Geometric Metrics Research Articles

Development and comprehensive evaluation of a national DBCG consensus-based auto-segmentation model for lymph node levels in breast cancer radiotherapy

Background and purposeThis study aimed at training and validating a multi-institutional deep learning (DL) auto segmentation model for nodal clinical target volume (CTVn) in high-risk breast cancer (BC) patients with both training and validation dataset created with multi-institutional participation, with the overall aim of national clinical implementation in Denmark. Materials and methodsA gold standard (GS) dataset and a high-quality training dataset were created by 21 BC delineation experts from all radiotherapy centres in Denmark. The delineations were created according to ESTRO consensus delineation guidelines. Four models were trained: One per laterality and extension of CTVn internal mammary nodes. The DL models were tested quantitatively in their own test-set and in relation to interobserver variation (IOV) in the GS dataset with geometrical metrics, such as the Dice Similarity Coefficient (DSC). A blinded qualitative evaluation was conducted with a national board, presented to both DL and manual delineations. ResultsA median DSC > 0.7 was found for all, except the CTVn interpectoral node in one of the models. In the qualitative evaluation ‘no corrections needed’ were acquired for 297 (36 %) in the DL structures and 286 (34 %) for manual delineations. A higher rate of ‘major corrections’ and ‘easier to start from scratch’ was found in the manual delineations. The models performed within the IOV of an expert group, with two exceptions. ConclusionDL models were developed on a national consensus cohort and performed on par with the IOV between BC experts and had a comparable or higher clinical acceptance than expert manual delineations.

Incorporating patient-specific information for the development of rectal tumor auto-segmentation models for online adaptive magnetic resonance Image-guided radiotherapy

Background and purposeIn online adaptive magnetic resonance image (MRI)-guided radiotherapy (MRIgRT), manual contouring of rectal tumors on daily images is labor-intensive and time-consuming. Automation of this task is complex due to substantial variation in tumor shape and location between patients. The aim of this work was to investigate different approaches of propagating patient-specific prior information to the online adaptive treatment fractions to improve deep-learning based auto-segmentation of rectal tumors. Materials and methods243 T2-weighted MRI scans of 49 rectal cancer patients treated on the 1.5T MR-Linear accelerator (MR-Linac) were utilized to train models to segment rectal tumors. As benchmark, an MRI_only auto-segmentation model was trained. Three approaches of including a patient-specific prior were studied: 1. include the segmentations of fraction 1 as extra input channel for the auto-segmentation of subsequent fractions, 2. fine-tuning of the MRI_only model to fraction 1 (PSF_1) and 3. fine-tuning of the MRI_only model on all earlier fractions (PSF_cumulative). Auto-segmentations were compared to the manual segmentation using geometric similarity metrics. Clinical impact was assessed by evaluating post-treatment target coverage. ResultsAll patient-specific methods outperformed the MRI_only segmentation approach. Median 95th percentile Hausdorff (95HD) were 22.0 (range: 6.1–76.6) mm for MRI_only segmentation, 9.9 (range: 2.5–38.2) mm for MRI+prior segmentation, 6.4 (range: 2.4–17.8) mm for PSF_1 and 4.8 (range: 1.7–26.9) mm for PSF_cumulative. PSF_cumulative was found to be superior to PSF_1 from fraction 4 onward (p = 0.014). ConclusionPatient-specific fine-tuning of automatically segmented rectal tumors, using images and segmentations from all previous fractions, yields superior quality compared to other auto-segmentation approaches.

Evolutions of semantic consistency in research topic via contextualized word embedding

Topic evolution has been studied extensively in the field of the science of science. This study first analyzes topic evolution pattern from topics’ semantic consistency in the semantic vector space, and explore its possible causes. Specifically, we extract papers in the computer science field from Microsoft Academic Graph as our dataset. We propose a novel method for encoding a topic with numerous Contextualized Word Embeddings (CWE), in which the title and abstract fields of papers studying the topic is taken as its context. Subsequently, we employ three geometric metrics to analyze topics’ semantic consistency over time, from which the influence of the anisotropy of CWE is excluded. The K-Means clustering algorithm is employed to identify four general evolution patterns of semantic consistency, that is, semantic consistency increases (IM), decreases (DM), increases first and then decreases (Inverted U-shape), and decreases first and then increases (U-shape). We also find that research methods tend to show DM and U-shape, but research questions tend to be IM and Inverted U-shape. Finally, we further utilize the regression analysis to explore whether and, if so, how a series of key features of a topic affect its semantic consistency. Importantly, semantic consistency of a topic varies inversely with the semantic similarity between the topic and other topics. Overall, this study sheds light on the evolution law of topics, and helps researchers to understand these patterns from a geometric perspective.

