Volume Rendering Research Articles

Distinct brain morphometry patterns revealed by deep learning improve prediction of post-stroke aphasia severity

BackgroundEmerging evidence suggests that post-stroke aphasia severity depends on the integrity of the brain beyond the lesion. While measures of lesion anatomy and brain integrity combine synergistically to explain aphasic symptoms, substantial interindividual variability remains unaccounted. One explanatory factor may be the spatial distribution of morphometry beyond the lesion (e.g., atrophy), including not just specific brain areas, but distinct three-dimensional patterns.MethodsHere, we test whether deep learning with Convolutional Neural Networks (CNNs) on whole brain morphometry (i.e., segmented tissue volumes) and lesion anatomy better predicts chronic stroke individuals with severe aphasia (N = 231) than classical machine learning (Support Vector Machines; SVMs), evaluating whether encoding spatial dependencies identifies uniquely predictive patterns.ResultsCNNs achieve higher balanced accuracy and F1 scores, even when SVMs are nonlinear or integrate linear or nonlinear dimensionality reduction. Parity only occurs when SVMs access features learned by CNNs. Saliency maps demonstrate that CNNs leverage distributed morphometry patterns, whereas SVMs focus on the area around the lesion. Ensemble clustering of CNN saliencies reveals distinct morphometry patterns unrelated to lesion size, consistent across individuals, and which implicate unique networks associated with different cognitive processes as measured by the wider neuroimaging literature. Individualized predictions depend on both ipsilateral and contralateral features outside the lesion.ConclusionsThree-dimensional network distributions of morphometry are directly associated with aphasia severity, underscoring the potential for CNNs to improve outcome prognostication from neuroimaging data, and highlighting the prospective benefits of interrogating spatial dependence at different scales in multivariate feature space.

The value of synthetic MRI in detecting the brain changes and hearing impairment of children with sensorineural hearing loss.

Sensorineural hearing loss (SNHL) can arise from a diverse range of congenital and acquired factors. Detecting it early is pivotal for nurturing speech, language, and cognitive development in children with SNHL. In our study, we utilized synthetic magnetic resonance imaging (SyMRI) to assess alterations in both gray and white matter within the brains of children affected by SNHL. The study encompassed both children diagnosed with SNHL and a control group of children with normal hearing {1.5-month-olds (n = 52) and 3-month-olds (n = 78)}. Participants were categorized based on their auditory brainstem response (ABR) threshold, delineated into normal, mild, moderate, and severe subgroups.Clinical parameters were included and assessed the correlation with SNHL. Quantitative analysis of brain morphology was conducted using SyMRI scans, yielding data on brain segmentation and relaxation time.Through both univariate and multivariate analyses, independent factors predictive of SNHL were identified. The efficacy of the prediction model was evaluated using receiver operating characteristic (ROC) curves, with visualization facilitated through the utilization of a nomogram. It's important to note that due to the constraints of our research, we worked with a relatively small sample size. Neonatal hyperbilirubinemia (NH) and children with inner ear malformation (IEM) were associated with the onset of SNHL both at 1.5 and 3-month groups. At 3-month group, the moderate and severe subgroups exhibited elevated quantitative T1 values in the inferior colliculus (IC), lateral lemniscus (LL), and middle cerebellar peduncle (MCP) compared to the normal group. Additionally, WMV, WMF, MYF, and MYV were significantly reduced relative to the normal group. Additionally, SNHL-children with IEM had high T1 values in IC, and LL and reduced WMV, WMF, MYV and MYF values as compared with SNHL-children without IEM at 3-month group. LL-T1 and WMF were independent risk factors associated with SNHL. Consequently, a prediction model was devised based on LL-T1 and WMF. ROC for training set, validation set and external set were 0.865, 0.806, and 0.736, respectively. The integration of T1 quantitative values and brain volume segmentation offers a valuable tool for tracking brain development in children affected by SNHL and assessing the progression of the condition's severity.

