Physics in medicine and biology最新文献

Objective.To evaluate the trade-offs among model resolution, anatomical fidelity, computational cost, and localization accuracy in electroencephalography (EEG) source imaging using personalized head models. Special emphasis was placed on segmentation-free modeling and sparse inverse analysis for rapid brain mapping.Approach. Forward problems were solved using a scalar-potential finite-difference method combined with orthogonal matching pursuit for inverse source localization. Tissue conductivity was directly estimated from magnetic resonance images using machine learning models, including CondNet and CondNet-TART, to generate personalized, continuous conductivity distributions without explicit segmentation. Localization accuracy and computational efficiency were compared across different model resolutions and against a finite-element method (FEM) using a five-tissue head model.Main results. Segmentation-free models achieved localization errors below 23 mm while maintaining stable accuracy even at coarser resolutions. Estimated current density distributions exhibited smooth transitions across tissue boundaries. Computational time was reduced by 87.9% compared with FEM at 2.0 mm resolution. Among segmentation-free models, CondNet-TART yielded the highest localization accuracy and stability.Significance.The proposed segmentation-free framework provides efficient and accurate EEG source localization with substantially reduced computational cost. These features support time-sensitive applications such as presurgical functional mapping and real-time neuroimaging.

Aortic dissection (AD) is a life-threatening cardiovascular emergency. Non-contrast-enhanced computed tomography (NCE-CT) could provide timely AD screening with fewer contraindications compared to CE-CT imaging. However, NCE-CT examinations lack distinctive imaging characteristics of AD, leading to high rates of missed diagnoses and misdiagnoses, and increased radiologist workload. In this paper, we propose a novel end-to-end multi-task framework for automated aortic segmentation and AD detection using NCE-CT images. The framework comprises three main components: a deformable feature extractor enhancing aorta tubular-feature attention, an adaptive geometric information extraction module to optimize feature sharing between segmentation and classification tasks via the transformer cross-attention mechanism, and a knowledge distillation module transferring diagnostic information from the CE-CT-based teacher model to the NCE-CT-based student model. Multi-center tests across 3 internal and 2 external centers demonstrated that our model outperformed existing methods both for segmenting the aorta and detecting AD. Specifically, for segmenting the aorta, our framework achieved dice of 0.928 and 0.909, Jaccard index (JI) of 0.867 and 0.858, mean intersection over union (MIoU) of 0.932 and 0.913, and frequency-weighted IoU (FWIoU) of 0.995 and 0.994, in internal and external testing datasets, respectively. For identifying AD patients from non-AD patients, our framework achieved accuracies of 0.911 and 0.840, sensitivities of 0.925 and 0.888, andF1-scores of 0.922 and 0.836, in internal and external testing datasets, respectively. Ablation experiment demonstrates the effectiveness of each module. The proposed model may serve as an effective diagnostic assistant for radiologists, acting as a 'second pair of eyes' to assist in AD screening using NCE-CT images.

Multimodal images contain complementary information that is valuable for deep learning (DL)-based image segmentation. To enable effective multimodal feature learning and fusion for accurate segmentation, multimodal images usually need to be registered to achieve anatomical alignment.However, in clinical settings, multimodal image registration is often challenging. For instance, to reduce radiation exposure, CT scans usually have a smaller field of view (FoV) than MR, i.e., inconsistent anatomical content in CT and MR images, hindering accurate registration. Using such misaligned multimodal images, segmentation performance could be significantly degraded. This study aims to develop a DL-based multimodal image segmentation method that is capable of learning high-quality and strongly related image features from misaligned multimodal images without registration and produce accurate segmentation results comparable to that obtained with well-aligned multimodal images. In our method, a unified body space (UBS) module is presented, where image patches cropped from misaligned modalities are encoded to positions and projected into a unified body space, thereby largely mitigating the misalignment among multimodal images. Built upon the UBS module, a new spatial-attention is proposed and integrated into a multilevel feature fusion (MFF) module, where features learned from misaligned multimodal images are effectively fused at internal-, spatial-, and modal-levels, leading the segmentation of misaligned multimodal images to a high accuracy level. We validate our method on both public and inhouse multimodal image datasets containing 1472 patients. Experimental results demonstrate that our method outperforms state-of-the-art (SOTA) methods. The ablation study further confirms that the UBS modules can accurately project image patches from different modalities into the unified body space. Moreover, the internal-, spatial-, and modal-level feature fusion in the MFF module substantially enhances segmentation accuracy for misaligned multimodal images. Codes are available at https://github.com/BH-MICom/Patch2Space.

