Med Phys [JOURNAL]
Sun
200 papers
A new in silico model to precisely design focused ultrasound brain therapies.
Conti A, Grossi M, De Andrade PC … +2 more , Duggento A, Toschi N
Med Phys · 2026 Jul · PMID 42339973 · Full text
BACKGROUND: Focused ultrasound (FUS) combined with microbubbles enables transient and noninvasive blood-brain barrier (BBB) opening, facilitating targeted drug delivery. However, accurate treatment planning remains diffi...
BACKGROUND: Focused ultrasound (FUS) combined with microbubbles enables transient and noninvasive blood-brain barrier (BBB) opening, facilitating targeted drug delivery. However, accurate treatment planning remains difficult due to inter-patient anatomical variability and the common assumption in simulations that brain tissues behave like water. PURPOSE: To develop and evaluate MODFUS ( ), an in silico acoustic simulation framework that integrates a high-resolution anatomical head model to quantify the impact of intracranial tissue heterogeneity and probe alignment on transmitted acoustic pressure, supporting treatment planning for BBB opening. METHODS: MODFUS integrates an anatomical head model comprising 115 tissue types and Computed Tomography (CT)-derived skull properties. FUS simulations were conducted using a 250 kHz single-element transducer at 13 distinct skull entry locations. Two modeling approaches were evaluated: a classical model, in which the skull was embedded in water and brain tissues were homogenized as water, and an heterogeneous anatomical model. To further assess robustness, an additional set of 50 simulations introduced controlled perturbations in probe positioning, consisting of angular deviations ( ) and translational offsets ( 7 mm) relative to the reference configuration. Model- and configuration-dependent differences were quantified using peak positive pressure (PPP), peak negative pressure (PNP), and potential therapeutic volume. Statistical significance was assessed using Wilcoxon rank-sum or Wilcoxon signed-rank tests ( = 0.05). For Classical vs Realistic models, Wilcoxon rank-sum tests were applied to PPP, PNP, and BBB exposure volume, and Levene's test assessed variance differences across 13 positions. Multiple testing was controlled using the Holm-Bonferroni procedure ( = 0.05) across all tests. Effect sizes were quantified using Cohen's d with 95% confidence intervals. RESULTS: Compared to the classical "skull+water" benchmark model, the heterogeneous model predicted up to 11% lower PPP, slightly lower PNP ( -5%), and 35% smaller potential BBB exposure volumes across the 13 paired sonication positions. After Bonferroni correction, Paired statistical testing (Wilcoxon signed-rank, two-sided) showed significant differences for PPP (p = 0.0270) and potential BBB exposure volume (p = 0.0015), while differences in PNP were not statistically significant (p = 0.8286). Levene's test for variance confirmed significant heteroscedasticity for PPP (p = 0.027) and PNP (p = 3.2 10), but not for BBB exposure volume (p = 0.2680). Cohen's d effect sizes indicated a large positive effect for PPP, a small negative effect for PNP, and a very large positive effect for BBB exposure volume. CONCLUSIONS: MODFUS demonstrates the influence of incorporating detailed tissue heterogeneity on simulation outcomes, including pressure distribution and potential BBB exposure volume. These results highlight the importance of realistic soft tissue modeling and stereotaxic probe alignment for safe and effective FUS treatment planning. The study serves as a preliminary proof-of-concept. Future studies incorporating in vivo experiments will be required to quantify the accuracy of this approach.
Blinded, bias-controlled multi-rater evaluation of human-versus-AI brain metastasis segmentation using a hybrid foundation-model framework.
Han Y, Zhu E, Mekdash MHA … +7 more , Awad O, Pathak P, Liang S, Hamstra DA, Zhang X, Siddiqui ZA, Sun B
Med Phys · 2026 Jul · PMID 42329664 · Full text
BACKGROUND: Accurate segmentation of brain metastases (BM) is essential for diagnosis, stereotactic radiosurgery planning, and longitudinal assessment. However, manual contouring is time-intensive, limiting clinical scal...
BACKGROUND: Accurate segmentation of brain metastases (BM) is essential for diagnosis, stereotactic radiosurgery planning, and longitudinal assessment. However, manual contouring is time-intensive, limiting clinical scalability, and exhibits substantial inter-observer variability. This variability complicates objective assessment of automated segmentation methods and challenges interpretation of model performance. PURPOSE: To address these limitations, we developed TUM-SAM, a hybrid foundation-model framework for fully automated BM segmentation, and introduced a bias-controlled, blinded multi-rater evaluation paradigm to determine whether AI-based BM segmentation has reached expert-level performance and whether AI-generated contours are preferred by human experts under unbiased assessment. METHODS: TUM-SAM integrates nnU-Net-based lesion detection with a tumor-adapted Med-SAM segmentation model to enable prompt-free, fully automated segmentation. Training used 301 patients (2548 lesions), and external evaluation used an independent cohort of 105 patients (397 lesions). Segmentation accuracy was benchmarked against DeepMedic and nnU-Net using Dice similarity coefficient (DSC) and 95th-percentile Hausdorff distance (HD95). Two physicians contoured all external cases, and a third physician contoured a 20-patient subset for a blinded, tumor-level, multi-rater preference study. Pairwise contour preferences were analyzed using a Bradley-Terry probabilistic model to obtain bias-adjusted estimates of relative contour quality while accounting for rater-specific tendencies and case difficulty. RESULTS: In the external cohort, TUM-SAM achieved a lesion-wise detection sensitivity of 0.94 and outperformed DeepMedic and nnU-Net across all tumor sizes, with a mean DSC of 0.84 and HD95 of 1.9 mm (nnU-Net/DeepMedic: DSC < 0.70, HD95 > 3.3 mm). Across voxel-wise evaluation, TUM-SAM's geometric performance fell within the range of inter-observer variability among physicians and was sensitive to reference construction. In contrast, in the blinded rater study, experts preferred TUM-SAM-generated contours over individual physician contours in 81-87% of raw comparisons; Bradley-Terry analysis yielded conservative, bias-corrected win probabilities of 55-56%, indicating consistent preference after adjustment for rater and case difficulty. CONCLUSION: Using a bias-controlled, blinded multi-rater evaluation framework, TUM-SAM demonstrates brain metastasis segmentation quality that is consistently preferred by expert physicians, highlighting the limitations of agreement-based voxel-wise metrics under inter-observer variability. These findings underscore the dependence of conventional evaluation on reference definition and support preference-based assessment as a complementary approach for evaluating AI segmentation quality in BM MRI.
Macro Monte Carlo dose calculation for very high energy electron (VHEE) radiotherapy.
Zhu C, Amstutz F, Costantini CM … +4 more , Volken W, Stampanoni MFM, Manser P, Fix MK
Med Phys · 2026 Jul · PMID 42329660 · Full text
BACKGROUND: Very high energy electron (VHEE) radiotherapy has gained growing interest owing to its potential to reach deep-seated targets and induce FLASH effect. Dose calculations can be performed using analytical or Mo...
