Duke University

The growing adoption of Digital Twin (DT) technologies in healthcare highlights the need to define how medical imaging tools (physical and digital phantoms, real and simulated CT studies, etc.) integrate with the virtual components of DTs. A recently introduc...

Limited-angle cone-beam CT (CBCT) reconstruction suffers from missing projection data, leading to severe streak artifacts, structural distortions, and degraded image quality. This study proposes a conditional diffusion-based projection extrapolation framework...

To validate plan-delivery log file-based 3D dose calculations with ArcCHECK (AC) 3DVH reconstruction for single-isocenter multiple-target (SIMT) brain plans

Diagnostic and Interventional Radiology Physics

Organ‑level breast dosimetry is essential for developing robust dose-response models of subsequent breast cancer (SBC) risk in female childhood cancer survivors. Historically, epidemiologic studies relied on surrogate metrics such as prescription dose due to...

Glioblastoma has a low median survival of 3cm from the tumor, where deformation is minimal. Vessels 80% of patients. Future work will incorporate this pathway into a full front-to-end deformable registration approach to precisely determine the origin of recur...

To develop a pixel-level uncertainty-aware consistency (UA-Cons) learning framework to optimize the feature compensation behavior of deep neural networks in scenarios where multi-parametric MRI modalities are incomplete.

Therapy Physics

Diagnostic and Interventional Radiology Physics

Mechanistic tumor growth models provide a framework for linking observed tumor kinetics to underlying biological processes; however, their application to real preclinical datasets is limited by incomplete longitudinal measurements and uncertainty in model par...

To propose a federated learning (FL) framework incorporating a novel deep ensemble strategy for multi-institutional brain metastasis (BM) segmentation, improving performance in limited local datasets while preserving privacy by avoiding large-scale data trans...

Standardizing the assessment of the Tumor Microenvironment (TME) is critical for personalizing therapy in Head and Neck Squamous Cell Carcinoma (HNSCC). As part of a larger multiomics effort to discover multiscale biomarkers, we developed a mathematical frame...

Effective radiotherapy for upper abdominal tumors requires dose escalation but is limited by respiratory motion and the low gastrointestinal radiation tolerance. Current clinical motion management on conventional Linacs relies on simple external surrogates th...

Studies show measurable changes in brain metastasis over 1-2 weeks, with delays between MRI and radiosurgery delivery reducing local control. We aim to minimize this gap by developing a multi-target radiosurgery workflow that eliminates need for CT simulation...

Mobile CT and CBCT systems improve flexibility and reduce patient transport burden in HDR gynecologic brachytherapy. However, competition for imaging resources remains a practical limitation. This work evaluates the Varian Ethos linear accelerator equipped wi...

To quantitatively benchmark dosimetric variation associated with tumor regression during head and neck (HN) radiotherapy and to evaluate the benefit of adaptive replanning as a basis for adapt-on-demand decision support.

Conventional filtered backprojection with a fixed Ram-Lak filter in cone-beam CT (CBCT) reconstruction method often amplifies noise and streak artifacts under sparse-view acquisition, limiting image quality for image-guided radiotherapy. This study investigat...

Understanding chemoradiation resistance remains a major barrier to improving outcomes in head and neck cancer. We developed a multiscale framework linking tumor architecture, cellular network topology, and immune context to therapeutic response. This approach...

The X-ray Heel Effect influences automatic exposure control (AEC) termination at different AEC cell locations. This study evaluates the Heel Effect’s influence on AEC cell balance testing methodologies in digital radiography.

While reinforcement-learning (RL) based needle placement optimization has been explored in prostate brachytherapy, its application to gynecologic high dose rate (HDR) brachytherapy remains limited. This work presents a reinforcement learning based framework f...

Augment immunogenic photothermal therapy (PTT) with a novel near-infrared (NIR) plasmonic nanoparticle platform to investigate photothermal immunotherapy treatment schemes. By leveraging the photothermal stability and biocompatibility of armored-core gold nan...

Monte Carlo simulations are widely used to derive effective dose conversion factors (k-factors) for CT; however, acquisition parameters such as helical pitch and tube starting angle are often randomly selected. The impact of these parameters on the variabilit...

