Dysphagia in Head and Neck Radiotherapy: The Influence of Pharyngeal Constrictor Anatomy and Dosimetry.

The goal of this study was to identify which anatomical and dosimetric changes correlated with late patient-reported dysphagia throughout the course of head and neck chemo-radiotherapy treatment. The patient cohort (n = 64) considered oropharyngeal and nasopharyngeal patients treated with curative intent, exhibiting no baseline dysphagia with a follow-up time greater than one year. Patients completed the MD Anderson Dysphagia Inventory during a follow-up visit. A composite score was measured ranging from 20 to 100, with a low score indicating a high symptom burden; a score ≤60 indicated patient-reported dysphagia. The pharyngeal (PCM) and cricopharyngeal constrictor muscles (CPM) were contoured on a planning CT image and adapted to weekly cone-beam CT anatomy using deformable image registration and dose was accumulated using weighted dose-volume histogram curves. The PCM and CPM were examined for volume, thickness, and dosimetric changes across treatment with the results correlated to symptom group. Anatomical evaluation indicated the PCM thickness increased more during treatment for patients with dysphagia, with base of C2 vertebrae (p = 0.04) and superior-inferior middle PCM (p = 0.01) thicknesses indicating a 1.0-1.5 mm increase. The planned and delivered mean dose and DVH metrics to PCM and CPM were found to be within random error measured for the dose accumulation, indicating delivered and planned dose are equivalent. The PCM and CPM organs were found to lie approximately 5 mm closer to high dose gradients in patients exhibiting dysphagia. The volume, thickness, and high dose gradient metrics may be useful metrics to identify patients at risk of late patient-reported dysphagia.

Developing and validating multi-omics prediction models for late patient-reported dysphagia in head and neck radiotherapy.

Background and purpose. To investigate models developed using radiomic and dosiomic (multi-omics) features from planning and treatment imaging for late patient-reported dysphagia in head and neck radiotherapy.Materials and methods. Training (n = 64) and testing (n = 23) cohorts of head and neck cancer patients treated with curative intent chemo-radiotherapy with a follow-up time greater than 12 months were retrospectively examined. Patients completed the MD Anderson Dysphagia Inventory and a composite score ≤60 was interpreted as patient-reported dysphagia. A chart review collected baseline dysphagia and clinical factors. Multi-omic features were extracted from planning and last synthetic CT images using the pharyngeal constrictor muscle contours as a region of interest. Late patient-reported dysphagia models were developed using a random forest backbone, with feature selection and up-sampling methods to account for the imbalanced data. Models were developed and validated for multi-omic feature combinations for both timepoints.Results. A clinical and radiomic feature model developed using the planning CT achieved good performance (validation: sensitivity = 80 ± 27% / balanced accuracy = 71 ± 23%, testing: sensitivity = 80 ± 10% / balanced accuracy = 73 ± 11%). The synthetic CT models did not show improvement over the plan CT multi-omics models, with poor reliability of the radiomic features on these images. Dosiomic features extracted from the synthetic CT showed promise in predicting late patient-reported dysphagia.Conclusion. Multi-omics models can predict late patient-reported dysphagia in head and neck radiotherapy patients. Synthetic CT dosiomic features show promise in developing successful models to account for changes in delivered dose distribution. Multi-center or prospective studies are required prior to clinical implementation of these models.

Can machine learning models improve early detection of brain metastases using diffusion weighted imaging-based radiomics?