BACKGROUND: Very high energy electron (VHEE) radiotherapy has gained growing interest owing to its potential to reach deep-seated targets and induce FLASH effect. Dose calculations can be performed using analytical or Monte Carlo (MC) methods. Analytical approaches enable rapid dose computation but suffer from limited accuracy in heterogeneous media, whereas MC methods provide high accuracy at the expense of substantial computational cost. Macro Monte Carlo (MMC) is a local-to-global method designed to improve dose calculation efficiency compared to general-purpose MC methods. In MMC, particle transport is based on precalculated transport data generated with general-purpose MC simulations on specific geometries, which is subsequently used to model particle transport over macroscopic steps within the absorber, avoiding computationally expensive microscopic tracking. MMC made it to a standard electron dose calculation engine in a commercial treatment planning system. However, to date, MMC has not been investigated for electron energies above 25 MeV. PURPOSE: To develop and validate an MMC framework for VHEE radiotherapy that improves dose calculation efficiency while preserving accuracy compared to general-purpose MC methods for electron energies up to 250 MeV. METHODS: Local simulations were performed using EGSnrc with monoenergetic electron pencil beams incident perpendicularly on spherical geometries (0.2-25 MeV) with radii of 0.5-3 mm, and slab geometries (25-250 MeV) of 2 mm thickness, composed of various materials. Physical quantities including energy loss, lateral displacement, and angular distributions of primary and secondary particles were scored and stored in a database. This database was subsequently used to transport electrons step-by-step in the global simulations, employing slab-based transport at energies ≥25 MeV and switching to spherical geometries for electron energies <25 MeV to account for increased scattering. Energy deposition was scored in a 3D dose grid. MMC dose calculations were validated against EGSnrc for monoenergetic VHEE beams (50-250 MeV) incident on homogeneous and heterogeneous slab phantoms, using pencil beams, parallel spot beams with 1 mm radius, and parallel beams with a field size of 5 × 5 cm. MMC and EGSnrc dose calculations were also performed for two patient CT datasets. Comparisons between MMC and EGSnrc were conducted using integrated depth dose curves, lateral dose profiles, and 3D gamma analysis with 2%/1 mm and 2%/2 mm (global) criteria and a 10% dose threshold. All simulations were performed with statistical uncertainties below 1%, and computation times were recorded. RESULTS: Integrated depth dose curves and lateral dose profiles agreed within 2% of the maximum dose for all cases considered. For homogeneous and heterogeneous phantoms, MMC dose distributions yielded gamma passing rates above 97% (2%/1 mm) and 99% (2%/2 mm), respectively, compared to EGSnrc. For patient CT datasets, gamma passing rates exceeded 94% (2%/1 mm) and 97% (2%/2 mm). Overall, MMC achieved up to a 27-fold improvement in dose calculation efficiency compared to EGSnrc. CONCLUSIONS: An MMC framework for VHEE dose calculation was successfully developed and validated for electron energies up to 250 MeV. The method demonstrated good agreement with EGSnrc while providing up to an order-of-magnitude improvement in dose calculation efficiency for the studied cases.
Development and validation of a novel pulse optimization and beam control system for conventional and ultra high dose-per-pulse (FLASH) irradiation.
Connell L, Esplen N, Lim R … +4 more , Baikalov A, Coupey N, Nguyen C, Schüler E
Med Phys · 2026 Jun · PMID 42284309 · Full text
BACKGROUND: FLASH radiotherapy requires precise control and minimal variation of dose per pulse (DPP). However, clinical linear accelerators and their beam control systems are designed to ensure accuracy of the temporall...
BACKGROUND: FLASH radiotherapy requires precise control and minimal variation of dose per pulse (DPP). However, clinical linear accelerators and their beam control systems are designed to ensure accuracy of the temporally integrated dose and do not control for transient variations in DPP during radiation delivery. PURPOSE: We introduce a robust external beam control system (EBCS) with radiofrequency optimization and beam monitoring that addresses this need. This system was designed to precisely control the output of FLASH-capable electron linear accelerators within a clinical range of energies (6-20 MeV) and to monitor the output by using a beam current transformer. METHODS: An EBCS, using either an internal transmission ion chamber or a multistage beam current transformer, was implemented to support delivery of conventional DPPs and ultrahigh DPPs (UH-DPPs) on a modified clinical linear accelerator. The EBCS was interfaced with the accelerator's gating system, and beam output and stability were maximized by optimizing the accelerating radiofrequency power efficiency through voltage inputs (V) to the automatic frequency control interface while the beam was held. The EBCS performance was tested by characterizing the beam-off latency; beam output stability within and between pulsed deliveries; sensitivity to deviations from optimization solutions; and beam current transformer linearity from conventional DPPs to UH-DPPs. RESULTS: The measured beam-off latency of the system was 56.7 µs (± 4.9 µs). The radiofrequency optimization was shown to reduce the DPP variability within the first five pulses from 26.7% to less than 0.5% for both conventional DPPs and UH-DPPs. Total output was reduced by up to 20% when V voltage inputs varied from the optimal solution by more than ± 10%. CONCLUSION: We developed an EBCS capable of delivering reproducible doses and implemented it on a modified clinical linear accelerator. Through real time readout of the beam current transformer signal and automatic radiofrequency optimization, the uncertainty in DPP within and between each delivery was reduced to < 0.5%, offering unprecedented precision and accuracy.
Sinogram-based flow estimation in computed tomography using a physics-informed neural network: Impact of gantry rotation speed, X-ray fluence and pulsed acquisition on accuracy.
Guo J, Khurana GS, Grande AG … +2 more , Alamo JCD, Contijoch F
Med Phys · 2026 Jun · PMID 42284278 · Full text
BACKGROUND: Non-invasive imaging-based assessment of blood flow plays a critical role in evaluating heart function and structure. Computed Tomography (CT) is a widely-used imaging modality that can robustly evaluate card...
BACKGROUND: Non-invasive imaging-based assessment of blood flow plays a critical role in evaluating heart function and structure. Computed Tomography (CT) is a widely-used imaging modality that can robustly evaluate cardiovascular anatomy and function, but direct methods to estimate blood flow velocity from movies of contrast dynamics have not been developed. PURPOSE: This study evaluates the impact of CT imaging parameters on Physics-Informed Neural Networks (PINN)-based flow estimation and proposes an improved PINN-based approach, SinoFlow, which uses sinogram data directly to estimate blood flow. METHODS: We generated pulsatile flow fields in an idealized 2D vessel bifurcation using computational fluid dynamics and simulated CT scans with varying gantry rotation speeds, photon flux, and pulse mode imaging settings. We compared the performance of PINN-based flow estimation using reconstructed images (ImageFlow) to SinoFlow. RESULTS: SinoFlow significantly improved flow estimation performance by avoiding temporal inconsistency errors introduced by filtered backprojection. SinoFlow was robust across all tested gantry rotation speeds and consistently produced lower mean squared error and velocity errors than ImageFlow. Additionally, SinoFlow was less susceptible to noise in the sinogram and was compatible with pulsed-mode imaging and maintained higher accuracy with shorter pulse widths. CONCLUSIONS: This study demonstrates the potential of SinoFlow for CT-based flow estimation, providing a more promising approach for non-invasive blood flow assessment. The findings inform future applications of PINNs to CT images and provide an alternative which avoids limitations associated with image-based estimation.
