points are sampled from the fixed image using a combination of the SIFT algorithm and random sampling. For each sampled point, its corresponding point in the moving image is estimated by cosine similarities between diffusion feature vectors of that point and all pixels in the moving image. Outliers in point correspondences are detected by an inverse consistency constraint, ensuring consistency in both forward and backward directions. Outliers with large distances between true and estimated points are further removed by a transformation-based outlier detector. The resulting point correspondences are then used to estimate a geometric transformation between the two images. We use a two-stage registration framework for robust and accurate alignment: the first stage estimates a homography for global alignment, and the second stage estimates a third-order polynomial transformation to capture local deformations. We evaluated RetinaRegNet on three imaging modalities: color fundus, fluorescein angiography, and laser speckle flowgraphy. Across these datasets, it consistently outperformed state-of-the-art methods, achieving AUC scores of 0.901, 0.868, and 0.861, respectively. RetinaRegNet's zero-shot performance highlights its potential as a valuable tool for tracking disease progression and evaluating treatment efficacy. Our code is publicly available at: https://github.com/mirthAI/RetinaRegNet.

The MR transverse relaxation rate, $R_2^{*}$, has been widely used to detect iron and myelin content in tissue. However, it is also sensitive to macroscopic B₀ inhomogeneities. One approach to correct for the B₀ effect is to fit gradient-echo signals with the three-parameter model, a sinc function-weighted monoexponential decay. However, such three-parameter models are subject to increased noise sensitivity. To address this issue, this study presents a two-stage fitting procedure based on the three-parameter model to mitigate the B₀ effect and reduce the noise sensitivity of $R_2^{*}$ measurement in the mouse brain at 7T. MRI scans were performed on eight healthy mice. The gradient-echo signals were fitted with the two-stage fitting procedure to generate $R_{2corr\_t}^{*}$. The signals were also fitted with the monoexponential and three-parameter models to generate $R_{2nocorr}^{*}$ and $R_{2corr}^{*}$, respectively. Regions of interest (ROIs), including the corpus callosum, internal capsule, somatosensory cortex, caudo-putamen, thalamus, and lateral ventricle, were selected to evaluate the within-ROI mean and standard deviation (SD) of the $R_2^{*}$ measurements. The results showed that the Akaike information criterion of the monoexponential model was significantly reduced by using the three-parameter model in the selected ROIs (p = 0.0039–0.0078). However, the within-ROI SD of $R_{2corr}^{*}$ using the three-parameter model was significantly higher than that of the $R_{2nocorr}^{*}$ in the internal capsule, caudo-putamen, and thalamus regions (p = 0.0039), a consequence partially due to the increased noise sensitivity of the three-parameter model. With the two-stage fitting procedure, the within-ROI SD of $R_{2corr}^{*}$ was significantly reduced by 7.7–30.2% in all ROIs, except for the somatosensory cortex region with a fast in-plane variation of the B₀ gradient field (p = 0.0039–0.0078). These results support the utilization of the two-stage fitting procedure to mitigate the B₀ effect and reduce noise sensitivity for $R_2^{*}$ measurement in the mouse brain.

Background/Objectives: Radiation therapy can often lead to structural and functional changes in the brain resulting in radiation-induced brain injury. This study investigates the MRI-detectable effects of whole-brain irradiation across all neuroanatomical structures in adult mice, with a specific focus on T2* MRI measurements, to evaluate regions that may be particularly sensitive to iron accumulation. Methods: One year following irradiation or sham treatment, mice were imaged with a 7T MRI to evaluate changes in regional volume and T2* relaxation times across more than 652 neuroanatomical using the DSURQE mouse brain atlas. Results: Statistical analysis identified 301 altered regions with respect to regional volume and 85 regions with respect to T2* relaxation showing significant differences relative to the control group (p < 0.05). Further data refinement, including the consolidation of redundant, bi-lateral structures revealed 18 subregions with significant changes in both volume and T2*. The data refinement revealed that the most represented system was the olfactory system (8/18 regions, 44%). The olfactory regions also showed the most pronounced changes and greatest correlation between the two metrics. Conclusions: These findings are suggestive that ionizing radiation may cause a pronounced disruption in the olfactory system that coincides with potential iron accumulation.

Over half a million Americans receive radiation therapy as part of their cancer treatment each year. Within several years of their treatment, many of these patients will present with clinically-significant disorders, such as cardiomyopathy, pulmonary fibrosis, and cognitive decline. These injuries are classified as “normal tissue injury” due to their resulting from radiation exposure to the healthy tissues surrounding the targeted cancer tissue. Acute radiation damage to the microvascular endothelium (termed “radiation microvasculopathy”) is thought to eventually lead to a loss of capillaries, tissue ischemia, and, consequently, a loss of normal tissue function. Portions of this proposed sequence have been observed in vivo in preclinical models like rabbits and miniature pigs; however, to date, clinical data have been limited to descriptive studies of histopathology specimens or indirect indicators of poor perfusion or ischemia. While early perfusion defects and late loss of capillaries have been observed clinically using methods like laser Doppler and single photon emission computed tomography (SPECT), the use of radiotracers and radiographic imaging have made serial measurements within a single individual over time impractical. Furthermore, noninvasive methods capable of visualizing the microvessels with high enough resolution to simultaneously assess changes in both structure and function are currently missing. To effectively understand normal tissue injury development, especially the role of radiation microvasculopathy, new imaging modalities capable of assessing microvascular structure and function are needed. The purpose of this thesis work was to obtain an in-depth characterization of radiation microvasculopathy and its related pathologies in patients treated with radiotherapy for uveal melanoma, the most common primary intraocular cancer in adults. A cross-sectional cohort of 25 patients was assessed to test the hypothesis that radiation damage to the microvasculature precedes a loss of capillaries, neural tissue, and visual function in a dose- and time-dependent manner. Evidence for damage to the microvascular endothelium was identified by detecting vessel leakage and fluid accumulation in the retinal layers within the first 2 years from radiotherapy. During this same time period, decreases in blood flow compared to the contralateral, nonirradiated eye were observed. Importantly, these decreases in flow were observed before a loss of capillaries was detected. These early changes in blood flow and retinal layer thickness were followed by a loss of capillaries, thinning of the retinal tissue, and loss of visual function, indicative of ischemia and loss of normal tissue function. To enable studies of radiation microvasculopathy in mice, complementary methods for detecting changes in retinal blood flow were established in a C57/Bl6J mouse model, in both normal control mice and in mice with angiotensin-II-induced hypertension. Additionally, methods for measuring blood flow in mice exposed to a light flicker stimulus, used to assess vasodilation during neurovascular coupling, were established to enable the development of a preclinical model of radiation microvasculopathy in mice. Together, these data suggest hemodynamic changes can be detected before a loss of capillaries and normal tissue function, and indicate a need for a longitudinal study to confirm this sequence of normal tissue injury development.