Objective.Incorporating iodine into deoxyribonucleic acid (DNA) bases offers a strategy to enhance radiotherapy. The iodine increases the photoabsorption cross-section and can promote DNA disruption and cell death in cancerous tissue. In this study, we investigate the local fragmentation mechanisms of iodinated DNA and the spatial extent of damage propagation following photoactivation.Approach.Single-stranded DNA oligonucleotides consisting of 2-5 bases, in which the methyl group of thymine is substituted with an iodine atom, were irradiated with synchrotron x-rays above the iodine L-shell ionisation threshold (4900 eV). Fragmentation patterns were extracted by subtracting background spectra obtained below the threshold (4500 eV), and the results were complemented by Born-Oppenheimer molecular dynamics simulations to resolve bond breaking at the atomic level.Main results.We find that longer oligonucleotide chains predominantly generate larger, high-m/zfragments, while shorter sequences produce a wider variety of small fragments. Backbone cleavage is observed in all sequences, with phosphate- and sugar-based ions dominating the spectra. Bond scission extends up to five bases from the iodination site, with the heaviest stable fragment containing two bases.Significance.Suppose this effect is extrapolated to genomic DNA, which includes about 29.5% thymine. In that case, the amount of thymine replaced by iodinated uracil can help estimate the extent of DNA damage that might occur during radiation therapy using iodine as a radiosensitiser.

Objective.To design and validate a single, reconfigurable gamma probe that overcomes the static compromise between spatial resolution and sensitivity in radio-guided surgery, enabling both rapid lesion detection and precise margin delineation.Approach.A dual-layer lead collimator was designed for a LaBr₃(Ce)-SiPM detector. A validated analytical model coupled with a multi-objective genetic algorithm (NSGA-II) was used to explore the theoretical performance limits and identify optimal geometries. A two-phase computational search identified a single, universal geometry that can be switched intraoperatively between a high-sensitivity (HS) mode and a high-resolution (HR) mode by adjusting collimator positions.Main results.The universal design, at a 30 mm distance, achieves a spatial resolution of 6.41 mm full width at half maximum (FWHM) in HR mode and a sensitivity of 1483 counts per second (cps)/MBq in HS mode. The optimization framework identified specialized, distance-specific theoretical designs with resolutions as fine as 3.26 mm FWHM. The underlying detector's energy resolution is sufficient to distinguish between ⁹⁹_mTc (140.5 keV) and¹²³I (159 keV).Significance.This work presents a practical, single-instrument solution that offers surgeons the intraoperative flexibility to prioritize either rapid detection or precise delineation. The developed design methodology provides a robust framework for creating next-generation, application-specific surgical guidance tools.

Objective.Adaptive radiotherapy aimed to account for inter- and intra-fractional anatomical changes to improve target coverage and spare normal tissues. Commercial softwares could quantify anatomical and setup differences between planning computed tomography (pCT) and image-guidance cone-beam computed tomography (CBCT) using density gamma passing rate (dGPR). This study evaluated the utility of dGPR as an automated, site-specific trigger for offline adaptive workflows, using statistical process control (SPC) to define tolerance thresholds and correlating dGPR with setup errors, planning target volume (PTV) coverage, and longitudinal stability.Approach.240 patients across six anatomical sites were retrospectively analysed. First-fraction CBCTs were compared to pCTs using MobiusCB to compute dGPR. SPC analysis was performed to establish site-specific tolerance limits. Rigid phantom studies were conducted to quantify dGPR sensitivity to setup errors. Correlation between dGPR and PTVV_100%was assessed in patients with repeat CTs, and longitudinal stability of dGPR across fractions was evaluated.Main results.SPC-derived lower action limits (Al) ranged from 97.3% (brain) to 82.2% (breast), reflecting site-specific anatomical variability and imaging protocols. Phantom studies verified dGPR sensitivity due to rigid shifts, with head dGPR decreasing to as low as 83.5% at 8 mm translation, while pelvis dGPR remained above 93.2% for the same shift, reflecting greater tolerance due to its larger volume. In head-and-neck patients, dGPR correlated moderately with changes in PTVV_100%(r= 0.56), with no coverage losses >10% observed when dGPR exceeded 90%. Longitudinal analysis showed stable dGPR for most sites, but gradual declines in head-and-neck patients, with some values falling below SPC-derived threshold of 94.3% towards the end of treatment.Significance.dGPR offered a practical, automated, and site-specific metric for detecting anatomical changes and setup errors during radiotherapy. SPC-derived thresholds provided robust action levels tailored to each site, and MobiusCB enabled automated alerts when thresholds were exceeded, reducing reliance on subjective image inspection.