Diffusing alpha-emitters radiation therapy dosimetry: Geant4 multi-stage Monte Carlo modeling of sources.
Lielkajis MD, Rosenfeld AB, Guatelli S … +4 more , Metcalfe PE, Poder J, Bucci JA, Cutajar DL
Med Phys · 2026 Jun · PMID 42273934 · Publisher ↗
BACKGROUND: Diffusing alpha-emitters Radiation Therapy (DaRT) is a brachytherapy treatment modality that utilizes the diffusing progeny of to treat solid tumors with therapeutic alpha radiation. The treatment is very co...
BACKGROUND: Diffusing alpha-emitters Radiation Therapy (DaRT) is a brachytherapy treatment modality that utilizes the diffusing progeny of to treat solid tumors with therapeutic alpha radiation. The treatment is very complex, with comprehensive dosimetry, micro- and nanodosimetry, and detector studies requiring temporal modelling of the entire decay chain, diffusion, energy-transfer physics, and the implementation of boundary conditions for both desorbed and source-bound nuclides. PURPOSE: To present and validate a 3D Monte Carlo (MC) model that combines the full decay chain, radionuclide diffusion, and particle energy-transfer physics, to enable temporal DaRT dosimetry for both source-bound and diffusing nuclides. METHODS: Using the Geant4 toolkit, a multi-stage Monte Carlo model (MSMCM) was developed to combine radioactive decay, Brownian motion, and particle energy-transfer physics into a single framework. Using this framework we performed three simulations, with different boundary condition variants, for point, single, and multi-source model configurations, with the dose distributions validated against published analytical models. The MSMCM's potential application for use in multi-source in-vivo detector analyses was also assessed. RESULTS: The MSMCM was successfully able to produce spatial and temporal dose distributions for all source types: point, single, and multi-source. Additionally, for the single source model, we produced full-spectrum dosimetry, dose-buildup, and dose-rate curves over clinically relevant timeframes. Comparing the MSMCMs benchmark models, we found that the measured dose depositions of point- and radial-source scenarios were within 5%. We also found that boundary conditions had a marked impact on the axial depth-dose for realistic source geometries. CONCLUSIONS: The MSMCM provides a flexible particle-by-particle solution to modelling DaRT, producing spatial and temporal dosimetry consistent with current analytical models. Given the MSMCM's capacity to model the complete decay chain, diffusion, track-level energy depositions, and complex boundaries in a single framework, it is suitable for micro- and nanodosimetry, multi-source in-vivo dosimetry, and detector analysis studies.
Physics-informed data augmentation to simulate low dose CT scans: Application to lung nodule detection.
Mostofa M, McIntosh J, Cao Q … +3 more , Sahiner B, Farhangi MM, Petrick N
Med Phys · 2026 Jun · PMID 42260270 · Publisher ↗
BACKGROUND: Convolutional neural networks (CNNs) can be sensitive to slight changes in input images, even when the differences are imperceptible to human observers. In artificial intelligence (AI) applications for medica...
BACKGROUND: Convolutional neural networks (CNNs) can be sensitive to slight changes in input images, even when the differences are imperceptible to human observers. In artificial intelligence (AI) applications for medical imaging, these variations can result from different imaging systems and acquisition parameters. PURPOSE: In this paper, we propose a new approach based on imaging physics principles for simulating the noise characteristics of images from specific CT scanners as part of data augmentation for AI training. METHODS: Our proposed Physics-Informed Data Augmentation (PIDA) method leverages the mAs and Noise Power Spectrum (NPS) profiles of various CT reconstruction kernels to simulate the effects of various dose exposures. In this approach, the NPS of a higher dose CT scan is used to generate correlated noise, which is then stochastically inserted into the training data. This simulates the noise characteristics of the lower dose exposure and enhances variability within the training set. To demonstrate PIDA's applicability in mitigating radiation dose-related domain shift, we applied PIDA in training a neural network designed to reduce false positives in a lung nodule detection algorithm. We evaluated the impact of the noise insertion training method by assessing lung nodule detection performance on low-dose CT scans. RESULTS: Our experimental results illustrate the effectiveness of our PIDA method in simulating noise characteristics of low dose CT scans from higher dose CT scans. Including PIDA in algorithm training, improved the performance of the algorithm when it was applied to low dose CT scans. The performance in terms of Competitive Performance Metric (CPM) for the low dose scans improved to 0.677 from a CPM = 0.586 when training was performed without PIDA. CONCLUSIONS: PIDA is designed to address the performance drop in CNNs due to acquisitional differences between training and testing datasets. Our findings indicate that it enhances the performance of a CNN in detecting nodules on low-dose CT scans when acquisition differences are present.
Cell survival comparison of proton, helium, and carbon ion interlaced minibeams in water phantom simulations.
Rousseti A, Dollinger G, Reindl J
Med Phys · 2026 Jun · PMID 42260269 · Full text
BACKGROUND: Radiotherapy is a key in cancer treatment, with particle therapy providing better tumor targeting and sparing healthy tissues. Particle Minibeam Radiotherapy (PMBT) integrates the advantages of spatial fracti...
BACKGROUND: Radiotherapy is a key in cancer treatment, with particle therapy providing better tumor targeting and sparing healthy tissues. Particle Minibeam Radiotherapy (PMBT) integrates the advantages of spatial fractionation into particle radiotherapy by employing submillimeter-sized beams, thus improving the therapeutic ratio by reducing side effects. Former simulation studies have shown that interlaced proton minibeams from opposing directions in Single Energy Distal-Edge (1E) mode better protect normal tissue compared to the conventional spread-out Bragg peak (SOBP) mode. PURPOSE: Helium and carbon ion minibeams may be an alternative to enhance the protection of healthy tissue, especially in deeper regions, due to less angular spread. This in silico study evaluates the potential for normal tissue sparing while preserving the same cell survival in the tumor in case of proton, helium and carbon minibeams in 1E mode. METHODS: Simulations were performed using TOPAS (Tool for Particle Simulation) by applying single-energy interlaced minibeams (beam size σ = 0.2 mm) from two opposing directions in a 250 mm-thick water phantom, assuming a 50 mm-thick tumor at the center. For the comparative analysis, cell survival rates were calculated across the whole phantom using the saturation-corrected Microdosimetric Kinetic Model (MKM-z*) implemented through MONAS (Microdosimetry-based modeling for RBE assessment). As a dose constraint, the minimum dose in the tumor was selected to ensure a maximum of 10% cell survival within the tumor. The sparing of healthy tissues was estimated using the Linear Quadratic (LQ) model, considering variable Relative Biological Effectiveness (RBE) via MONAS. RESULTS: The findings show that helium and carbon minibeams offer enhanced protection of the normal tissues only ∼ 1-2 mm close to the tumor borders, while protons achieve an overall better sparing in the rest of the phantom for the 1E mode, when looking purely at the cell survival. CONCLUSIONS: Although protons achieved the highest mean cell survival in normal tissue, the actual sparing effect is strongly influenced by beam size and valley dose, with helium and carbon ions showing enhanced confinement of damage near the tumor edge. These findings highlight the need for the implementation of more anatomically accurate phantoms with more concise biological data as a basis.
