Contour Variations Research Articles

FBDPN: CNN-Transformer hybrid feature boosting and differential pyramid network for underwater object detection

Despite advancements in underwater object detection (UOD) from optical underwater images in recent years, the task still poses significant challenges due to the chaotic underwater environment, as well as the substantial variations in scale and contour of objects. Existing deep learning-based schemes generally overlook the enhancement and refinement between multi-scale features of densely distributed underwater objects, leading to inaccurate localization and classification predictions with excessive information redundancy. To tackle the above issues, this article presents a novel feature boosting and differential pyramid network (FBDPN) for precise and efficient UOD. The salient properties of our paper are: (1) a heuristic feature pyramid network (FPN)-inspired architecture is constructed, which employs a convolutional neural network (CNN)-Transformer hybrid strategy to simultaneously facilitate the learning of multi-scale features and the capture of long-distance dependencies among pixels. (2) A neighborhood-scale feature boosting module (NSFBM) is developed to enhance contextual information between features of neighborhood scales. (3) A cross-scale feature differential module (CSFDM) is designed further to achieve effective information redundancy between features of different scales. Extensive experiments are conducted to reveal that our proposed FBDPN can outperform other state-of-the-art methods in both UOD performance and computational complexity. In addition, sufficient ablation studies are also performed to demonstrate the effectiveness of each component in our FBDPN. The source code is available at https://github.com/jixun-dmu/FBDPN.

Adaptive contour-tracking to aid wide-field swept-source optical coherence tomography imaging of large objects with uneven surface topology.

High-speed and wide-field optical coherence tomography (OCT) imaging is increasingly essential for clinical applications yet faces challenges due to its inherent sensitivity roll-off and limited depth of focus, particularly when imaging samples with significant variations in surface contour. Here, we propose one innovative solution of adaptive contour tracking and scanning methods to address these challenges. The strategy integrates an electrically tunable lens and adjustable optical delay line control with real-time surface contour information, enabling dynamic optimization of imaging protocols. It rapidly pre-scans the sample surface to acquire a comprehensive contour map. Using this map, it generates a tailored scanning protocol by partitioning the entire system ranging distance into depth-resolved segments determined by the optical Raleigh length of the objective lens, ensuring optimal imaging at each segment. Employing short-range imaging mode along the sample contour minimizes data storage and post-processing requirements, while adaptive adjustment of focal length and reference optical delay line maintains high imaging quality throughout. Experimental demonstrations show the effectiveness of the adaptive contour tracking OCT in maintaining high contrast and signal-to-noise ratio across the entire field of view, even in samples with significantly uneven surface curvatures. Notably, this approach achieves these results with reduced data volume compared to traditional OCT methods. This advancement holds promise for enhancing OCT imaging in clinical settings, particularly in applications requiring rapid, wide-field imaging of tissue structures and blood flow.

Comparing immobilisation devices in gynaecological external beam radiotherapy: improving inter-fraction reproducibility of pelvic tilt.

The aim was to determine which immobilisation device improved inter-fraction reproducibly of pelvic tilt and required the least pre-treatment setup and planning interventions. Sixteen patients were retrospectively reviewed, eight immobilised using the BodyFIX system (BodyFIX®, Elekta, Stockholm, Sweden) and eight using the Butterfly Board (BB) (Bionix Radiation Therapy, Toledo, OH, USA). The daily pre-treatment images were reviewed to assess setup variations between each patient and groups for pelvic tilt, pubic symphysis, sacral promontory and the fifth lumbar spine (L5). Compared with the planning CT, pelvic tilt for most patients was within ±2° using the BodyFIX and ± 4° for the BB. The Butterfly Board had a slightly higher variance both for patient-to-patient (standard deviation of the systematic error) and day-to-day error (standard deviation of the random error). Variance in position between individual patients and the two stabilisation devices were minimal in the anterior-posterior (AP) and superior-inferior (SI) direction for the pubic symphysis, sacral promontory and L5 spine. Re-imaged fractions due to pelvic tilt reduced by about half when BodyFIX was used (39.1% BB, 19.4% BodyFIX). One patient treated with the BB required a re-scan for pelvic tilt. Three patients required a re-scan for body contour variations (two using BodyFIX and one with the BB). BodyFIX resulted in a more accurate inter-fraction setup and efficient treatment and is used as the standard stabilisation for gynaecological patients at our centre. It reduced the pelvic tilt variance and reduced the need for re-imaging pre-treatment by half.