Analyzing the Relationship between Dose and Geometric Agreement Metrics for Auto-Contouring in Head and Neck Normal Tissues.

This study aimed to determine the relationship between geometric and dosimetric agreement metrics in head and neck (H&N) cancer radiotherapy plans. A total 287 plans were retrospectively analyzed, comparing auto-contoured and clinically used contours using a Dice similarity coefficient (DSC), surface DSC (sDSC), and Hausdorff distance (HD). Organs-at-risk (OARs) with ≥200 cGy dose differences from the clinical contour in terms of Dmax (D0.01cc) and Dmean were further examined against proximity to the planning target volume (PTV). A secondary set of 91 plans from multiple institutions validated these findings. For 4995 contour pairs across 19 OARs, 90% had a DSC, sDSC, and HD of at least 0.75, 0.86, and less than 7.65 mm, respectively. Dosimetrically, the absolute difference between the two contour sets was <200 cGy for 95% of OARs in terms of Dmax and 96% in terms of Dmean. In total, 97% of OARs exhibiting significant dose differences between the clinically edited contour and auto-contour were within 2.5 cm PTV regardless of geometric agreement. There was an approximately linear trend between geometric agreement and identifying at least 200 cGy dose differences, with higher geometric agreement corresponding to a lower fraction of cases being identified. Analysis of the secondary dataset validated these findings. Geometric indices are approximate indicators of contour quality and identify contours exhibiting significant dosimetric discordance. For a small subset of OARs within 2.5 cm of the PTV, geometric agreement metrics can be misleading in terms of contour quality.

Tales of twin cities: what are climate analogues good for?

This article provides an epistemological assessment of climate analogue methods, with specific reference to the use of spatial analogues in the study of the future climate of target locations. Our contention is that, due to formal and conceptual inadequacies of geometrical dissimilarity metrics and the loss of relevant information, especially when reasoning from the physical to the socio-economical level, purported inferences from climate analogues of the spatial kind we consider here prove limited in a number of ways. Indeed, we formulate five outstanding problems concerning the search for best analogues, which we call the problem of non-uniqueness of the source, problem of non-uniqueness of the target, problem of average, problem of non-causal correlations and problem of inferred properties, respectively. In the face of such problems, we then offer two positive recommendations for a fruitful application of this methodology to the assessment of impact, adaptation and vulnerability studies of climate change, especially in the context of what we may prosaically dub “twin cities”. Arguably, such recommendations help decision-makers constrain the set of plausible climate analogues by integrating local knowledge relevant to the locations of interest.

Unsupervised domain adaptation with hard-sample dividing and processing strategy

In Unsupervised Domain Adaptation(UDA), the hard samples cause serious negative transfer to model performance. However, the existing UDA methods don’t fully pay attention to the fine-grained processing of hard samples in target and source domains. To address these problems, the UDA framework with Hard-sample Dividing and Processing Strategy (HDPS) is proposed in this paper. In HDPS, we define sample division criteria in target domain and source domain respectively and divide the samples into easy samples, hard samples with low confidence and hard samples with high confidence. We further design systematic processing strategy of the three types of samples. Specifically, we firstly define geometric metrics based on the class prototype to divide the hard samples and eliminate them in source domain. We further design two weight allocation strategies based on sample classification entropy in target domain. Finally, we build a teacher-student guidance mechanism to learn the discriminative features of hard samples. The HDPS framework can be easily loaded to other methods as a plug-in, can promote the diversity feature learning of samples, and significantly improve the discriminant performance of the model. Extensive experiments on six benchmark datasets verify that HDPS is effective and superior to most existing UDA methods.