Background: Metastatic complications are a major cause of cancer-related morbidity, with up to 40% of cancer patients experiencing at least one brain metastasis. Earlier detection may significantly improve patient outcomes and overall survival. We investigated machine learning (ML) models for early detection of brain metastases based on diffusion weighted imaging (DWI) radiomics. Methods: Longitudinal diffusion imaging from 116 patients previously treated with stereotactic radiosurgery (SRS) for brain metastases were retrospectively analyzed. Clinical contours from 600 metastases were extracted from radiosurgery planning computed tomography, and rigidly registered to corresponding contrast enhanced-T1 and apparent diffusion coefficient (ADC) maps. Contralateral contours located in healthy brain tissue were used as control. The dataset consisted of (I) radiomic features using ADC maps, (II) radiomic feature change calculated using timepoints before the metastasis manifested on contrast enhanced-T1, (III) primary cancer, and (IV) anatomical location. The dataset was divided into training and internal validation sets using an 80/20 split with stratification. Four classification algorithms [Linear Support Vector Machine (SVM), Random Forest (RF), AdaBoost, and XGBoost] underwent supervised classification training, with contours labeled either 'control' or 'metastasis'. Hyperparameters were optimized towards balanced accuracy. Various model metrics (receiver operating characteristic curve area scores, accuracy, recall, and precision) were calculated to gauge performance. Results: The radiomic and clinical data set, feature engineering, and ML models developed were able to identify metastases with an accuracy of up to 87.7% on the training set, and 85.8% on an unseen test set. XGBoost and RF showed superior accuracy (XGBoost: 0.877±0.021 and 0.833±0.47, RF: 0.823±0.024 and 0.858±0.045) for training and validation sets, respectively. XGBoost and RF also showed strong area under the receiver operating characteristic curve (AUC) performance on the validation set (0.910±0.037 and 0.922±0.034, respectively). AdaBoost performed slightly lower in all metrics. SVM model generalized poorly with the internal validation set. Important features involved changes in radiomics months before manifesting on contrast enhanced-T1. Conclusions: The proposed models using diffusion-based radiomics showed encouraging results in differentiating healthy brain tissue from metastases using clinical imaging data. These findings suggest that longitudinal diffusion imaging and ML may help improve patient care through earlier diagnosis and increased patient monitoring/follow-up. Future work aims to improve model classification metrics, robustness, user-interface, and clinical applicability.

Machine learning-generated decision boundaries for prediction and exploration of patient-specific quality assurance failures in stereotactic radiosurgery plans.

INTRODUCTION: Stereotactic radiosurgery (SRS) is a form of radiotherapy treatment during which high radiation dose is delivered in a single or few fractions. These treatments require highly conformal plans with steep dose gradients, which can result in an increase in plan complexity prompting the need for stringent pretreatment patient-specific quality assurance (QA) measurements to ensure the planned and measured dose distributions agree within clinical standards. Complexity scores and machine learning (ML) techniques may help with prediction of QA outcomes; however interpretability and usability of those results continues to be an area of study. This study investigates the use of plan complexity metrics as input for an ML model to allow for prediction of QA outcomes for SRS plans as measured via three-dimension (3D) phantom dose verification. Explorations into interpretability and predictive ability, as well as a prospective in-clinic implementation using the resulting model were performed. METHODS: Four hundred ninety-eight plans (1571 volumetric modulated arc therapy arcs) were processed via in-house script to generate several complexity scores. 3D phantom dose verification measurement results were extracted and classified as pass or failure (with failures defined as below 95% voxel agreement passing 3%/1-mm gamma criteria with 10% threshold,) and 1472 of the arcs were split into training and testing sets, with 99 arcs as a sequential holdout set. A z-score scaler was trained on the training set and used to scale all other sets. Variations of multi-leaf collimator (MLC) leaf movement variability, aperture complexity, and leaf size, and monitor unit (MU) at control point weighted target area scores were used as input to a support vector classifier to generate a series of 1D, 2D, and 5D decision boundaries. The best performing 5D model was then used within a prospective in-clinic study providing predictions to physicists prior to ordering 3D phantom dose verification measurements for 38 patient plans (112 arcs). The decision to order 3D phantom dose verification measurements was recorded before and after prediction. RESULTS: Best performing 1D threshold and 2D prediction models with best performance produced a QA failure recall and QA passing recall of 1.00 and 0.55, and 0.82 and 0.82, respectively. Best performing 5D prediction model produced a QA failure recall (sensitivity) of 1.00 and QA passing recall (specificity) of 0.72. This model was then used within a prospective in-clinic study providing predictions to physicists prior to ordering 3D phantom dose verification measurements and achieved a QA failure recall of 1.00 and QA passing recall of 0.58. The decision to order 3D phantom dose verification measurements was recorded before and after measurement. A single initially unidentified failing plan of the prospective cohort was successfully predicted to fail by the model. CONCLUSION: Implementation of complexity score-based prediction models for SRS would allow for support of a clinician's decision to reduce time spent performing QA measurements and avoid patient treatment delays (i.e., in case of QA failure).

Investigating diffusion patterns of brain metastases pre- and post-stereotactic radiosurgery: a feasibility study.