To develop an in-silico tumor model that incorporates nutrient-driven growth and radiotherapy response to generate spatio-temporal proliferating (P), quiescent (Q), and necrotic (N) cell maps for radiomics-based heterogeneity analysis.

To investigate the suitability of three different materials as background media for use with in-house phantoms for HDR interstitial brachytherapy. In-house phantoms are designed to supplement commercial phantom capabilities and enable interstitial needle plac...

SRS MapCheck has been widely used for SRS pretreatment patient-specific quality assurance (PSQA); however, it requires multiple machine deliveries due to limited array size. To improve the pretreatment PSQA efficiency for Single Isocenter Multiple Target (SIM...

Radiation induced bowel toxicity is challenging for abdominal SBRT due to proximity to PTV and high dose/fraction. Remote magnetic navigation systems (RMNS) have been developed for magnetic capsule endoscopy, in which the RMNS mounted on a robotic arm control...

High-pitch dual-source acquisition (“FLASH”) with tin spectral filtration has substantially reduced radiation dose while minimizing patient motion, both of which are particularly important for pediatric imaging. However, in adults, tin-filtered spectra have b...

Personalized oncology requires bridging the gap between macro-scale imaging and micro-scale cellular architecture. We developed a radiopathomic fusion framework to integrate PET/CT metabolic heterogeneity with graph-based topological modeling of the tumor mic...

To evaluate the uses of beryllium oxide optically stimulated luminescent dosimeters (OSLDs) and solid-state micro-diamond detectors as viable dosimetry tools for ultra-high dose rate electron FLASH treatments.

Low grade diffuse gliomas are slow-growing tumors that must be monitored longitudinally to detect progression to more aggressive higher-grade disease. However, the slow growth of these tumors can make it difficult for radiologists to appreciate subtle tumor c...

Head and neck squamous cell carcinoma (HNSCC) is a clinically aggressive malignancy with heterogeneous responses to chemoradiation, driven in part by dynamic immune remodeling. While immune infiltration has been studied at different time points, the spatiotem...

Murine models of head and neck squamous cell carcinoma (HNSCC) are routinely monitored using external caliper measurements to estimate tumor volume, most often relying on simplified geometric assumptions such as cylindrical approximations. While convenient, t...

To develop a multi-parametric MRI (mp-MRI) radiomics framework for predicting post-resection glioblastoma (GBM) survival by integrating conventional MR modalities with a quantum mechanics–inspired imaging representation.

Daily cone-beam CT (CBCT) is widely used in adaptive proton therapy; however, scatter artifacts can degrade image quality and introduce proton dose calculation inaccuracies. We developed a region-of-interest (ROI)–guided Swin-Transformer deep learning (DL) ne...

This study aimed to evaluate the timing of incident occurrence and detection across the radiation oncology workflow to identify error-prone workflow processes and to assess the effectiveness of quality assurance (QA) processes. The findings provide insights t...

Reinforcement learning (RL) constitutes a strong candidate for AI-guided treatment planning for two distinguishing reasons: it differs from greedy algorithms by optimizing strategy over the complete history of a Markovian process; and it contrasts with superv...

Discrepancy between the diaphragm position in PET and CT during PET/CT studies can lead to artifacts in the PET images due to improper attenuation correction. This artifact is an artificially depleted region at the liver dome that can potentially obscure lesi...

The novel treatment delivery technique RapidArc Dynamic (RAD), which combines dynamic gantry motion with static angle–modulated ports (STAMPs), offers strong potential for advanced planning and improved plan quality. However, a major challenge in clinical imp...

To retrospectively identify potential x-ray anode overheating events and to prospectively design protocols and policies to mitigate heat-related risks that can lead to tube downtime and workflow disruption.

To quantify how SAR-constrained saturation design, Dixon water–fat separation, and distortion/motion correction affect quantitative abdominal APT/CEST using a time-efficient (seg-)EPI acquisition, and to define a robust operating range for water-only APT mapp...

To develop and evaluate a federated learning (FL) framework for brain metastasis (BM) segmentation that integrates an uncertainty score into a novel FL objective, improving segmentation robustness and potentially performance when training on limited-size data...