Deep learning promoted target volumes delineation of total marrow and total lymphoid irradiation for accelerated radiotherapy: A multi-institutional study

Background and purposeOne of the current roadblocks to the widespread use of Total Marrow Irradiation (TMI) and Total Marrow and Lymphoid Irradiation (TMLI) is the challenging difficulties in tumor target contouring workflow. This study aims to develop a hybrid neural network model that promotes accurate, automatic, and rapid segmentation of multi-class clinical target volumes. Materials and methodsPatients who underwent TMI and TMLI from January 2018 to May 2022 were included. Two independent oncologists manually contoured eight target volumes for patients on CT images. A novel Dual-Encoder Alignment Network (DEA-Net) was developed and trained using 46 patients from one internal institution and independently evaluated on a total of 39 internal and external patients. Performance was evaluated on accuracy metrics and delineation time. ResultsThe DEA-Net achieved a mean dice similarity coefficient of 90.1 % ± 1.8 % for internal testing dataset (23 patients) and 91.1 % ± 2.5 % for external testing dataset (16 patients). The 95 % Hausdorff distance and average symmetric surface distance were 2.04 ± 0.62 mm and 0.57 ± 0.11 mm for internal testing dataset, and 2.17 ± 0.68 mm, and 0.57 ± 0.20 mm for external testing dataset, respectively, outperforming most of existing state-of-the-art methods. In addition, the automatic segmentation workflow reduced delineation time by 98 % compared to the conventional manual contouring process (mean 173 ± 29 s vs. 12168 ± 1690 s; P < 0.001). Ablation study validate the effectiveness of hybrid structures. ConclusionThe proposed deep learning framework achieved comparable or superior target volume delineation accuracy, significantly accelerating the radiotherapy planning process.

Gray matters: ViT-GAN framework for identifying schizophrenia biomarkers linking structural MRI and functional network connectivity

Brain disorders are often associated with changes in brain structure and function, where functional changes may be due to underlying structural variations. Gray matter (GM) volume segmentation from 3D structural MRI offers vital structural information for brain disorders like schizophrenia, as it encompasses essential brain tissues such as neuronal cell bodies, dendrites, and synapses, which are crucial for neural signal processing and transmission; changes in GM volume can thus indicate alterations in these tissues, reflecting underlying pathological conditions. In addition, the use of the ICA algorithm to transform high-dimensional fMRI data into functional network connectivity (FNC) matrices serves as an effective carrier of functional information. In our study, we introduce a new generative deep learning architecture, the conditional efficient vision transformer generative adversarial network (cEViT-GAN), which adeptly generates FNC matrices conditioned on GM to facilitate the exploration of potential connections between brain structure and function. We developed a new, lightweight self-attention mechanism for our ViT-based generator, enhancing the generation of refined attention maps critical for identifying structural biomarkers based on GM. Our approach not only generates high quality FNC matrices with a Pearson correlation of 0.74 compared to real FNC data, but also uses attention map technology to identify potential biomarkers in GM structure that could lead to functional abnormalities in schizophrenia patients. Visualization experiments within our study have highlighted these structural biomarkers, including the medial prefrontal cortex (mPFC), dorsolateral prefrontal cortex (DL-PFC), and cerebellum. In addition, through cross-domain analysis comparing generated and real FNC matrices, we have identified functional connections with the highest correlations to structural information, further validating the structure-function connections. This comprehensive analysis helps to understand the intricate relationship between brain structure and its functional manifestations, providing a more refined insight into the neurobiological research of schizophrenia.

A joint ESTRO and AAPM guideline for development, clinical validation and reporting of artificial intelligence models in radiation therapy

Background and purposeArtificial Intelligence (AI) models in radiation therapy are being developed with increasing pace. Despite this, the radiation therapy community has not widely adopted these models in clinical practice. A cohesive guideline on how to develop, report and clinically validate AI algorithms might help bridge this gap. Methods and materialsA Delphi process with all co-authors was followed to determine which topics should be addressed in this comprehensive guideline. Separate sections of the guideline, including Statements, were written by subgroups of the authors and discussed with the whole group at several meetings. Statements were formulated and scored as highly recommended or recommended. ResultsThe following topics were found most relevant: Decision making, image analysis, volume segmentation, treatment planning, patient specific quality assurance of treatment delivery, adaptive treatment, outcome prediction, training, validation and testing of AI model parameters, model availability for others to verify, model quality assurance/updates and upgrades, ethics. Key references were given together with an outlook on current hurdles and possibilities to overcome these. 19 Statements were formulated. ConclusionA cohesive guideline has been written which addresses main topics regarding AI in radiation therapy. It will help to guide development, as well as transparent and consistent reporting and validation of new AI tools and facilitate adoption.