Purpose.Metastatic complications are responsible for 90% of cancer-associated mortality. Magnetic resonance imaging (MRI) can be used to observe the brain's microstructure and potentially correlate changes with metastasis occurrence. Diffusion weighted imaging (DWI) is an MRI technique that utilizes the kinetics of water molecules within the body. The aim of this study is to use DWI to characterize diffusion changes within brain metastases in cancer patients pre- and post-stereotactic radiosurgery (SRS).Methods.We retrospectively analyzed 113 metastases from 13 patients who underwent SRS for brain metastasis recurrence. Longitudinal apparent diffusion coefficient (ADC) maps were registered to Gd-T1 images and CT, and clinical metastasis ROIs from all SRS treatments were retrospectively transferred onto these ADC maps for analysis. Metastases were characterized based on pre-SRS diffusion pattern, primary cancer site, and post-SRS outcome. ADC values were calculated pre- and post-SRS.Results.ADC values were significantly elevated (980.2 × 10-6mm2s-1and 1040.3 × 10-6mm2s-1pre- and post-SRS, respectively) when compared to healthy brain tissue (826.8 × 10-6mm2s-1) for all metastases. Three identified pre-SRS patterns were significantly different before SRS and within 6 months post-SRS. No significant differences were observed between different primaries pre-SRS. Post-SRS, Lung metastases ADC decreased by 86.2 × 10-6mm2s-1, breast metastases increased by 116.7 × 10-6mm2s-1, and genitourinary metastases showed no significant ADC change. SRS outcomes showed ADC variability pre-treatment but no significant differences pre- and post-SRS, except at 6-9 months post-SRS where progressing metastases were elevated when compared to other response groups.Conclusion. This study provided a unique opportunity to characterize diffusion changes in brain metastases before their manifestation on standard Gd-T1 images and post-SRS. Identified patterns may improve early detection of brain metastases as well as predict their response to treatment.

Cardiovascular risk and encoding-related hippocampal connectivity in older adults.

BACKGROUND: Cardiovascular conditions contribute to brain volume loss, reduced cerebrovascular health, and increased dementia risk in aging adults. Altered hippocampal connectivity has also been observed in individuals with cardiovascular conditions, yet the functional consequences of these changes remain unclear. In the present study, we collected functional magnetic resonance imaging data during memory encoding and used a psychophysiological interaction analysis to examine whether cardiovascular burden, indexed using the Framingham risk score, was associated with encoding-related hippocampal connectivity and task performance in cognitively-intact older adults between 65 and 85 years of age. Our goal was to better understand the early functional consequences of vascular and metabolic dysfunction in those at risk for cognitive decline. RESULTS: High Framingham risk scores were associated with lower total brain volumes. In addition, those with high Framingham risk scores showed an altered relationship between left hippocampal-medial prefrontal coupling and task performance compared to those with low Framingham risk scores. Specifically, we found a significant interaction of Framingham risk and learning on connectivity between the left hippocampus and primarily left midline prefrontal regions comprising the left ventral anterior cingulate cortex and medial prefrontal cortex. Those with lower Framingham risk scores showed a pattern of weaker connectivity between left hippocampal and medial prefrontal regions associated with better task performance. Those with higher Framingham risk scores showed the opposite pattern; stronger connectivity was associated with better performance. CONCLUSIONS: Findings from the current study show that amongst older adults with cardiovascular conditions, higher Framingham risk is associated with lower brain volume and altered left hippocampal-medial prefrontal coupling during task performance compared to those with lower Framingham risk scores. This may reflect a compensatory mechanism in support of memory function and suggests that older adults with elevated cardiovascular risk are vulnerable to early Alzheimer disease-like dysfunction within the episodic memory system.

Brain Function Is Linked to LDL Cholesterol in Older Adults with Cardiovascular Risk.