Accurate identification of individuals with subjective cognitive decline using 3D regional fractal dimensions on structural magnetic resonance imaging

Background and ObjectiveAccurate identification of individuals with subjective cognitive decline (SCD) is crucial for early intervention and prevention of neurodegenerative diseases. Fractal dimensionality (FD) has emerged as a robust and replicable measure, surpassing traditional geometric metrics, in characterizing the intricate fractal geometrical properties of brain structure. Nevertheless, the effectiveness of FD in identifying individuals with SCD remains largely unclear. A 3D regional FD method can be suggested to characterize and quantify the spatial complexity of the precise gray matter, providing insights into cognitive aging and aiding in the automated identification of individuals with SCD. MethodsThis study introduces a novel integer ratio based 3D box-counting fractal analysis (IRBCFA) to quantify regional fractal dimensions (FDs) in structural magnetic resonance imaging (MRI) data. The innovative method overcomes limitations of conventional box-counting techniques by accommodating arbitrary box sizes, thereby enhancing the precision of FD estimation in small, yet neurologically significant, brain regions. ResultsThe application of IRBCFA to two publicly available datasets, OASIS-3 and ADNI, consisting of 520 and 180 subjects, respectively. The method identified discriminative regions of interest (ROIs) predominantly within the limbic system, fronto-parietal region, occipito-temporal region, and basal ganglia-thalamus region. These ROIs exhibited significant correlations with cognitive functions, including executive functioning, memory, social cognition, and sensory perception, suggesting their potential as neuroimaging markers for SCD. The identification model trained on these ROIs demonstrated exceptional performance achieving over 93 % accuracy on the discovery dataset and exceeding 87 % on the independent testing dataset. Furthermore, an exchange experiment between datasets revealed a substantial overlap in discriminative ROIs, highlighting the robustness of our method across diverse populations. ConclusionOur findings indicate that IRBCFA can serve as a valuable tool for quantifying the spatial complexity of gray matter, providing insights into cognitive aging and aiding in the automated identification of individuals with SCD. The demonstrated generalizability and robustness of this method position it as a promising tool for neurodegenerative disease research and offer potential for clinical applications.

Enhancing Precision in Cardiac Segmentation for Magnetic Resonance-Guided Radiation Therapy Through Deep Learning

IntroductionCardiac substructure dose metrics are more strongly linked to late cardiac morbidities than whole-heart metrics. MR-guided radiation therapy (MRgRT) enables substructure visualization during daily localization, allowing potential for enhanced cardiac sparing. We extend a publicly available state-of-the-art deep learning (DL) framework, nnU-Net, to incorporate self-distillation (nnU-Net.wSD) for substructure segmentation for MRgRT. MethodsEighteen (Institute A) patients who underwent thoracic or abdominal radiation therapy on a 0.35 T MR-guided linac were retrospectively evaluated. On each image, one of two radiation oncologists delineated reference contours of 12 cardiac substructures (chambers, great vessels, and coronary arteries) used to train (n=10), validate (n=3), and test (n=5) nnU-Net.wSD leveraging a teacher-student network and comparing to standard 3D U-Net. The impact of using simulation data or including 3-4 daily images for augmentation during training was evaluated for nnU-Net.wSD. Geometric metrics (Dice similarity coefficient (DSC), mean distance to agreement (MDA), and 95% Hausdorff distance (HD95)), visual inspection, and clinical dose volume histograms (DVHs) were evaluated. To determine generalizability, Institute A's model was tested on an unlabeled dataset from Institute B (n=22) and evaluated via consensus scoring and volume comparisons. ResultsnnU-Net.wSD yielded a DSC (reported mean ± standard deviation) of 0.65±0.25 across the 12 substructures (Chambers: 0.85±0.05, Great Vessels: 0.67±0.19, and Coronary Arteries 0.33±0.16, mean MDA <3 mm, and mean HD95 <9 mm) while outperforming the 3D U-Net (0.583±0.28, p<0.01). Leveraging fractionated data for augmentation improved over a single MR-SIM timepoint (0.579±0.29, p<0.01). Predicted contours yielded DVHs that closely matched the clinical treatment plans where mean and D0.03cc doses deviated by 0.32±0.5 Gy and 1.42±2.6 Gy respectively. No statistically significant differences between Institute A and B volumes (p>0.05) for 11 of 12 substructures with larger volumes requiring minor changes and coronary arteries exhibiting more variability. ConclusionsThis work is a critical step to rapid and reliable cardiac substructure segmentation to improve cardiac sparing in low-field MRgRT.