No-Reference-Based and Noise Level Evaluations of Cinematic Rendering in Bone Computed Tomography.

Cinematic rendering (CR) is a new 3D post-processing technology widely used to produce bone computed tomography (CT) images. This study aimed to evaluate the performance quality of CR in bone CT images using blind quality and noise level evaluations. Bone CT images of the face, shoulder, lumbar spine, and wrist were acquired. Volume rendering (VR), which is widely used in the field of diagnostic medical imaging, was additionally set along with CR. A no-reference-based blind/referenceless image spatial quality evaluator (BRISQUE) and coefficient of variation (COV) were used to evaluate the overall quality of the acquired images. The average BRISQUE values derived from the four areas were 39.87 and 46.44 in CR and VR, respectively. The difference between the two values was approximately 1.16, and the difference between the resulting values increased, particularly in the bone CT image, where metal artifacts were observed. In addition, we confirmed that the COV value improved by 2.20 times on average when using CR compared to VR. This study proved that CR is useful in reconstructing bone CT 3D images and that various applications in the diagnostic medical field will be possible.

Improving the Feasibility of Mitigating Phase Errors in 4DCT caused by a Random Reference Breathing Pattern in the Quasar Phantom and the Role of Slice Thickness.

The study aimed to validate a method for minimizing phase errors by combining full-length lung 4DCT (f4DCT) scans with shorter tumor-restricted 4DCT (s4DCT) scans. It assessed the feasibility of integrating two scans one covering the entire phantom length and the other focused on the tumor area. The study also evaluated the impact of Maximum Intensity Projection (MIP) volume and imaging dose for different slice thicknesses (2.5mm and 1.25mm) in both full-length and short target-restricted 4DCT scans. The study utilized the Quasar Programmable Respiratory Motion Phantom, simulating tumor motion with a variable lung insert. The setup included a tumor replica and a six-dot IR reflector marker on the breathing platform. The objective was to analyze volume differences in fMIP_2.5mm compared to sMIP_1.25mm within their respective 4D_MIP CT series. This involved varying breathing periods (2.5s, 3.0s, 4.0s, and 5.0s) and longitudinal tumor sizes (6mm, 8mm, and 10mm). The study also assessed exposure time and expected CTDIvol of s4D_2.5mm and s4D_1.25mm for different breathing periods (5.0s to 2.0s) in the sinusoidal wave motion of the six-dot marker on the breathing platform. Conducting two consecutive 4DCT scans is viable for patients with challenging breathing patterns or when the initial lung tumor scan is in close proximity to the tumor location, eliminating the need for an additional full-length 4DCT. The analysis involves assessing MIP volume, imaging dose (CTDIvol), and exposure time. Longitudinal tumor shifts for 6mm are [16.6-17.2] in fMIP_2.5mm and [16.8-17.5] in sMIP_1.25mm, for 8mm [17.2-18.3] in fMIP_2.5mm and [17.8-18.4] in sMIP_1.25mm, and for 10mm [19-19.9] in fMIP_2.5mm and [19.4-20] in sMIP_1.25mm (p≥ 0.005), respectively. The Quasar Programmable Respiratory Motion Phantom accurately replicated varied breathing patterns and tumor motions. Comprehensive analysis was facilitated through detailed manual segmentation of Internal Target Volumes and Internal Gross Target Volumes.