OBJECTIVES: To determine how cardiovascular risk is associated with working memory task performance and task-related suppression of default-mode network (DMN) activity in cognitively intact older adults. DESIGN: A cross-sectional functional magnetic resonance imaging study of older adults with cardiovascular risk factors. SETTING: Rotman Research Institute, Baycrest Health Sciences. PARTICIPANTS: Thirty older adults with cardiovascular risk factors. MEASUREMENTS: Participants provided health information and a blood sample, and underwent functional magnetic resonance imaging during a working memory task and during a breath-hold task to assess cerebrovascular reactivity. RESULTS: Higher plasma low-density lipoprotein cholesterol (LDL-C) was associated with poorer working memory task performance (P = 0.008) and reduced task-related DMN suppression (P = 0.005). A composite index of cardiovascular risk, the Framingham General Cardiovascular Risk Profile, showed no associations with task performance or task-related DMN suppression. These findings were independent of white matter burden and cerebrovascular reactivity and thus cannot be accounted for by individual differences in neurovascular health. CONCLUSION: These findings suggest a deleterious effect of elevated LDL-C on working memory task performance and task-related DMN suppression in older adults with cardiovascular risk. The relations between the Framingham General Cardiovascular Risk Profile, cognitive task performance, and DMN function require further study.

Summative effects of vascular risk factors on cortical thickness in mild cognitive impairment.

Vascular risk factors (VRFs) increase the risk of Alzheimer's disease (AD) and contribute to neurodegenerative processes. The purpose of this study was to investigate whether increasing number of VRFs contributes to within-cohort differences in cortical thickness (CThk) among adults with mild cognitive impairment (MCI) and cognitively intact older controls from the AD Neuroimaging Initiative 1, GO, and 2 data sets. Multivariate partial least squares analysis was used to investigate the effect of VRF index on regional CThk measurements, which produced a significant latent variable and identified patterns of cortical thinning in the MCI group but not controls. Subsequent analyses tested the interaction effects between VRF index and cognitive grouping and examined 1-year follow-up data. There was evidence of a VRF index by cognitive group interaction. Partial least squares results were replicated at 1-year follow-up among MCI cohort in a subset of baseline CThk regions. This study provides evidence that a summative VRF index accounts for some of the variance in brain tissue loss in regions implicated in AD among MCI adults.

Vascular risk factor burden correlates with cerebrovascular reactivity but not resting state coactivation in the default mode network.

PURPOSE: White matter hyperintensities (WMH) are prevalent among older adults and are often associated with cognitive decline and increased risk of stroke and dementia. Vascular risk factors (VRFs) are linked to WMH, yet the impact of multiple VRFs on gray matter function is still unclear. The goal of this study was to test for associations between the number of VRFs and cerebrovascular reactivity (CVR) and resting state (RS) coactivation among individuals with WMH. MATERIALS AND METHODS: Twenty-nine participants with suspected WMH were grouped based on the number of VRFs (subgroups: 0, 1, or ≥2). CVR and RS coactivation were measured with blood oxygenation level-dependent (BOLD) imaging on a 3T magnetic resonance imaging (MRI) system during hypercapnia and rest, respectively. Default-mode (DMN), sensory-motor, and medial-visual networks, generated using independent component analysis of RS-BOLD, were selected as networks of interest (NOIs). CVR-BOLD was analyzed using two methods: 1) a model-based approach using CO2 traces, and 2) a dual-regression (DR) approach using NOIs as spatial inputs. Average CVR and RS coactivations within NOIs were compared between VRF subgroups. A secondary analysis investigated the correlation between CVR and RS coactivation. RESULTS: VRF subgroup differences were detected using DR-based CVR in the DMN (F20,2 = 5.17, P = 0.015) but not the model-based CVR nor RS coactivation. DR-based CVR was correlated with RS coactivation in the DMN (r(2) = 0.28, P = 0.006) but not the sensory-motor nor medial-visual NOIs. CONCLUSION: In individuals with WMH, CVR in the DMN was inversely associated with the number of VRFs and correlated with RS coactivation.

A systematic review of type 2 diabetes mellitus and hypertension in imaging studies of cognitive aging: time to establish new norms.