Exploring the impact of tube rotation on the melting performance of multi-tube latent heat storage systems: A numerical investigation

This study explores the impact of tube rotation on the melting performance of a multi-tube latent heat energy storage system through numerical analysis. The system involves a phase change material (PCM) contained within a tube, which absorbs heat from hot water flowing through a tube surrounding the PCM, as well as two tubes enveloped by the PCM. The investigation considers the angle between the line connecting the centers of the two inner tubes and the horizon (α = 0°, 45°, 90°, and 135°), along with the direction of rotation of inner tubes. Results are compared with data from scenarios involving only outer tube rotation and stationary systems. Temporal variations in temperature and liquid fraction contours, along with the temporal evolution of average temperature and average liquid fraction of PCM are presented. It was determined that for each α, the melting performance varied from best to worst across the system configurations as follows: the system with all three rotating tubes, the system with only the outer rotating tube, and the stationary system. The minimum time required to complete PCM melting for cases at α = 0°, 45°, 90°, and 135° is respectively 44.92 %, 42.11 %, 41.32 %, and 40.62 % less than that of the stationary system.

Evaluating the relationship between contouring variability and modelled treatment outcome for prostate bed radiotherapy.

Objectives.Contouring similarity metrics are often used in studies of inter-observer variation and automatic segmentation but do not provide an assessment of clinical impact. This study focused on post-prostatectomy radiotherapy and aimed to (1) identify if there is a relationship between variations in commonly used contouring similarity metrics and resulting dosimetry and (2) identify the variation in clinical target volume (CTV) contouring that significantly impacts dosimetry.Approach.The study retrospectively analysed CT scans of 10 patients from the TROG 08.03 RAVES trial. The CTV, rectum, and bladder were contoured independently by three experienced observers. Using these contours reference simultaneous truth and performance level estimation (STAPLE) volumes were established. Additional CTVs were generated using an atlas algorithm based on a single benchmark case with 42 manual contours. Volumetric-modulated arc therapy (VMAT) treatment plans were generated for the observer, atlas, and reference volumes. The dosimetry was evaluated using radiobiological metrics. Correlations between contouring similarity and dosimetry metrics were calculated using Spearman coefficient (Γ). To access impact of variations in planning target volume (PTV) margin, the STAPLE PTV was uniformly contracted and expanded, with plans created for each PTV volume. STAPLE dose-volume histograms (DVHs) were exported for plans generated based on the contracted/expanded volumes, and dose-volume metrics assessed.Mainresults. The study found no strong correlations between the considered similarity metrics and modelled outcomes. Moderate correlations (0.5 <Γ< 0.7) were observed for Dice similarity coefficient, Jaccard, and mean distance to agreement metrics and rectum toxicities. The observations of this study indicate a tendency for variations in CTV contraction/expansion below 5 mm to result in minor dosimetric impacts.Significance. Contouring similarity metrics must be used with caution when interpreting them as indicators of treatment plan variation. For post-prostatectomy VMAT patients, this work showed variations in contours with an expansion/contraction of less than 5 mm did not lead to notable dosimetric differences, this should be explored in a larger dataset to assess generalisability.

Deep Learning-Derived Myocardial Strain

BackgroundEchocardiographic strain measurements require extensive operator experience and have significant intervendor variability. Creating an automated, open-source, vendor-agnostic method to retrospectively measure global longitudinal strain (GLS) from standard echocardiography B-mode images would greatly improve post hoc research applications and may streamline patient analyses. ObjectivesThis study was seeking to develop an automated deep learning strain (DLS) analysis pipeline and validate its performance across multiple applications and populations. MethodsInterobserver/-vendor variation of traditional GLS, and simulated effects of variation in contour on speckle-tracking measurements were assessed. The DLS pipeline was designed to take semantic segmentation results from EchoNet-Dynamic and derive longitudinal strain by calculating change in the length of the left ventricular endocardial contour. DLS was evaluated for agreement with GLS on a large external dataset and applied across a range of conditions that result in cardiac hypertrophy. ResultsIn patients scanned by 2 sonographers using 2 vendors, GLS had an intraclass correlation of 0.29 (95% CI: −0.01 to 0.53, P = 0.03) between vendor measurements and 0.63 (95% CI: 0.48-0.74, P < 0.001) between sonographers. With minor changes in initial input contour, step-wise pixel shifts resulted in a mean absolute error of 3.48% and proportional strain difference of 13.52% by a 6-pixel shift. In external validation, DLS maintained moderate agreement with 2-dimensional GLS (intraclass correlation coefficient [ICC]: 0.56, P = 0.002) with a bias of −3.31% (limits of agreement: −11.65% to 5.02%). The DLS method showed differences (P < 0.0001) between populations with cardiac hypertrophy and had moderate agreement in a patient population of advanced cardiac amyloidosis: ICC was 0.64 (95% CI: 0.53-0.72), P < 0.001, with a bias of 0.57%, limits of agreement of −4.87% to 6.01% vs 2-dimensional GLS. ConclusionsThe open-source DLS provides lower variation than human measurements and similar quantitative results. The method is rapid, consistent, vendor-agnostic, publicly released, and applicable across a wide range of imaging qualities.