Quality Assurance in Radiotherapy (RT)-Specific Trials: Indian Scenario

AimsThere is evidence that proper radiotherapy trial quality assurance (RTTQA) translates into improved outcomes for patients. However, the practice of RTTQA is heterogeneous and implemented in a diverse manner across trials. In this paper, we review the RTTQA report for randomised trials (RCT) conducted in India and present our experience with RTTQA for various clinical trials and highlight the key achievements and challenges. Materials and methodsSearch was performed using the keywords and the variations thereof for “radiotherapy” and author affiliations from India, its states and major metropolitan cities. Pubmed search filters were used to restrict results to RCT published in the past 5 years (2019–2024). Reporting of RTTQA procedures from publications and protocols was documented along with the protocol-specified dosimetric goals. We also evaluated a few clinical trials performed in the Department of Radiation Oncology at Tata Medical Center. The different RTTQA procedures and results for four representative clinical trials have been described. ResultsA formal RTTQA process was reported by only one out of 24 randomised controlled trials and formal dosimetric goals were pre-specified by 9 of 13 trials where IMRT was used as treatment. RTTQA requirements were tailored for each clinical trial at Tata Medical Center. For the HYPORT trial, the RTTQA process focused on ensuring the matchline doses were homogenous. HYPORT B trial commissioned the use of a simultaneous integrated boost technique which emphasised conformal avoidance of dose spillage to contralateral breast and lung. HYPORT Adjuvant and PROPARA trials are multicentre clinical trials. While HYPORT Adjuvant focussed on ensuring that the dose delivery met the predefined constraints, segmentation of the target volume was important for the PROPARA trial. ConclusionWe demonstrate different RTTQA procedures required for representative clinical trials and highlight key challenges encountered.

Gender Estimation from Morphometric Measurements of Mandibular Lingula by Using Machine Learning Algorithms and Artificial Neural Networks

Background: Sex determination from the bones is of great importance for forensic medicine and anthropology. The mandible is highly valued because it is the strongest, largest and most dimorphic bone in the skull. Aim: Our aim in this study is gender estimation with morphometric measurements taken from mandibular lingula, an important structure on the mandible, by using machine learning algorithms and artificial neural networks. Methods: Cone beam computed tomography images of the mandibular lingula were obtained by retrospective scanning from the Picture Archiving Communication Systems of the Department of Oral, Dental and Maxillofacial Radiology, Faculty of Dentistry, İnönü University. Images scanned in Digital Imaging and Communications in Medicine (DICOM) format were transferred to RadiAnt DICOM Viewer (Version: 2020.2). The images were converted to 3-D format by using the 3D Volume Rendering console of the program. Eight anthropometric parameters were measured bilaterally from these 3-D images based on the mandibular lingula. Results: The results of the machine learning algorithms analyzed showed that the highest accuracy was 0.88 with Random Forest and Gaussian Naive Bayes algorithm. Accuracy rates of other parameters ranged between 0.78 and 0.88. Conclusions: As a result of the study, it is thought that mandibular lingula-centered morphometric measurements can be used for gender determination as well as bones such as the pelvis and skull as they were found to be highly accurate. This study also provides information on the anatomical position of the lingula according to gender in Turkish society. The results can be important for oral-dental surgeons, anthropologists, and forensic experts.

3D revisualization: a new method to revisit segmented data.

3D visualization and segmentation are increasingly widely used in physical, biological and medical science, facilitating advanced investigative methodologies. However, the integration and reproduction of segmented volumes or results across the spectrum of mainstream 3D visualization platforms remain hindered by compatibility constraints. These barriers not only challenge the replication of findings but also obstruct the process of cross-validating the accuracy of 3D visualization outputs. To address this gap, we developed an innovative revisualization method implemented within the open-source framework of Drishti, a 3D visualization software. Leveraging four animal samples alongside three mainstream 3D visualization platforms as case studies, our method demonstrates the seamless transferability of segmented results into Drishti. This capability effectively fosters a new avenue for authentication and enhanced scrutiny of segmented data. By facilitating this interoperability, our approach underscores the potential for significant advancements in accuracy validation and collaborative research efforts across diverse scientific domains.

Uncertainty estimation- and attention-based semi-supervised models for automatically delineate clinical target volume in CBCT images of breast cancer