Matching the building footprints of different vector spatial datasets at a similar scale based on one-class support vector machines

Automatic matching of multisource data is an important technique for achieving change detection, fusion and updating spatial data. However, most current learning methods for building footprint matching require a large number of samples, and labeling these samples is costly in terms of labor and time. Moreover, multisource building footprint data are complex and diverse leading to recognizing the different matching relationships is a hard task. Thus, this study proposes a learning-based method for recognizing multisource building footprints matching relationships by using a one-class support vector machine (OCSVM). The OCSVM was trained using only positive samples. First, a set of geometric indicators was designed to train a model and realize initial matching recognition. Then, a contextual metric was calculated based on the rough matching results, and geometric and contextual metrics were combined to train the model and realize relaxed matching recognition. Relaxed matching is an optimization process implemented after initial matching to recognize more relaxed matching relationships. In relaxed matching, a convex hull is used to recognize matching relationships besides 1:1, such as 1:n, m:1 and m:n. The experimental results showed that the proposed method outperformed indicator-weighted (weighted average) and learning-based matching methods, such as traditional SVMs and decision trees (DTs). The precision scores of the proposed model were 97.1%, 95% and 97.2% for the Wuhan (China), Beijing (China) and Richmond Hill (Canada) datasets, respectively. Furthermore, the proposed model identified the matching relationships of buildings with complex geometric features and high-density spatial distributions.

OATracker: Object-aware anti-occlusion 3D multiobject tracking for autonomous driving

Efficient and accurate 3D multiobject tracking (MOT) is essential for autonomous driving. However, existing methods still struggle with multiobject occlusion, in which interruptions or even loss of crucial tracked objects is likely. Therefore, a more intelligent object-aware anti-occlusion 3D MOT for autonomous driving, named OATracker, is developed in this paper. First, an updated 3D cascaded detection (3DCD) is adopted as the detection backbone, with incorporation of an updatable component based on the OpenPCDet framework to increase the longevity. Second, a reliable object state-constrained affinity (OSA) metric is designed that adaptively combines 3D geometric and motion metrics to dynamically measure objects’ size and distance in both spatial and temporal dimensions. Third, an object-aware trajectory management (OTM) module is developed for trajectory management and status updating. Specifically, the object identity-aware adaptive tracking life manager (ATLM) allows the tracker to adaptively manage the trajectories of captured objects through a confidence metric for each objects, and the ghost-aware trajectory postoptimizer (GTP) associates the historical and detected motion states, further overcoming id-switch and fragmentation problems. Finally, experiments on the KITTI-Car tracking benchmark and practical autonomous driving car application platform show that the OATracker can achieve competitive and robust tracking performance even in occlusion conditions.

Dosimetric impact of contour editing on CT and MRI deep-learning autosegmentation for brain OARs.

To establish the clinical applicability of deep-learning organ-at-risk autocontouring models (DL-AC) for brain radiotherapy. The dosimetric impact of contour editing, prior to model training, on performance was evaluated for both CT and MRI-based models. The correlation between geometric and dosimetric measures was also investigated to establish whether dosimetric assessment is required for clinical validation. CT and MRI-based deep learning autosegmentation models were trained using edited and unedited clinical contours. Autosegmentations were dosimetrically compared to gold standard contours for a test cohort. D1%, D5%, D50%, and maximum dose were used as clinically relevant dosimetric measures. The statistical significance of dosimetric differences between the gold standard and autocontours was established using paired Student's t-tests. Clinically significant cases were identified via dosimetric headroom to the OAR tolerance. Pearson's Correlations were used to investigate the relationship between geometric measures and absolute percentage dose changes for each autosegmentation model. Except for the right orbit, when delineated using MRI models, the dosimetric statistical analysis revealed no superior model in terms of the dosimetric accuracy between the CT DL-AC models or between the MRI DL-AC for any investigated brain OARs. The number of patients where the clinical significance threshold was exceeded was higher for the optic chiasm D1% than other OARs, for all autosegmentation models. A weak correlation was consistently observed between the outcomes of dosimetric and geometric evaluations. Editing contours before training the DL-AC model had no significant impact on dosimetry. The geometric test metrics were inadequate to estimate the impact of contour inaccuracies on dose. Accordingly, dosimetric analysis is needed to evaluate the clinical applicability of DL-AC models in the brain.