Objective.This study introduces a novel non-invasive wound healing method that generates Lorentz fields (LFs) in the wound area using ultrasonic transducers under a static magnetic field, enabling localized stimulation without direct electrode contact.Approach.Theoretical derivations of the governing equations, supported by numerical simulations, demonstrate the feasibility and potential effectiveness of this technique. The model includes the two-dimensional geometry of the wound, skin layers, gel, a single-element ultrasonic probe, or a 16-element linear phased array (LPA) transducer. The pressure and velocity current density distributions in the wound area were analyzed under three different excitation configurations: (i) excitation using a single-element ultrasonic probe, (ii) beam steering of the LPA transducer at 5^∘intervals between-30∘and+30∘at 13 different angles, and (iii) focusing of the LPA transducer at 0^∘. In each configuration, distinct pressure distributions and velocity current density patterns were obtained in the wound region. In addition,in vivoanimal experiments were conducted using the single-element ultrasonic probe to evaluate the biological effects of LF-based stimulation on wound healing. The study included four experimental groups: a static magnetic field (SMF) group, an ultrasound (US) group, a combined LF group, and a control group without any stimulation.Main results.In the single-element probe configuration, the simulated velocity current density reached approximately 4.51μAcm-2, corresponding to a pressure of 0.17 MPa. These values remained within the established safety limits while being sufficient to promote wound healing. For the LPA transducer, electronic beam steering enabled a uniform distribution of acoustic pressure and induced current density over a wider wound area. The pressure ranged between ±(0.118-0.203)  MPa, and the corresponding velocity current density varied between ±(2.33-2.69) μAcm-2. In the focusing configuration (0^∘), the maximum pressure in the wound region reached 0.285 MPa, while the peak absolute velocity current density was 6.72 μAcm-2, both remaining within safe limits. Animal experiments were conducted for 14 d, with each group receiving a 5 min daily treatment. The Lorentz-field group exhibited the fastest wound closure, followed by the US and magnetic-field groups, whereas the control group showed the least improvement.Significance.The proposed method offers an innovative and safe alternative for accelerating wound healing by combining US and SMFs to generate Lorentz-induced current densities in the wound, providing localized and non-invasive therapeutic stimulation.

Objective.Patients with locally advanced non-small cell lung cancer (LA-NSCLC) exhibit heterogeneous prognoses despite receiving standard treatments, highlighting the need for more reliable prognostic biomarkers. This study aims to develop and validate OmicsMap model, a deep radiomics biomarkers derived from computed tomography images for the prediction of progression-free survival (PFS) in LA-NSCLC patients.Approach.We retrospectively analyzed data from 329 LA-NSCLC patients who underwent definitive radiotherapy. The cohort was randomly divided into development (N= 220) and independent testing set (N= 109). The prognostic signature was derived from integrated radiomics features extracted from both the primary tumor and involved lymph nodes, and inter-patient radiomics feature interactions. To achieve this, high-dimensional radiomics data from all patients were transformed into structured two-dimensional representations, termed OmicsMap, wherein radiomics feature interactions were encoded within the pixelated configuration. Deep radiomics features from the OmicsMaps were then extracted using a convolutional neural network for prognostic prediction. Model performance was evaluated by time-dependent area under the receiver operating characteristic curves area under the curve (AUC). Kaplan-Meier curves were plotted and hazard ratios (HR) were calculated via Cox proportional hazards model.Main results.The OmicsMap model achieved time-dependent AUCs of 0.76, 0.78 and 0.76 at 1, 2 and 3 years in the independent testing set, significantly outperforming the clinical model (AUC: 0.57, 0.57, 0.64;p< 0.05). The proposed model improved predictive discrimination with 7.69% increase in C-index over conventional radiomics approaches. It effectively stratified patients into high-risk and low-risk subgroups for both PFS (p< 0.001, HR = 0.380) and overall survival (p= 0.0021, HR = 0.525) in the testing set.Significance.The proposed OmicsMap model provides a novel paradigm for enhancing prognostic prediction in patients with LA-NSCLC. By improving risk stratification, the framework may help inform clinical decision-making and support future efforts toward more individualized management strategies.