ObjectivesAccurate segmentation of the clinical target volume (CTV) of CBCT images can observe the changes of CTV during patients' radiotherapy, and lay a foundation for the subsequent implementation of adaptive radiotherapy (ART). However, segmentation is challenging due to the poor quality of CBCT images and difficulty in obtaining target volumes. An uncertainty estimation- and attention-based semi-supervised model called residual convolutional block attention-uncertainty aware mean teacher (RCBA-UAMT) was proposed to delineate the CTV in cone-beam computed tomography (CBCT) images of breast cancer automatically.MethodsA total of 60 patients who undergone radiotherapy after breast-conserving surgery were enrolled in this study, which involved 60 planning CTs and 380 CBCTs. RCBA-UAMT was proposed by integrating residual and attention modules in the backbone network 3D UNet. The attention module can adjust channel and spatial weights of the extracted image features. The proposed design can train the model and segment CBCT images with a small amount of labeled data (5%, 10%, and 20%) and a large amount of unlabeled data. Four types of evaluation metrics, namely, dice similarity coefficient (DSC), Jaccard, average surface distance (ASD), and 95% Hausdorff distance (95HD), are used to assess the model segmentation performance quantitatively.ResultsThe proposed method achieved average DSC, Jaccard, 95HD, and ASD of 82%, 70%, 8.93, and 1.49 mm for CTV delineation on CBCT images of breast cancer, respectively. Compared with the three classical methods of mean teacher, uncertainty-aware mean-teacher and uncertainty rectified pyramid consistency, DSC and Jaccard increased by 7.89–9.33% and 14.75–16.67%, respectively, while 95HD and ASD decreased by 33.16–67.81% and 36.05–75.57%, respectively. The comparative experiment results of the labeled data with different proportions (5%, 10% and 20%) showed significant differences in the DSC, Jaccard, and 95HD evaluation indexes in the labeled data with 5% versus 10% and 5% versus 20%. Moreover, no significant differences were observed in the labeled data with 10% versus 20% among all evaluation indexes. Therefore, we can use only 10% labeled data to achieve the experimental objective.ConclusionsUsing the proposed RCBA-UAMT, the CTV of breast cancer CBCT images can be delineated reliably with a small amount of labeled data. These delineated images can be used to observe the changes in CTV and lay the foundation for the follow-up implementation of ART.

Probability maps for deep learning-based head and neck tumor segmentation: Graphical User Interface design and test

BackgroundThe different tumor appearance of head and neck cancer across imaging modalities, scanners, and acquisition parameters accounts for the highly subjective nature of the manual tumor segmentation task. The variability of the manual contours is one of the causes of the lack of generalizability and the suboptimal performance of deep learning (DL) based tumor auto-segmentation models. Therefore, a DL-based method was developed that outputs predicted tumor probabilities for each PET-CT voxel in the form of a probability map instead of one fixed contour. The aim of this study was to show that DL-generated probability maps for tumor segmentation are clinically relevant, intuitive, and a more suitable solution to assist radiation oncologists in gross tumor volume segmentation on PET-CT images of head and neck cancer patients. MethodA graphical user interface (GUI) was designed, and a prototype was developed to allow the user to interact with tumor probability maps. Furthermore, a user study was conducted where nine experts in tumor delineation interacted with the interface prototype and its functionality. The participants’ experience was assessed qualitatively and quantitatively. ResultsThe interviews with radiation oncologists revealed their preference for using a rainbow colormap to visualize tumor probability maps during contouring, which they found intuitive. They also appreciated the slider feature, which facilitated interaction by allowing the selection of threshold values to create single contours for editing and use as a starting point. Feedback on the prototype highlighted its excellent usability and positive integration into clinical workflows. ConclusionsThis study shows that DL-generated tumor probability maps are explainable, transparent, intuitive and a better alternative to the single output of tumor segmentation models.

A Three-Dimensional Morphological Assessment for Evaluating Pre- and Postsurgery in Cervical Laminoplasty

ObjectiveThe efficacy of medical treatments and the changes in radiological imaging before and after treatment have consistently remained pivotal factors. This is particularly critical for surgical procedures, where precise evaluation of disparities pre- and post-surgery or the accuracy of implantation is paramount. Based on three-dimensional morphological interests, we provide an automatic quantification evaluation method that delivers an evident base for assessing the outcomes of a widely employed surgical technique, cervical laminoplasty. MethodsThe sample study included patients who underwent cervical laminoplasty for cervical spondylotic myelopathy/ossification of the longitudinal ligament. We present a superimposition method that facilitates a unique and precise assessment between pre and post-surgery. The degree of expansion was evaluated by the canal volume increase and canal expansion rate after surgery. ResultsThere were 31 patients with 112 vertebral segments measured. The target cervical's pre- and post-operative canal areas were 122.63 ± 30.34 and 196.50 ± 37.10 mm2, respectively (P < 0.001). The average cervical canal expansion rate was 64.42%. The expansion effect of C5 cervical laminoplasty was the maximum (71.01%), and the canal volume of other segments expanded by approximately 60%. The functional outcomes demonstrated significant improvements in symptoms. ConclusionsThe quantification evaluation method can be utilized for any morphology changes before and after laminoplasty, as it does not lead to errors or variations from different inspection machines or human factors. The automatic method delivers an evident base for assessing the outcomes of a widely employed surgical technique.