Automated visual quality assessment for virtual and augmented reality based digital twins

Virtual and augmented reality digital twins are becoming increasingly prevalent in a number of industries, though the production of digital-twin systems applications is still prohibitively expensive for many smaller organisations. A key step towards reducing the cost of digital twins lies in automating the production of 3D assets, however efforts are complicated by the lack of suitable automated methods for determining the visual quality of these assets. While visual quality assessment has been an active area of research for a number of years, few publications consider this process in the context of asset creation in digital twins. In this work, we introduce an automated decimation procedure using machine learning to assess the visual impact of decimation, a process commonly used in the production of 3D assets which has thus far been underrepresented in the visual assessment literature. Our model combines 108 geometric and perceptual metrics to determine if a 3D object has been unacceptably distorted during decimation. Our model is trained on almost 4, 000 distorted meshes, giving a significantly wider range of applicability than many models in the literature. Our results show a precision of over 97% against a set of test models, and performance tests show our model is capable of performing assessments within 2 minutes on models of up to 25, 000 polygons. Based on these results we believe our model presents both a significant advance in the field of visual quality assessment and an important step towards reducing the cost of virtual and augmented reality-based digital-twins.

Using Dry Dispersion Laser Diffraction to Assess Dispersibility in Spheronized Agglomerate Formulations.

In this study, dry dispersion laser diffraction was used to study the dispersibility of spheronized agglomerate formulations and identify geometric particle size metrics that correlated well with aerodynamic particle size distribution (APSD). Eleven unique batches of agglomerates were prepared for both laser diffraction and cascade impaction testing. Correlations between the particle size distribution (PSD) and aerodynamic particle size distribution (APSD) metrics for the eleven agglomerate batches were determined in a semi-empirical manner. The strongest correlation between APSD and PSD was observed between the impactor-sized mass (%ISM) and the cumulative PSD fraction <14.5 µm. The strongest correlation with fine particle fraction (FPF) was observed with the cumulative PSD fraction <0.99 micron (R-squared = 0.974). In contrast to the other APSD metrics, good correlations were not found between the mass median aerodynamic diameter (MMAD) and the cumulative PSD fractions. Overall, the implementation of laser diffraction as a surrogate for cascade impaction has the potential to streamline product development. Laser diffraction measurements offer savings in labor and turnaround time compared to cascade impaction.

Assessment of morphological similarities for the conservative Allen–Cahn and Cahn–Hilliard equations

Phase separations occur in various natural environments, such as materials sciences and biochemistry, and have been scientifically modeled in various fields. We often need to distinguish between two different phase separation phenomena induced by the conservative Allen–Cahn (CAC) and Cahn–Hilliard (CH) equations, which are used to describe natural phenomena. Both phase-field equations share the same properties, including phase-coarsening, mass conservation, and energy dissipation. These similarities make it difficult to determine which equation is responsible for the observed dynamics. However, there are morphological differences between the evolutionary behaviors of the two dynamics, despite their similarities.This work investigates how the dynamics can be characterized by the CAC and CH equations and proposes evaluation criteria to distinguish their features. We examine phase-coarsening in light of the morphological patterns of the phases. Specifically, we train a Convolutional Neural Network (CNN) with morphological patterns generated by the CAC and CH equations. We use Alexnet as one of the CNNs to classify the patterns. In addition, we introduce other geometric properties to characterize the phase-separation dynamics. For a comprehensive understanding of these dynamics, we propose geometrical evaluations such as a distance measure by inner product, area, arclength and anatomical complexity, as measures of similarity metrics. Furthermore, we test and investigate geometrical similarity metrics to discriminate between the CAC and CH equations through numerical experiments. The numerical evaluations of the similarity metrics show that the morphological difference between these two CAC and CH dynamics appears to be large in the middle stages of evolution, whereas the distinction is vague in the early and late stages of evolution. In addition, the geometric aspects of morphological difference and similarity are also confirmed numerically.

Real and virtual environments have comparable spatial memory distortions after scale and geometric transformations

ABSTRACT Boundaries define space, impacting spatial memory and neural representations. Unlike rodents, impact in humans is often tested using desktop virtual-reality (VR). This lacks self-motion cues, diminishing path-integration input. We replicated a desktop-VR study testing boundary impact on spatial memory for object locations using a physical, desktop-VR, and head-mounted-display-VR environment. Performance was measured by comparing participant responses to seven spatial distribution models using geometric or walking-path metrics. A weighted-linear combination of geometric models and a “place-cell-firing” model performed best, with identical fits across environments. Spatial representation appears differentially influenced by different boundary changes, but similarly across virtual and physical environments.