A python framework for single-image characterization of X-ray focal spot distribution and detector point spread function.
Altieri J, Cardarelli P, Di Domenico G … +1 more , Taibi A
Med Phys · 2026 Jun · PMID 42260268 · Publisher ↗
PURPOSE: Accurate 2D characterization of X-ray tube focal spot dimensions (FS) and detector Point Spread Function (PSF) is essential for radiographic quality assurance, yet traditional methods (pinhole, slit cameras) are...
PURPOSE: Accurate 2D characterization of X-ray tube focal spot dimensions (FS) and detector Point Spread Function (PSF) is essential for radiographic quality assurance, yet traditional methods (pinhole, slit cameras) are either limited to 1D characterization or require impractical setups. This article introduces SCOPE-XR, an open-source Python framework that implements and generalizes a previously established reconstruction technique, providing fully automated 2D estimation of FS distributions and detector PSF from a single radiograph of a basic test object. The software is targeted at medical physicists, researchers, and clinical quality control personnel to streamline and enhance routine acceptance testing. DEVELOPMENT AND VALIDATION METHODS: SCOPE-XR processes radiographic images to estimate the shape and dimensions of FS and PSF distributions. The underlying algorithm utilizes automatic circle detection, derivation, pseudo-CT reconstruction and incorporates an oversampling strategy to improve PSF reconstruction accuracy at limited sampling densities. The software was validated against both virtually simulated datasets and experimental clinical acquisitions, demonstrating high fidelity in characterizing source morphology and detector responses. DATA FORMAT AND USAGE NOTES: SCOPE-XR is implemented in Python and is cross-platform compatible (Windows, macOS, Linux), requiring minimal computational resources. The software accepts standard radiographic image formats (e.g., [DICOM, TIFF, RAW]) as input and outputs 2D emission profiles, quantitative dimensional metrics, and performance plots. A small dataset of virtual and experimental acquisitions is included as an example for benchmarking and reproducibility. The source code, datasets, and comprehensive documentation are publicly accessible via its public repository: https://doi.org/10.15161/oar.it/hrrqs-cn059. POTENTIAL APPLICATIONS: SCOPE-XR provides a practical, fully automated alternative to traditional measurement techniques. Its primary clinical and scientific applications include the streamlined evaluation of imaging system performance during acceptance testing, routine quality control, and system design characterization.
Monte Carlo modeling and phantom studies show Cherenkov emission per unit dose during total skin electron therapy is a function of tissue optical properties.
Zhu Y, Ong YH, Byrd BK … +4 more , Alexander DA, Johnson M, Sourvanos D, Zhu TC
Med Phys · 2026 Jun · PMID 42259769 · Full text
BACKGROUND: Cherenkov imaging provides a noninvasive approach to visualize total skin electron therapy (TSET) radiation dose deposition on patient. Accurate conversion of Cherenkov intensity to radiation dose is necessar...
BACKGROUND: Cherenkov imaging provides a noninvasive approach to visualize total skin electron therapy (TSET) radiation dose deposition on patient. Accurate conversion of Cherenkov intensity to radiation dose is necessary for in vivo dosimetry to assess the spatial dose distribution in TSET. Studies have shown a linear correlation between Cherenkov intensity and absorbed dose, but the effect of tissue optical properties on the Cherenkov emission per dose is not well understood. PURPOSE: This work uses Monte Carlo simulations and experiments to assess how tissue optical properties affect the Cherenkov emission per dose detected during TSET. METHODS: Monte Carlo modeling was used to simulate Cherenkov generation during total skin electron therapy and quantify the effect of tissue optical properties on the detected Cherenkov emission. The study examined a clinically relevant range of absorption coefficients (0.01-1 cm) and reduced scattering coefficients (2-40 cm) at 665 nm. The effect of tissue optical properties, depth of origin and the angular distribution of Cherenkov emission on tissue surface were systematically evaluated. The Monte Carlo results are compared to measurements for a series of solid phantoms. An analytical function is proposed to fit the optical properties dependence (µ and µ') of Cherenkov emission for tissue. RESULTS: Simulation results show that Cherenkov emission decreases with increasing tissue absorption and effective attenuation coefficients but increases then decreases with tissue scattering coefficients. 80% of the surface-detected emission originated from superficial layers 0.17- 2.0 cm beneath the surface. Angle-specific generation of Cherenkov radiation in tissue has not resulted in preferential exiting angle as the propagation directions of most Cherenkov photons are randomized prior to reaching the surface. Monte Carlo simulation (max dev 0.23%) agrees with experiments to within a standard (maximum) % deviation of 2.8% (7.6%). CONCLUSION: Our findings indicate that tissue optical properties exert a substantial influence on the surface Cherenkov emission. The optical properties dependence of Cherenkov emission per dose can be expressed as a two-dimensional function of µ and µ'. An analytical expression is presented. Monte Carlo simulation agrees with experiments for TSE electrons used in a series of tissue simulating phantoms.
Alpha DaRT source activity confirmation using a reentrant well-type ionization chamber.
Deufel CL, Ordog JL, Underwood K … +4 more , Cohen Y, Cohen GN, Damato AL, Malouff TD
Med Phys · 2026 Jun · PMID 42253233 · Publisher ↗
BACKGROUND: Alpha-emitting radionuclides enable precise cancer therapy through high linear energy transfer and limited tissue penetration, damaging tumor cells while sparing healthy tissue. Diffusing Alpha-emitters Radia...