A temporal enhanced semi-supervised training framework for needle segmentation in 3D ultrasound images

Objective. Automated biopsy needle segmentation in 3D ultrasound images can be used for biopsy navigation, but it is quite challenging due to the low ultrasound image resolution and interference similar to the needle appearance. For 3D medical image segmentation, such deep learning networks as convolutional neural network and transformer have been investigated. However, these segmentation methods require numerous labeled data for training, have difficulty in meeting the real-time segmentation requirement and involve high memory consumption. Approach. In this paper, we have proposed the temporal information-based semi-supervised training framework for fast and accurate needle segmentation. Firstly, a novel circle transformer module based on the static and dynamic features has been designed after the encoders for extracting and fusing the temporal information. Then, the consistency constraints of the outputs before and after combining temporal information are proposed to provide the semi-supervision for the unlabeled volume. Finally, the model is trained using the loss function which combines the cross-entropy and Dice similarity coefficient (DSC) based segmentation loss with mean square error based consistency loss. The trained model with the single ultrasound volume input is applied to realize the needle segmentation in ultrasound volume. Main results. Experimental results on three needle ultrasound datasets acquired during the beagle biopsy show that our approach is superior to the most competitive mainstream temporal segmentation model and semi-supervised method by providing higher DSC (77.1% versus 76.5%), smaller needle tip position (1.28 mm versus 1.87 mm) and length (1.78 mm versus 2.19 mm) errors on the kidney dataset as well as DSC (78.5% versus 76.9%), needle tip position (0.86 mm versus 1.12 mm) and length (1.01 mm versus 1.26 mm) errors on the prostate dataset. Significance. The proposed method can significantly enhance needle segmentation accuracy by training with sequential images at no additional cost. This enhancement may further improve the effectiveness of biopsy navigation systems.

Intraretinal Hyper-Reflective Foci Are Almost Universally Present and Co-Localize With Intraretinal Fluid in Diabetic Macular Edema.

In diabetic macular edema (DME), hyper-reflective foci (HRF) has been linked to disease severity and progression. Using an automated approach, we aimed to investigate the baseline distribution of HRF in DME and their co-localization with cystoid intraretinal fluid (IRF). Baseline spectral-domain optical coherence tomography (SD-OCT) volume scans (N = 1527) from phase III clinical trials YOSEMITE (NCT03622580) and RHINE (NCT03622593) were segmented using a deep-learning-based algorithm (developed using B-scans from BOULEVARD NCT02699450) to detect HRF. The HRF count and volume were assessed. HRF distributions were analyzed in relation to best-corrected visual acuity (BCVA), central subfield thickness (CST), and IRF volume in quartiles, and Diabetic Retinopathy Severity Scores (DRSS) in groups. Co-localization of HRF with IRF was calculated in the central 3-mm diameter using the en face projection. HRF were present in most patients (up to 99.7%). Median (interquartile range [IQR]) HRF volume within the 3-mm diameter Early Treatment Diabetic Retinopathy Study ring was 1964.3 (3325.2) pL, and median count was 64.0 (IQR = 96.0). Median HRF volumes were greater with decreasing BCVA (nominal P = 0.0109), and increasing CST (nominal P < 0.0001), IRF (nominal P < 0.0001), and DRSS up to very severe nonproliferative diabetic retinopathy (nominal P < 0.0001). HRF co-localized with IRF in the en face projection. Using automated HRF segmentation of full SD-OCT volumes, we observed that HRF are a ubiquitous feature in DME and exhibit relationships with BCVA, CST, IRF, and DRSS, supporting a potential link to disease severity. The spatial distribution of HRF closely followed that of IRF.

A Comparative Study of AI-Based Automated and Manual Postprocessing of Head and Neck CT Angiography: An Independent External Validation with Multi-Vendor and Multi-Center Data.