KarstNSim: A graph-based method for 3D geologically-driven simulation of karst networks

In karst aquifers, groundwater flow is highly influenced by the interconnected underground cavities and conduits that form the karst network. Modeling karst flows requires the use of spatially distributed approaches accounting for these networks. Their exploration is, however, often complex, and mapping them using indirect methods such as geophysical ones has proven challenging. To overcome these limitations, stochastically simulating discrete karst networks should account for the uncertainties on conduit position and geometry. Only a few existing methods can reproduce realistic and diverse karst morphologies. In this work, we propose a new algorithm, KarstNSim, for simulating discrete karst networks, that incorporates field data to generate a range of possible karst network geometries. It relies on the computation of the shortest path between the inlets and outlets of the network with the use of an anisotropic cost function defined on an n-nearest neighbor graph conformal to geological and structural heterogeneities. This cost function represents the physico-chemical processes that govern speleogenesis – such as erosion and chemical weathering – providing simplified control over the geometry of the generated networks. Our approach reproduces the vadose-phreatic partition visible in the karst networks, by generating sub-vertical conduits in the unsaturated zone and sub-horizontal ones in the saturated part. It encompasses geological parameters such as inception surfaces, fractures, permeability, and solubility of layers, along with considering the hydrological context of recharge by assigning relative weights to the inlets. We simulate various synthetic models to demonstrate the influence of different input parameters on the spatial organization of the past and present karst flows. We also apply KarstNSim to a real case study – the Ribeaucourt system (Meuse, France) – and use available data to generate realistic karst networks. The lack of knowledge on the Ribeaucourt network geometry is alleviated by the use of multiple scenarios corresponding to different sets of weights in the cost function, resulting in a variety of network morphologies which represent the uncertainty on the real geometry. The geometrical and topological metrics computed on the simulated networks generally overlap with those of real ones, suggesting the reliability of the method.

Custom-Trained Deep Learning-Based Auto-Segmentation for Male Pelvic Iterative CBCT on C-Arm Linear Accelerators

Purpose: To evaluate the clinical applicability of a commercial artificial intelligence (AI)-driven deep learning auto-segmentation (DLAS) tool on enhanced iterative cone-beam CT (iCBCT) acquisitions for intact prostate and prostate bed treatments. Methods and MaterialsDLAS models were trained using 116 iCBCT datasets with manually delineated organs-at-risk (OARs – bladder, femoral heads, and rectum) and target volumes (intact prostate and prostate bed) adhering to institution-specific contouring guidelines. An additional 25 intact prostate and prostate bed iCBCT datasets were utilized for model testing. Segmentation accuracy relative to a reference structure set was quantified using various geometric comparison metrics and qualitatively evaluated by trained physicists and physicians. These results were compared to those obtained for an additional DLAS-based model trained on planning CT (pCT) datasets and for a deformable image registration (DIR)-based automatic contour propagation method. Results: In most instances, statistically significant differences in the Dice similarity coefficient (DSC), 95% directed Hausdorff distance, and mean surface distance metrics were observed between the models, as the iCBCT-trained DLAS model outperformed the pCT-trained DLAS model and DIR-based method for all OARs and the intact prostate target volume. Mean DSC values for the proposed method were ≥0.90 for these volumes-of-interest. The iCBCT-trained DLAS model demonstrated a relatively suboptimal performance for the prostate bed segmentation, as the mean DSC value was <0.75 for this target contour. Overall, 90% of bladder, 93% of femoral head, 67% of rectum, and 92% of intact prostate contours generated by the proposed method were deemed clinically acceptable based on qualitative scoring, while approximately 63% of prostate bed contours required moderate or major manual editing to adhere to institutional contouring guidelines. Conclusion: The proposed method presents the potential for improved segmentation accuracy and efficiency when compared to the DIR-based automatic contour propagation method as commonly applied in CBCT-based dose evaluation and calculation studies.

Deep learning application for abdominal organs segmentation on 0.35 T MR-Linac images.