BACKGROUND: Alpha-emitting radionuclides enable precise cancer therapy through high linear energy transfer and limited tissue penetration, damaging tumor cells while sparing healthy tissue. Diffusing Alpha-emitters Radiation Therapy (Alpha DaRT) features Ra-224 sources that are implanted directly into the tumor and emit alpha particles during radioactive decay. Alpha DaRT has demonstrated efficacy and safety in preclinical and early clinical trials across multiple tumor types, including skin, head and neck, and pancreatic cancers. PURPOSE: Reliable and efficient methods for verifying Alpha DaRT source activity prior to treatment can help support accurate and consistent radiation delivery. The direct measurement of alpha particles from sources within an Alpha DaRT applicator is impractical due to their short range; however, gamma emissions from the Ra-224 sources can be used to infer radioactivity. This study established a protocol for verifying the source activity within Ra-224 Alpha DaRT applicators using a reentrant well-type ionization chamber, providing users with a practical method for detecting errors in source manufacturing or certificate paperwork without compromising applicator sterility. METHODS: Ra-224 Alpha DaRT sources in sterile packaging (Flex and Needle applicators with 1-4 sources) were assessed. Source energy spectra and activities were verified using a high-purity germanium (HPGe) radiation detector. Calibration factors (kBq/pA) were established using an IVB1000 well-type ionization chamber with measurements conducted by placing single applicator sterile packages into the chamber with sources centered in the chamber's sweet spot and corrected for temperature, pressure, and leakage current. Quality assurance was performed on 26 Flex applicators using the established calibrations before the first clinical procedure. RESULTS: HPGe measurements agreed with vendor-stated activities. The average calibration coefficient using the IVB1000 chamber was 233 ± 3 kBq/pA for Flex and 597 ± 7 kBq/pA for Needle applicators. Calibration coefficients were consistent across two IVB1000 chambers. Source number dependence was observed, with calibration factors increasing by 1.7% ± 0.7% per source (Needle) and 2.7% ± 0.6% per source (Flex). Measurement repeatability was 3.3%. Applying the calibration to 26 Flex applicators before the first patient treatment yielded a 1.1 ± 5.8% (range: -7.5% to 11.4%) difference relative to the vendor's stated activity. CONCLUSION: A reentrant well-type ionization chamber is suitable for pre-treatment quality assurance of Ra-224 Alpha DaRT applicators, enabling verification of the vendor-stated activity while maintaining sterility within sealed packaging.
Toward universal dose prediction: A multi-scale, multi-objective framework for sequential boost radiotherapy.
Maniscalco AM, Zhong X, Domal S … +5 more , Lin YC, Sher D, Lin MH, Jiang S, Nguyen D
Med Phys · 2026 Jun · PMID 42253229 · Full text
BACKGROUND: Sequential boost radiotherapy (RT) poses a challenge in allocating dose across multiple plans while protecting organs at risk (OARs). Clinicians must decide whether OAR sparing should occur primarily in the i...
BACKGROUND: Sequential boost radiotherapy (RT) poses a challenge in allocating dose across multiple plans while protecting organs at risk (OARs). Clinicians must decide whether OAR sparing should occur primarily in the initial plan, the boost plan(s), or all plans, resulting in a time-intensive, iterative optimization process. PURPOSE: Current dose prediction frameworks are limited to single plans and do not account for complexities introduced by sequential boosts. We propose a multi-plan dose prediction framework that models both individual plan doses and the cumulative plan-sum dose. By integrating the full course's context, this approach may help planners establish optimization objectives with fewer iterative adjustments. Ultimately, this framework aims to enable a versatile dose prediction approach adaptable to any RT course, regardless of treatment site, fractionation scheme, or the number of sequential plans. METHODS: We developed a U-Net-based Hybrid Convolutional Neural Network (CNN) that processes CT images, OARs, PTVs, and dosimetric goals to predict dose distributions for each plan and the plan-sum. It incorporates five pooling layers, skip connections, and a transformer bottleneck to capture global context. Deep supervision is applied in the decoder to encourage robust feature representation at deeper network layers, acting as a form of regularization. The single-plan model used for comparison differs from the multi-plan model in several key ways: (1) it only processes a single treatment plan at a time, (2) plan-sum dosimetric goals are re-scaled to reflect the plan fractions, (3) plan-sum PTVs are omitted, (4) it only predicts plan doses, and (5) plan-sum dose distributions were generated by summing plan dose predictions. Models were trained using a multi-objective loss vector of Mean Squared Error (MSE) for voxel-wise accuracy and Multi-Scale Structural Similarity Index (MS-SSIM) for regional coherence. Loss was averaged across all output levels to encourage global coherence. To improve training stability and ensure that parameter updates benefit both loss objectives, we used a Jacobian descent strategy via the TorchJD package. We collected a site-agnostic dataset of 64 patients that underwent sequential boost RT (38/6/20 training/validation/testing split). RESULTS: The multi-plan model and single-plan model were trained to convergence. Evaluation metrics included Mean Absolute Error as a percent of prescription dose (MAE/Rx) and Structural Similarity Index Measure (SSIM). The multi-plan model achieved statistically significant differences versus the single-plan model in plan dose distributions with MAE/Rx (1.244 ± 0.151% vs. 1.650 ± 0.172%, p < 0.001) and SSIM (0.964 ± 0.005 vs. 0.944 ± 0.009, p < 0.001), and plan-sum dose distributions with MAE/Rx (1.146 ± 0.174% vs. 1.525 ± 0.188%, p < 0.001) and SSIM (0.972 ± 0.006 vs. 0.960 ± 0.006, p < 0.001). CONCLUSION: Our multi-plan dose prediction framework improves voxel-wise accuracy and perceptual consistency by incorporating plan-sum information. Unlike traditional single-plan prediction models, our approach processes data from multiple treatment plans simultaneously, allowing it to consider cumulative dose requirements across the full treatment course. This approach can streamline treatment planning by providing clinicians with an accurate, comprehensive strategy for dose allocation in sequential boost RT. This framework lays the foundation for a universal RT dose prediction model capable of handling any fractionation scheme, disease site, or number of plans, which we plan to demonstrate in future work.
Radiomics signatures based on computed tomography for noninvasive prediction of ICOS expression and prognosis in head and neck squamous cell carcinoma.
Peng C, Liu F, Wang B … +2 more , Fan G, Li Z
Med Phys · 2026 Jun · PMID 42253227 · Publisher ↗
BACKGROUND: Head and neck squamous cell carcinoma (HNSCC) is highly invasive and heterogeneous, with significant differences in patients prognosis and immunotherapy efficacy. Studies have shown that inducible co-stimulat...