To externally validate the performance of automated postprocessing (AP) on head and neck CT Angiography (CTA) and compare it with manual postprocessing (MP). This retrospective study included head and neck CTA-exams of patients from three tertiary hospitals acquired on CT scanners from five manufacturers. AP was performed by CerebralDoc. The image quality was assessed using Likert scales, and the qualitative and quantitative diagnostic performance of arterial stenosis and aneurysm, postprocessing time, and scanning radiation dose were also evaluated. A total of 250 patients were included. Among these, 55 patients exhibited significant stenosis (≥ 50%), and 33 patients had aneurysms, diagnosed using original CTA datasets and corresponding multiplanar reconstructions as the reference. While the scores of the V4 segment and the edge of the M1 segment on volume rendering (VR), as well as the C4 segment on maximum intensity projection (MIP), were significantly lower with AP compared to MP across vendors (all P < 0.05), most scores in AP demonstrated image quality that was either superior to or comparable with that of MP. Furthermore, the diagnostic performance of AP was either superior to or comparable with that of MP. Moreover, AP also exhibited advantages in terms of postprocessing time and radiation dose when compared to MP (P < 0.001). The AP of CerebralDoc presents clear advantages over MP and holds significant clinical value. However, further optimization is required in the image quality of the V4 and M1 segments on VR as well as the C4 segment on MIP.

A transformer-guided cross-modality adaptive feature fusion framework for esophageal gross tumor volume segmentation

Background and ObjectiveAccurate segmentation of esophageal gross tumor volume (GTV) indirectly enhances the efficacy of radiotherapy for patients with esophagus cancer. In this domain, learning-based methods have been employed to fuse cross-modality positron emission tomography (PET) and computed tomography (CT) images, aiming to improve segmentation accuracy. This fusion is essential as it combines functional metabolic information from PET with anatomical information from CT, providing complementary information. While the existing three-dimensional (3D) segmentation method has achieved state-of-the-art (SOTA) performance, it typically relies on pure-convolution architectures, limiting its ability to capture long-range spatial dependencies due to convolution's confinement to a local receptive field. To address this limitation and further enhance esophageal GTV segmentation performance, this work proposes a transformer-guided cross-modality adaptive feature fusion network, referred to as TransAttPSNN, which is based on cross-modality PET/CT scans. MethodsSpecifically, we establish an attention progressive semantically-nested network (AttPSNN) by incorporating the convolutional attention mechanism into the progressive semantically-nested network (PSNN). Subsequently, we devise a plug-and-play transformer-guided cross-modality adaptive feature fusion model, which is inserted between the multi-scale feature counterparts of a two-stream AttPSNN backbone (one for the PET modality flow and another for the CT modality flow), resulting in the proposed TransAttPSNN architecture. ResultsThrough extensive four-fold cross-validation experiments on the clinical PET/CT cohort. The proposed approach acquires a Dice similarity coefficient (DSC) of 0.76 ± 0.13, a Hausdorff distance (HD) of 9.38 ± 8.76 mm, and a Mean surface distance (MSD) of 1.13 ± 0.94 mm, outperforming the SOTA competing methods. The qualitative results show a satisfying consistency with the lesion areas. ConclusionsThe devised transformer-guided cross-modality adaptive feature fusion module integrates the strengths of PET and CT, effectively enhancing the segmentation performance of esophageal GTV. The proposed TransAttPSNN has further advanced the research of esophageal GTV segmentation.

FaceFolds: Meshed Radiance Manifolds for Efficient Volumetric Rendering of Dynamic Faces

3D rendering of dynamic face captures is a challenging problem, and it demands improvements on several fronts---photorealism, efficiency, compatibility, and configurability. We present a novel representation that enables high-quality volumetric rendering of an actor's dynamic facial performances with minimal compute and memory footprint. It runs natively on commodity graphics soft- and hardware, and allows for a graceful trade-off between quality and efficiency. Our method utilizes recent advances in neural rendering, particularly learning discrete radiance manifolds to sparsely sample the scene to model volumetric effects. We achieve efficient modeling by learning a single set of manifolds for the entire dynamic sequence, while implicitly modeling appearance changes as temporal canonical texture. We export a single layered mesh and view-independent RGBA texture video that is compatible with legacy graphics renderers without additional ML integration. We demonstrate our method by rendering dynamic face captures of real actors in a game engine, at comparable photorealism to state-of-the-art neural rendering techniques at previously unseen frame rates.