VMAT optimization is a non-convex problem with tightly coupled parameters and machine constraints, which limits the development of transparent and extensible frameworks outside commercial treatment planning systems. This work introduces a new perspective on V...

General radiography lacks a clinically relevant, vendor-agnostic metric to indicate for-presentation image quality across anatomical views. While exposure index is commonly used to indicate exposure appropriateness, it is biased by vendor-specific region-of-i...

To evaluate whether a federated learning (FL) scheme that leverages adult glioma patient data improves multi-parametric MRI (mp-MRI) based pediatric glioma segmentation.

Targeted radiopharmaceuticals have become of widespread interest in the scientific community given their ability to detect and treat metastatic disease. We aim to validate the use of novel chelate-tethered eHSP-90 inhibitors as potential agents for detection...

To enable consistent COPD quantification in multi-center studies across energy-integrating (EICT) and photon-counting (PCCT) CT technologies, using a physics-based image harmonization technique.

Upright patient positioners coupled with diagnostic-quality vertical CTs at treatment isocenter introduce significant opportunities for online adaptive particle therapy. However, limited upright data constrains our ability to rigorously test and optimize this...

Image noise is a fundamental determinant of CT image quality and directly impacts diagnostic performance. Conventional noise metrics are typically global and do not account for spatial or organ-specific variations introduced by modern CT techniques such as tu...

Therapy Physics

Automated segmentation of lung nodules in chest CT is critical for early cancer screening but remains challenging due to the small size and variable morphology of nodules, which often resemble vessels or pleura. This study proposes a novel framework integrati...

To study the cardiac and respiratory heart motion together. Cardiorespiratory motion management is a significant topic in stereotactic arrhythmia radiotherapy (STAR). Cardiorespiratory motion consists of the heart beating and respiration. Commonly studied sep...

To update U.S. national adult CT DRLs and ADs using 2025 data and to demonstrate a scalable, acquisition-level, size-normalized framework for generating timely and clinically relevant national benchmarks.

Professional

Our previous work proposed a Neural ODE–based U-Net (NODE-UNet) that generates continuous trajectories to visualize the evolution of feature representations from the initial input to the terminal state. We hypothesize that modeling contextual consistency alon...

It was critical yet impossible to quantitatively assess the impact of patient-specific cardiorespiratory motion on dose plans for stereotactic arrhythmia radioablation (STAR) treatments. To fill the gap, a fast algorithm, SPM (spatial probability map), was de...

Diagnostic and Interventional Radiology Physics

Diagnostic and Interventional Radiology Physics

To advance the clinical application of a fully automatic planning system for functional lung avoidance radiotherapy (AP-FLART) through clinical implementation and comprehensive evaluation.

Radiotherapy for upper abdominal cancers is limited by respiratory motion and the low radiation tolerance, restricting safe dose escalation. Conventional linear accelerators rely on kV X-ray and CBCT imaging but lack real-time internal motion tracking capabil...

Therapy Physics

Therapy Physics

Managing cardiac and respiratory motion is critical for stereotactic arrhythmia radiotherapy (STAR), thoracic, and breast treatments. Breath-hold cardiac 4DCT (c4DCT) is commonly required for assessing cardiac motion, while respiratory 4DCT (r4DCT) is used fo...

Professional

Therapy Physics

Therapy Physics

Professional

RapidArc Dynamic (RAD) is a novel treatment delivery technique that combines rotational arc therapy with static angle modulated ports (STAMPs). This study evaluates the feasibility, dosimetric performance, and delivery efficiency of RAD for single-isocenter m...

Reference dosimetry forms the foundation of accurate dose delivery in radiation therapy. Two primary calibration protocols dominate clinical practice: AAPM’s TG-51, used throughout North America and select countries, and IAEA’s TRS-398, adopted widely across...

To investigate mechanisms of chemoradiation resistance in murine head and neck squamous cell carcinoma using a multiscale framework integrating metabolic imaging (μPET), tissue heterogeneity (μCT), and cellular topology (digital pathology).

To develop an in silico clinical trial framework based on mathematical modeling of tumor response to chemoradiation therapy as a prognostic indication of patient outcomes.