The rising prevalence of type 2 diabetes (T2DM) and hypertension in older adults, and the deleterious effect of these conditions on cerebrovascular and brain health, is creating a growing discrepancy between the "typical" cognitive aging trajectory and a "healthy" cognitive aging trajectory. These changing health demographics make T2DM and hypertension important topics of study in their own right, and warrant attention from the perspective of cognitive aging neuroimaging research. Specifically, interpretation of individual or group differences in blood oxygenation level dependent magnetic resonance imaging (BOLD MRI) or positron emission tomography (PET H2O(15)) signals as reflective of differences in neural activation underlying a cognitive operation of interest requires assumptions of intact vascular health amongst the study participants. Without adequate screening, inclusion of individuals with T2DM or hypertension in "healthy" samples may introduce unwanted variability and bias to brain and/or cognitive measures, and increase potential for error. We conducted a systematic review of the cognitive aging neuroimaging literature to document the extent to which researchers account for these conditions. Of the 232 studies selected for review, few explicitly excluded individuals with T2DM (9%) or hypertension (13%). A large portion had exclusion criteria that made it difficult to determine whether T2DM or hypertension were excluded (44 and 37%), and many did not mention any selection criteria related to T2DM or hypertension (34 and 22%). Of all the surveyed studies, only 29% acknowledged or addressed the potential influence of intersubject vascular variability on the measured BOLD or PET signals. To reinforce the notion that individuals with T2DM and hypertension should not be overlooked as a potential source of bias, we also provide an overview of metabolic and vascular changes associated with T2DM and hypertension, as they relate to cerebrovascular and brain health.

Combined effects of type 2 diabetes and hypertension associated with cortical thinning and impaired cerebrovascular reactivity relative to hypertension alone in older adults.

OBJECTIVE: Type 2 diabetes mellitus is characterized by metabolic dysregulation in the form of hyperglycemia and insulin resistance and can have a profound impact on brain structure and vasculature. The primary aim of this study was to identify brain regions where the combined effects of type 2 diabetes and hypertension on brain health exceed those of hypertension alone. A secondary objective was to test whether vascular impairment and structural brain measures in this population are associated with cognitive function. RESEARCH DESIGN AND METHODS: We enrolled 18 diabetic participants with hypertension (HTN + T2DM, 7 women, 71.8 ± 5.6 years) and 22 participants with hypertension only (HTN, 12 women, 73.4 ± 6.2 years). Cerebrovascular reactivity (CVR) was assessed using blood oxygenation level dependent (BOLD) MRI during successive breath holds. Gray matter structure was evaluated using cortical thickness (CThk) measures estimated from T1-weighted images. Analyses of cognitive and blood data were also performed. RESULTS: Compared to HTN, HTN + T2DM had decreased CVR and CThk in a spatially overlapping region of the right occipital lobe (P < 0.025); CVR group differences were more expansive and included bilateral occipito-parietal areas (P < 0.025). Whereas CVR showed no significant associations with measures of cognitive function (P > 0.05), CThk in the right lingual gyrus ROI and regions resulting from a vertex-wise analysis (including posterior cingulate, precuneus, superior and middle frontal, and middle and inferior temporal regions (P < 0.025) were associated with executive function. CONCLUSIONS: Individuals with T2DM and HTN showed decreased CVR and CThk compared to age-matched HTN controls. This study identifies brain regions that are impacted by the combined effects of comorbid T2DM and HTN conditions, with new evidence that the corresponding cortical thinning may contribute to cognitive decline.

Evaluation of a new commercial Monte Carlo dose calculation algorithm for electron beams.

PURPOSE: In this report the authors present the validation of a Monte Carlo dose calculation algorithm (XiO EMC from Elekta Software) for electron beams. METHODS: Calculated and measured dose distributions were compared for homogeneous water phantoms and for a 3D heterogeneous phantom meant to approximate the geometry of a trachea and spine. Comparisons of measurements and calculated data were performed using 2D and 3D gamma index dose comparison metrics. RESULTS: Measured outputs agree with calculated values within estimated uncertainties for standard and extended SSDs for open applicators, and for cutouts, with the exception of the 17 MeV electron beam at extended SSD for cutout sizes smaller than 5 × 5 cm(2). Good agreement was obtained between calculated and experimental depth dose curves and dose profiles (minimum number of measurements that pass a 2%/2 mm agreement 2D gamma index criteria for any applicator or energy was 97%). Dose calculations in a heterogeneous phantom agree with radiochromic film measurements (>98% of pixels pass a 3 dimensional 3%/2 mm γ-criteria) provided that the steep dose gradient in the depth direction is considered. CONCLUSIONS: Clinically acceptable agreement (at the 2%/2 mm level) between the measurements and calculated data for measurements in water are obtained for this dose calculation algorithm. Radiochromic film is a useful tool to evaluate the accuracy of electron MC treatment planning systems in heterogeneous media.