BACKGROUND: Head and neck squamous cell carcinoma (HNSCC) is highly invasive and heterogeneous, with significant differences in patients prognosis and immunotherapy efficacy. Studies have shown that inducible co-stimulator (ICOS) is a favorable prognostic factor for HNSCC. PURPOSE: This study aims to investigate the relationship between ICOS expression and the prognosis of HNSCC patients. Specifically, we aim to explore the potential of radiomic models, developed through radiomic feature extraction and selection, in predicting ICOS expression levels in HNSCC patients. By evaluating the predictive efficacy of these models, we seek to establish a noninvasive method for assessing ICOS expression, which may serve as a valuable prognostic factor in HNSCC and aid in personalized treatment strategies. METHODS: A number of 483 HNSCC samples were extracted from The Cancer Genome Atlas (TCGA) database to investigate the relevance between ICOS expression and the survival of HNSCC patients. Moreover, 139 intersection cases from TCGA and The Cancer Imaging Archive (TCIA) databases were chosen for the extraction radiomic features and the development of radiomic models. Following the selection of radiomic features by recursive feature elimination (RFE), radiomic models were developed via logistic regression (LR) and support vector machine (SVM). Receiver operating characteristic (ROC) curves, precision-recall (PR) curves, calibration curves, and decision curve analysis (DCA) were applied to evaluate the prediction efficacy of radiomic models. RESULTS: ICOS was markedly relevant to the survival of HNCSS patients, with high expression of ICOS serving as a protective factor for their overall survival (HR = 0.584, 95%CI = 0.439-0.776, P < 0.001). After extraction and selection of radiomic features, LR and SVM radiomic models were developed based on the optimal five features. Furthermore, both radiomic models demonstrated strong predictive effectiveness for ICOS expression, with the SVM radiomic model exhibiting superior predictive performance. CONCLUSIONS: Radiomic models can noninvasively predict the expression of ICOS, which influences the prognosis of HNSCC patients.
Comparison of deep learning and particle smoother EM methods for estimation of Rb-82 myocardial perfusion PET kinetic parameters.
Chin M, Zou SJ, Chinn G … +1 more , Levin CS
Med Phys · 2026 Jun · PMID 42253226 · Publisher ↗
BACKGROUND: Positron emission tomography (PET) enables quantification of dynamic physiological processes through time-resolved imaging. In myocardial perfusion PET, kinetic compartment modeling is used to estimate physi...
BACKGROUND: Positron emission tomography (PET) enables quantification of dynamic physiological processes through time-resolved imaging. In myocardial perfusion PET, kinetic compartment modeling is used to estimate physiological parameters and derive myocardial blood flow. However, conventional nonlinear least squares (NLLS) estimation is sensitive to model misspecification when not all parameters can be reliably estimated and must instead be fixed or initialized using population averages, which can degrade accuracy. PURPOSE: This work develops and evaluates two alternative kinetic analysis approaches for PET: a particle smoother-based Expectation-Maximization method (PSEM) and a convolutional neural network (CNN). METHODS: Both methods were evaluated using simulated dynamic myocardial perfusion studies and compared against NLLS and a Kalman-smoother-based Expectation-Maximization (KEM) algorithm across multiple frame durations and noise levels. RESULTS: Across 2-10 s frames, the CNN achieved the lowest relative errors for all parameters ( : 8.78%-4.98%, : 26.05%-25.50%, : 34.34%-22.76%), significantly outperforming NLLS, KEM, and PSEM (Holm-adjusted at 1.0 noise, 2-s frames), although performance degraded under out-of-distribution input-function conditions. CONCLUSIONS: Overall, the CNN provided the most accurate and robust in-distribution kinetic parameter estimates across frame durations. In contrast, PSEM exhibited parameter-dependent behavior, improving estimation while underperforming for , suggesting that further methodological refinement is needed.
A coincidence-based response matrix for correction of charge sharing spectral distortions in photon counting detectors.
Monaco V, Brombal L, Delogu P … +7 more , Feruglio A, Fiorini M, Longo R, Marchetti L, Poli AM, Rigon L, Rosso V
Med Phys · 2026 Jun · PMID 42253218 · Full text
BACKGROUND: Charge sharing between pixels distorts the count and spectral information of X-ray photon counting detectors. Compensation methods for charge sharing effects are required to exploit the full potentiality of t...
BACKGROUND: Charge sharing between pixels distorts the count and spectral information of X-ray photon counting detectors. Compensation methods for charge sharing effects are required to exploit the full potentiality of these detectors in medical diagnosis. PURPOSE: A statistical method is proposed to correct charge sharing effects in a pixellated photon counting detector by applying a spectral response matrix determined with a coincidence-based acquisition. METHODS: The technique is based on a preliminary calibration with a uniform irradiation and an arbitrary polychromatic spectrum, during which the number of coincidences between a pixel and its eight neighbours are collected for different combinations of energy bins. A coincidence-based response matrix (CBRM) is determined and afterwards applied to correct other spectra acquired with the same detector and conventional multi-comparator electronics. The technique was validated with Geant4 Monte Carlo simulations of a 1 mm thick CdTe detector and with data collected with a pixel hybrid detector consisting of a 300 thick silicon sensor readout by a Timepix4 chip. The effect of pulse pileup was not analyzed in this study. RESULTS: The response matrix restores the spectral information with a performance comparable to analog charge summing (ACS) algorithms. For example, for a simulation of a spectrum from a 120 kV X-ray tube attenuated by a solution of water and iodine and a CdTe detector with a pixel size of 200 µm, the mean absolute percentage errors (MAPE) from the comparison of the corrections with an ideal spectrum are 20.0% for the CBRM method and 22.8% for ACS. The ACS method is more sensitive to electronic noise than the CBRM correction, thus requiring a higher noise discrimination threshold. For experimental acquisitions of monochromatic spectra with the silicon sensor, the mean values and standard deviations of Gaussian fits of the restored energy peaks provide results close to those from a clustering algorithm based on 3 3 pixel blocks. The MAPE value from the comparison of the CBRM correction and clustering distributions for a polychromatic spectrum of an X-ray tube at 50 kV attenuated by an Ag solution is 6.2% It is also shown that a CBRM matrix determined with a fine division of the energy range can be adapted to match a lower number of energy bins employed for subsequent acquisitions without affecting the accuracy of the spectrum correction. A preliminary reconstruction of a nonuniform irradiation demonstrates the potentiality of the method to restore general spectral images. CONCLUSIONS: The proposed method allows the experimental determination of a response matrix which is independent of physics models or parameterizations and is realizable with a simple coincidence electronic circuit involving a limited number of pixels in a calibration stage. With respect to ACS techniques, the application of the response matrix requires only the number of counts collected with existing readout systems with multiple comparators, without introducing additional dead-times during the acquisition.
A constraint-normalized robustness index for HDR brachytherapy.
Portocarrero-Bonifaz A, Viviers E, Mclemore L … +3 more , Sowards KT, Silva SR, Nieto K
Med Phys · 2026 Jun · PMID 42253208 · Full text
BACKGROUND: Geometric uncertainties in high-dose-rate brachytherapy, including applicator displacement, catheter deflection, and interstitial needle spread, produce dose deviations that are patient and plan specific. Unl...