Linear accelerator (linac) incorporating a magnetic resonance (MR) imaging device providing enhanced soft tissue contrast is particularly suited for abdominal radiation therapy. In particular, accurate segmentation for abdominal tumors and organs at risk (OARs) required for the treatment planning is becoming possible. Currently, this segmentation is performed manually by radiation oncologists. This process is very time consuming and subject to inter and intra operator variabilities. In this work, deep learning based automatic segmentation solutions were investigated for abdominal OARs on 0.35 T MR-images. One hundred and twenty one sets of abdominal MR images and their corresponding ground truth segmentations were collected and used for this work. The OARs of interest included the liver, the kidneys, the spinal cord, the stomach and the duodenum. Several UNet based models have been trained in 2D (the Classical UNet, the ResAttention UNet, the EfficientNet UNet, and the nnUNet). The best model was then trained with a 3D strategy in order to investigate possible improvements. Geometrical metrics such as Dice Similarity Coefficient (DSC), Intersection over Union (IoU), Hausdorff Distance (HD) and analysis of the calculated volumes (thanks to Bland-Altman plot) were performed to evaluate the results. The nnUNet trained in 3D mode achieved the best performance, with DSC scores for the liver, the kidneys, the spinal cord, the stomach, and the duodenum of 0.96 ± 0.01, 0.91 ± 0.02, 0.91 ± 0.01, 0.83 ± 0.10, and 0.69 ± 0.15, respectively. The matching IoU scores were 0.92 ± 0.01, 0.84 ± 0.04, 0.84 ± 0.02, 0.54 ± 0.16 and 0.72 ± 0.13. The corresponding HD scores were 13.0 ± 6.0mm, 16.0 ± 6.6mm, 3.3 ± 0.7mm, 35.0 ± 33.0mm, and 42.0 ± 24.0mm. The analysis of the calculated volumes followed the same behavior. Although the segmentation results for the duodenum were not optimal, these findings imply a potential clinical application of the 3D nnUNet model for the segmentation of abdominal OARs for images from 0.35 T MR-Linac.

Drone Multiline Light Detection and Ranging Data Filtering in Coastal Salt Marshes Using Extreme Gradient Boosting Model

Quantitatively characterizing coastal salt-marsh terrains and the corresponding spatiotemporal changes are crucial for formulating comprehensive management plans and clarifying the dynamic carbon evolution. Multiline light detection and ranging (LiDAR) exhibits great capability for terrain measuring for salt marshes with strong penetration performance and a new scanning mode. The prerequisite to obtaining the high-precision terrain requires accurate filtering of the salt-marsh vegetation points from the ground/mudflat ones in the multiline LiDAR data. In this study, a new alternative salt-marsh vegetation point-cloud filtering method is proposed for drone multiline LiDAR based on the extreme gradient boosting (i.e., XGBoost) model. According to the basic principle that vegetation and the ground exhibit different geometric and radiometric characteristics, the XGBoost is constructed to model the relationships of point categories with a series of selected basic geometric and radiometric metrics (i.e., distance, scan angle, elevation, normal vectors, and intensity), where absent instantaneous scan geometry (i.e., distance and scan angle) for each point is accurately estimated according to the scanning principles and point-cloud spatial distribution characteristics of drone multiline LiDAR. Based on the constructed model, the combination of the selected features can accurately and intelligently predict the category of each point. The proposed method is tested in a coastal salt marsh in Shanghai, China by a drone 16-line LiDAR system. The results demonstrate that the averaged AUC and G-mean values of the proposed method are 0.9111 and 0.9063, respectively. The proposed method exhibits enhanced applicability and versatility and outperforms the traditional and other machine-learning methods in different areas with varying topography and vegetation-growth status, which shows promising potential for point-cloud filtering and classification, particularly in extreme environments where the terrains, land covers, and point-cloud distributions are highly complicated.

Deep Efficient Continuous Manifold Learning for Time Series Modeling.

Modeling non-euclidean data is drawing extensive attention along with the unprecedented successes of deep neural networks in diverse fields. Particularly, a symmetric positive definite matrix is being actively studied in computer vision, signal processing, and medical image analysis, due to its ability to learn beneficial statistical representations. However, owing to its rigid constraints, it remains challenging to optimization problems and inefficient computational costs, especially, when incorporating it with a deep learning framework. In this paper, we propose a framework to exploit a diffeomorphism mapping between Riemannian manifolds and a Cholesky space, by which it becomes feasible not only to efficiently solve optimization problems but also to greatly reduce computation costs. Further, for dynamic modeling of time-series data, we devise a continuous manifold learning method by systematically integrating a manifold ordinary differential equation and a gated recurrent neural network. It is worth noting that due to the nice parameterization of matrices in a Cholesky space, training our proposed network equipped with Riemannian geometric metrics is straightforward. We demonstrate through experiments over regular and irregular time-series datasets that our proposed model can be efficiently and reliably trained and outperforms existing manifold methods and state-of-the-art methods in various time-series tasks.