Objective.Current dosimetry protocols typically recommend multiple measurements to determine recombination correction factors (ks), increasing the time required for dose measurements in the quality assurance workflow. We propose a novel dual-gap ionization chamber (DGIC) design for reference dosimetry featuring two air gaps of different thicknesses within a single device. This design enables the determination ofksdirectly from the same measurements required to determine absorbed dose-to-water. Thus, eliminating the need for separate measurements to correct for recombination losses. The approach relies on analyzing the charge ratio between the two gaps, under ultra high dose rates (UHDR) and dose per pulse (DPP) under ultra high DPP (UHDPP) conditions.Approach.A DGIC prototype with electrode distances of 1 and 0.6 mm was developed and tested using different beam qualities: (1) a 240 MeV u^-1clinical carbon ion beam at conventional field dose rates of 25 Gy min^-1, (2) a 226 MeV continuous proton beam with a current between 5 and 800 nA at the cyclotron exit, where the maximum approximately corresponds to 200 Gy s^-1in the treatment room and (3) a 9 MeV electron beam with a DPP from 0.03 to 4.2 Gy, a frequency of 60 Hz and a pulse duration between 0.7-3.9μs.ks-factors were derived for the top cavity of 1 mm gap using the DGIC method and compared against the following: for proton and carbon ions, comparisons were made with the Jaffé plot method. For the electron beam, it was compared with a dose rate independent device, a flashDiamond detector, and the integrated current transformer of the LINAC.Main results.A DGIC prototype was able to successfully correct for recombination losses under different beam modalities: for initial recombination in a clinical carbon ion beam, volume recombination in UHDR proton beam with field dose rates of 200 Gy s^-1and in UHDPP electron beams, where four pulses were delivered with DPP up to 4.2 Gy (this DPP corresponds to an effective pulse duration of 3.9μs).Significance.A DGIC design and its inherent method provides a practical and accurate way of determining dose and dose rate in emerging radiotherapy treatment modalities.

Purpose: Radiation pneumonitis occurs in approximately 10-30% of lung cancer patients treated with radiation therapy, posing a significant dose-limiting factor. The assessment of regional ventilation changes from functional ventilation data can provide essential information regarding treatment response. However, this task can be challenging since ventilation maps contain noisy measurements and artifacts. Methods: We introduce a framework that estimates physiological changes from a set of longitudinal ventilation scans. Our method identifies changes as more plausible if they follow a monotonic trend while attributing smaller confidence in regions where large fluctuations are observed. Our algorithm outputs the estimated volumes of significant function increase and decline. The proposed framework was calibrated and validated using synthetic datasets. We also applied our model to a dataset comprising 11 lung cancer patients for whom multiple 4DCT scans were obtained during the course of radiotherapy treatment. CT-derived ventilation maps were generated and used as input to the proposed framework. In order to create a control dataset where no functional changes were expected, we also shuffled the time points for the 11 patients in every possible way that discarded as much temporal information as possible resulting in 128 functional map sequences. Results: In the patient dataset, 3/11 patients were identified with significant functional decline and 4/11 with functional increase that was associated with tumor regression. Finally, in the control dataset the frequency of occurrence of significant changes was 1.6% (4/256) compared to 32% (7/22) for the original patient dataset. Conclusion: We have developed a framework for analyzing functional ventilation changes from longitudinal data. The results of the lung cancer patient dataset indicate that significant functional increase and decline can occur during the course of radiotherapy treatment. More generally, the developed framework can be used to assess ventilation changes with the potential of guiding adaptive treatment strategies. .