BACKGROUND: Geometric uncertainties in high-dose-rate brachytherapy, including applicator displacement, catheter deflection, and interstitial needle spread, produce dose deviations that are patient and plan specific. Unlike external beam radiotherapy, where robustness evaluation frameworks are increasingly standardized, high dose rate brachytherapy lacks a normalized aggregate metric for quantifying plan sensitivity to geometric perturbations. PURPOSE: To define a bounded, dimensionless robustness index for per-fraction evaluation of high dose rate brachytherapy plans that is normalized to clinically meaningful dose-limit constraints, incorporates dose-dependent penalty sensitivity, and provides both a full-scenario severity measure and a dose-threshold-filtered variant. METHODS: Geometric perturbation scenarios (translations, rotations, radial expansion/contraction, catheter deflection, and combined modes) are applied to source dwell positions. For each organ at risk, a convex-transformed, constraint-normalized penalty quantifies the fraction of the remaining margin to the dose-limit constraint consumed per scenario. For target volumes, a prescription-normalized piecewise penalty captures under-coverage and excessive dose escalation. The Robustness Index is defined as one minus the mean scenario penalty, yielding a value in [0,1]. RESULTS: The framework produces structure-independent, interpretable indices where a Robustness Index indicates that scenarios consume on average only 10% of the normalized remaining margin to the dose-limit constraint. A threshold-filtered variant evaluates robustness among only those scenarios that exceed a clinically specified dose level. CONCLUSIONS: The proposed robustness index provides a standardized, clinically anchored metric for high dose rate brachytherapy plan evaluation under geometric uncertainty, complementing existing dose volume histogram based reporting with a single interpretable number per structure.
Evaluating the role of plan complexity metrics in online adaptive radiotherapy for pancreatic cancer patients.
Cavinato S, Galetto M, Scaggion A … +7 more , Bettinelli A, Nardini M, Chiloiro G, Romano A, Paiusco M, Boldrini L, Placidi L
Med Phys · 2026 Jun · PMID 42253183 · Full text
PURPOSE: To systematically investigate the behavior of plan complexity metrics (PCMs) in an MR-Linac online adaptive radiotherapy (oART) workflow for pancreatic cancer, and to evaluate their potential as surrogate indica...
PURPOSE: To systematically investigate the behavior of plan complexity metrics (PCMs) in an MR-Linac online adaptive radiotherapy (oART) workflow for pancreatic cancer, and to evaluate their potential as surrogate indicators of delivery accuracy. METHODS: Thirty-seven patients with locally advanced pancreatic cancer were retrospectively analyzed, yielding 222 MR-Linac plans (37 reference and 185 delivered fractions). Fifteen PCMs were extracted from plans generated with three optimizers: Penalty, Objectives and Constraints, and A3i (current clinical practice). Plan specific quality assurance (PSQA) has been performed through an independent dose calculation algorithm. Statistical analyses included: (i) inter-optimizer comparisons (ANOVA and mixed-effects models), (ii) variance decomposition of adapted-plan complexity metrics using linear mixed-effects models (LMEMs), and (iii) evaluation of PSQA stability using statistical process control (SPC) and leave-one-patient-out (LOPO) cross-validation. RESULTS: Optimizer choice strongly influenced plan complexity. The Penalty optimizer generated higher-complexity plans, whereas Objectives and Constraints and A3i produced more modulation-efficient configurations with fewer small, low-MU segments. Variance decomposition identified a subset of metrics that exhibited consistent behavior across all optimizers, serving as robust descriptors independent of the algorithm. Metrics dominated by between-patient variance (σ ) emerged as reliable surrogates for patient-specific complexity Tongue & Groove Index, Average Leaf Gap and Number of Active Leaves consistently showed high between-patient contributions (σ > 74%) among others. In contrast, metrics related to low-MU segments (Seg< 5 and Seg< 5 [%]) were dominated by within-patient variance (σ ), with A3i showing the most pronounced fluctuations (85.8% and 84.8%, respectively). These descriptors are therefore more sensitive to plan-specific or optimizer-related stochasticity than to stable patient factors. SPC analyses demonstrated that the current adaptive workflow is robustly stable for most patients: 11 of 13 never experienced a fraction below the tolerance level (TL), and 12 of 13 never exceeded the action level (AL), even when thresholds were dynamically recalculated within the LOPO-CV. CONCLUSION: This study provides the first systematic assessment of PCMs in MR-guided oART, demonstrating optimizer-specific complexity signatures, predominant inter-patient variability, and the predictive value of selected metrics for delivery accuracy. Although limited to a single tumor site and workflow, the methodology supports the development of institution-specific, complexity-aware scorecards to enhance adaptive planning and quality assurance.
VIGA (VIrtual GAster): An integrated analysis platform for quantitative assessments of gastric geometry and motility using MRI.
Liao D, Mark EB, Poulsen MB … +9 more , Davidsen L, Hagn-Meincke R, Rosca V, Rusu R, Gheorghe L, Drug V, Brock C, Frøkjær JB, Drewes AM
Med Phys · 2026 May · PMID 42230338 · Publisher ↗
BACKGROUND: Current methodologies, such as scintigraphy and gastric manometry, for assessing gastric physiology exhibit notable limitations, particularly in capturing the dynamic changes of the gastric wall during fillin...
BACKGROUND: Current methodologies, such as scintigraphy and gastric manometry, for assessing gastric physiology exhibit notable limitations, particularly in capturing the dynamic changes of the gastric wall during filling and emptying. PURPOSE: To address this gap, we introduce a novel integrated analysis platform named VIGA (VIrtual GAster), designed to quantify stomach geometry and motility using Magnetic Resonance Imaging (MRI). METHODS: The platform, along with its standardized analytical pipeline, was developed and applied to both static and dynamic MRI datasets from 12 healthy individuals after the ingestion of a 600 mL test meal, enabling the evaluation of both morphology and dynamic changes of the stomach wall. Gastric morphology was quantified by measuring the volume and surface geometry of the fundus, corpus, and antrum. Gastric motility was visualized using a comprehensive motility map, from which contraction frequency, amplitude, and speed were derived. RESULTS: The gastric morphology and motility properties can be quantified and visualized interactively on the VIGA platform using individual-specific MRI datasets. The analysis revealed region-specific differences in gastric contractions, with the distal region demonstrating greater occlusion of the contractions (15.8 ± 4.0% in distal stomach vs. 8.9 ± 4.0% in proximal stomach, p < 0.001). Additionally, the distal stomach tended to exhibit higher contraction propagation speed (4.87 ± 1.7 mm/s in distal vs. 4.01 ± 0.9 mm/s in proximal stomach, p = 0.07). Gastric configurations also varied among the compartments, with the fundus having the highest inverse mean curvature: an indicator of the ratio of the wall tension to intragastric pressure (p = 0.003). CONCLUSIONS: The VIGA interface effectively integrates gastric functional assessment using non-invasive imaging, offering potential to advance the clinical evaluation of gastric function in diseases.
Multivariate statistical process control for daily linear accelerator quality assurance: A framework for early detection and diagnosis.