MRI radiomics captures early treatment response in patient-derived organoid endometrial cancer mouse models.

Radiomics can capture microscale information in medical images beyond what is visible to the naked human eye. Using a clinically relevant mouse model for endometrial cancer, the objective of this study was to develop and validate a radiomic signature (RS) predicting response to standard chemotherapy. Mice orthotopically implanted with a patient-derived grade 3 endometrioid endometrial cancer organoid model (O-PDX) were allocated to chemotherapy (combined paclitaxel/carboplatin, n=11) or saline/control (n=13). During tumor progression, the mice underwent weekly T2-weighted (T2w) magnetic resonance imaging (MRI). Segmentation of primary tumor volume (vMRI) allowed extraction of radiomic features from whole-volume tumor masks. A radiomic model for predicting treatment response was derived employing least absolute shrinkage and selection operator (LASSO) statistics at endpoint images in the orthotopic O-PDX (RS_O), and subsequently applied on the earlier study timepoints (RS_O at baseline, and week 1-3). For external validation, the radiomic model was tested in a separate T2w-MRI dataset on segmented whole-volume subcutaneous tumors (RS_S) from the same O-PDX model, imaged at three timepoints (baseline, day 3 and day 10/endpoint) after start of chemotherapy (n=8 tumors) or saline/control (n=8 tumors). The RS_O yielded rapidly increasing area under the receiver operating characteristic (ROC) curves (AUCs) for predicting treatment response from baseline until endpoint; AUC=0.38 (baseline); 0.80 (week 1), 0.85 (week 2), 0.96 (week 3) and 1.0 (endpoint). In comparison, vMRI yielded AUCs of 0.37 (baseline); 0.69 (w1); 0.83 (week 2); 0.92 (week 3) and 0.97 (endpoint). When tested in the external validation dataset, RS_S yielded high accuracy for predicting treatment response at day10/endpoint (AUC=0.85) and tended to yield higher AUC than vMRI (AUC=0.78, p=0.18). Neither RS_S nor vMRI predicted response at day 3 in the external validation set (AUC=0.56 for both). We have developed and validated a radiomic signature that was able to capture chemotherapeutic treatment response both in an O-PDX and in a subcutaneous endometrial cancer mouse model. This study supports the promising role of preclinical imaging including radiomic tumor profiling to assess early treatment response in endometrial cancer models.

Distinguishing Laterality in Brain Injury in Rabbit Fetal Magnetic Resonance Imaging Using Novel Volume Rendering Techniques

Introduction: Our laboratory has been exploring the MRI detection of fetal brain injury, which previously provided a prognostic biomarker for newborn hypertonia in an animal model of cerebral palsy (CP). The biomarker relies on distinct patterns of diffusion-weighted imaging-defined apparent diffusion coefficient (ADC) in fetal brains during uterine hypoxia-ischemia (H-I). Despite the challenges posed by small brains and tissue acquisition, our objective was to differentiate between left and right brain ADC changes. Methods: A novel aspect involved utilizing three-dimensional rendering techniques to refine ADC measurements within spheroids encompassing fetal brain tissue. 25-day gestation age of rabbit fetuses underwent global hypoxia due to maternal uterine ischemia. Results: Successful differentiation of left and right brain regions was achieved in 28% of the fetal brains. Ordinal analysis revealed predominantly higher ADC on the left side compared to the right at baseline and across the entire time series. During H-I and reperfusion-reoxygenation, the right side exhibited a favored percentage change. Among these fetal brains, 73% exhibited the ADC pattern predictive of hypertonia. No significant differences between left and right sides were observed in patterns predicting hypertonia, except for one timepoint during H-I. This study also highlights a balance between left-sided and right-sided alterations within the population. Conclusion: This study emphasizes the importance of investigating laterality and asymmetric hemispheric lesions for early diagnosis of brain injury, leading to CP. The technological limitations in obtaining a clear picture of the entire fetal brain for every fetus mirror the challenges encountered in human studies.