Legrand C, Tiplica T, Kobi A
Med Phys · 2026 May · PMID 42219659 · Publisher ↗
BACKGROUND: Statistical Process Control (SPC) has gained increasing interest in radiotherapy Quality Assurance (QA), particularly following recommendations from American Association of Physicists in Medicine Task Group 2...
BACKGROUND: Statistical Process Control (SPC) has gained increasing interest in radiotherapy Quality Assurance (QA), particularly following recommendations from American Association of Physicists in Medicine Task Group 218. However, SPC is still predominantly applied via univariate charts such as Shewhart and Exponentially Weighted Moving Average (EWMA), even though many linac QA parameters (dose, symmetry, and flatness) are correlated and may drift together. Monitoring each parameter independently increases workload and overlooks the covariance structure, potentially reducing sensitivity to emerging faults and contributing to both false alarms and missed deviations. Multivariate Statistical Process Control (MSPC) techniques, such as Hotelling's T and Multivariate EWMA (MEWMA), address these limitations but remain underused in clinical practice. PURPOSE: This study aims to (1) provide a clear, practical framework for implementing T and MEWMA charts in radiotherapy machine QA, (2) compare their detection performance with standard univariate Shewhart and EWMA charts, and (3) demonstrate the diagnostic added value of variable-level contribution analysis for identifying the root causes of out-of-control (OC) conditions. METHODS: Daily QA measurements were collected over 168 days on a TrueBeam STx accelerator using a Daily QA3 device, yielding four dosimetric beam parameters: dose deviation, flatness, axial symmetry, and transverse symmetry. Univariate (I-chart, EWMA) and multivariate (T, MEWMA) charts were constructed following standard SPC methodology. Phase I data (90 observations) were validated for stationarity, independence, and normality using Kwiatkowski-Phillips-Schmidt-Shin, Augmented-Dickey-Fuller, Ljung-Box, Shapiro-Wilk, and several multivariate normality tests. The upper control limits for T and MEWMA were derived from the theoretical F distribution and χ approximation, respectively. Contribution analysis based on Cholesky decomposition was used to quantify variable-level responsibility for each multivariate alarm. Detection performance was evaluated for three real clinical events: an abrupt dose anomaly, a progressive drift in symmetry/flatness, and a sudden monitor-chamber failure. RESULTS: Univariate charts detected the major deviations but required monitoring eight separate charts, increasing cognitive and operational burden. Both T and MEWMA charts detected all clinically relevant events identified by univariate charts, and often earlier. The MEWMA chart detected the onset of transverse symmetry degradation two measurements before the univariate EWMA and identified coordinated deviations across variables that were not yet individually out of control. T was particularly effective for abrupt shifts, mirroring univariate Shewhart performance but within a consolidated multivariate framework. Contribution analysis consistently identified the variable(s) driving each OC signal, providing clear diagnostic insight. Multivariate monitoring also revealed that the process never returned to an in-control state between Events 2 and 3, a finding not apparent from the ±2% specification limit or univariate charts. This demonstrates the importance of accounting for covariance when interpreting QA data. CONCLUSIONS: Hotelling's T and MEWMA charts enhance early detection of deviations in linac QA by integrating the covariance structure among beam parameters and reducing the family-wise false-alarm rate inherent to multiple univariate charts. Their use streamlines workflow, improves diagnostic clarity through contribution analysis, and provides earlier warnings of progressive faults. MSPC represents a clinically valuable and operationally efficient extension to current radiotherapy QA practice, particularly as QA datasets become increasingly multidimensional.
Resilience-depletion hypothesis: Unifying BED from conventional to ablative radiotherapy.
Oliveira JM
Med Phys · 2026 May · PMID 42219650 · Publisher ↗
BACKGROUND: Cell-survival modeling remains fundamental to radiobiology because it underlies both mechanistic interpretation of radiosensitivity and practical isoeffect quantities such as the biologically effective dose (...
BACKGROUND: Cell-survival modeling remains fundamental to radiobiology because it underlies both mechanistic interpretation of radiosensitivity and practical isoeffect quantities such as the biologically effective dose (BED). The linear-quadratic (LQ) model remains the standard low-dose framework, yet its direct extrapolation to the large doses per fraction used in stereotactic body radiotherapy (SBRT) is problematic. Piecewise corrections such as the universal survival curve (USC) recover the expected high-dose tail by explicit stitching, but no single continuous formalism has achieved broad acceptance across the therapeutic range while also providing a clear language for heterogeneous survival-curve phenotypes. PURPOSE: To present the resilience-depletion (RD) framework as a macroscopic theory of acute cell survival under stress, and to develop from it a unified BED formalism with direct clinical applicability in radiotherapy. METHODS: Starting from a single postulate for passive cells under acute exposure, we derive a continuous hazard-based survival law in which cumulative stress progressively depletes a resilience reserve. The framework is formulated in generalized homogeneous and heterogeneous forms, its limiting behaviors are analyzed, and exact algebraic mappings from the historical descriptors to RD parameters are obtained. Over the therapeutic range, the generalized law reduces to an effective three-parameter form from which unified BED, EQD2, and exact isoeffect inversions follow. The reduced model is then examined against curated clonogenic datasets and through back-projection of accepted NSCLC hypofractionated regimens to a conventional reference scale. RESULTS: RD yields a continuous cumulative hazard governed in its therapeutic form by three macroscopic parameters: initial resilience ( ), sensitization rate ( ), and killing efficiency ( ). In the homogeneous single-target limit, the low-dose expansion recovers LQ behavior, whereas the high-dose limit approaches the expected shifted linear tail without piecewise construction. The generalized formulation further shows how homogeneous multi-target structure produces extended initial near-linearity and how heterogeneous mixtures generate inverse-shoulder ensemble behavior, including compact treatment-range representations with effective . Under acute irradiation, the therapeutic single-target law is mathematically equivalent to the acute Linear Quadratic-Linear (LQ-L) class after reparameterization. Clinically, the resulting BED framework provides exact schedule inversion and maps accepted NSCLC regimens to a comparatively coherent EQD2 neighborhood around the conventional reference. CONCLUSIONS: RD provides a continuous macroscopic theory of acute radiation survival that unifies low-dose LQ structure, high-dose linear-tail behavior, and key homogeneous and heterogeneous survival-curve topologies within a single hazard-based formalism. In this sense, it is not only a replacement BED expression, but a broader statistical framework linking observed survival geometry to underlying effective structure. Its reduced therapeutic form retains immediate clinical utility through unified BED, EQD2, exact isoeffect inversion, and direct translation from historical radiobiological descriptors. These results support RD as a consistent acute baseline for both radiobiological interpretation and practical radiotherapy modeling. Extension to dose-rate effects, incomplete recovery, brachytherapy, and other prolonged-delivery settings will require explicit time-dependent recovery kinetics beyond the present acute formulation.
About
- Frequency
- Sun
- Papers found
- 200
- RSS feed
- Subscribe