Evaluating 3D Simulation Accuracy for Breast Augmentation Outcomes: A Volumetric and Surface Contour Analysis in Chinese Patients.

The use of three-dimensional imaging in breast augmentation with silicone implants has revolutionized the surgery planning process by providing detailed visualizations of expected post-surgical outcomes. This technology enhances the decision-making process, enabling patients to choose their implants with greater confidence and ultimately leading to higher satisfaction with the postoperative outcome. This study aims to assess the accuracy of 3D imaging simulations using the Canfield Vectra XT 3D system in predicting breast augmentation outcomes in Chinese patients, focusing on volume, surface contour, breast anterior-posterior (AP) Projection, and breast internal angle. Our study analyzed female patients who received breast augmentation, documenting their preoperative and three-month postoperative conditions with 3D Vectra XT system images. Exclusions were made for patients undergoing concurrent breast surgeries or those with tuberous or ptotic breasts, due to limitations of the imaging system. Implants used were either round textured or anatomically shaped cohesive silicone gel, inserted subpectorally through trans-axillary or inframammary incisions, based on personalized evaluations. A detailed comparison between preoperative simulations and actual postoperative outcomes was conducted, focusing on volume, surface contour, AP projection, and internal angle variations. Statistical significance was determined through paired T tests, P < 0.05. In the analysis of preoperative simulations for determining postoperative outcomes in breast surgery, our study involving 42 Chinese patients, a total of 84 breasts, was conducted. The results indicated a mean volumetric discrepancy of 21.5 ± 10.3 (SD) cubic centimeters between the simulated and actual postoperative outcomes, achieving an accuracy rate of 91.9%. The root mean square deviation for the breast surface geometry was calculated to be 4.5 ± 1.1 (SD) millimeters (mm), demonstrating a low variance between the predicted and observed outcomes. The investigation found no significant variations across any specific areas of the breast surface, highlighting the uniform accuracy of the simulations across the entire breast. Additionally, the mean differences in Anterior-Posterior (AP) projection and internal angle were determined to be 8.82 ± 5.64 mm and 0.48 ± 1.91 (SD) degrees, respectively. These findings collectively attest to the efficacy of preoperative simulations in accurately predicting the postoperative physical appearance of breasts, thereby providing a valuable tool for surgical planning and improving the consultation process for patients. The Canfield Vectra XT 3D system has proven to be remarkably accurate in predicting the volumetric outcomes of breast augmentation surgery, with an accuracy rate exceeding 91.9%. It stands as a valuable tool for surgeons and patients alike, enhancing the preoperative planning process by offering a realistic preview of surgical results. This advancement not only facilitates a deeper understanding and setting of realistic expectations for patients but also strengthens the communication between patients and surgeons, ultimately leading to higher satisfaction rates with the surgical outcomes. It also emphasizes the significance of detailed documentation and consent processes in protecting against legal repercussions. This journal requires that authors assign a level of evidence to each article. For a full description of these Evidence-Based Medicine ratings, please refer to the Table of Contents or the online Instructions to Authors www.springer.com/00266 .

Exploring and explaining variation in phrase-final f0 movements in spontaneous Papuan Malay.

This study investigates the variation in phrase-final f0 movements found in dyadic unscripted conversations in Papuan Malay, an Eastern Indonesian language. This is done by a novel combination of exploratory and confirmatory classification techniques. In particular, this study investigates the linguistic factors that potentially drive f0 contour variation in phrase-final words produced in a naturalistic interactive dialogue task. To this end, a cluster analysis, manual labelling and random forest analysis are carried out to reveal the main sources of contour variation. These are: taking conversational interaction into account; turn transition, topic continuation, information structure (givenness and contrast), and context-independent properties of words such as word class, syllable structure, voicing and intrinsic f0. Results indicate that contour variation in Papuan Malay, in particular f0 direction and target level, is best explained by turn transitions between speakers, corroborating similar findings for related languages. The applied methods provide opportunities to further lower the threshold of incorporating intonation and prosody in the early stages of language documentation.

Abstract 6224: Drop-like deformation of nuclei in cancer patient tissue

Abstract In this study, we investigated nuclear deformations in cancer tissues and their clinical relevance to diagnostics. Introduction: Cancers are graded by pathologists examining biopsy slides stained with hematoxylin and eosin for factors including nuclear atypia, which can include features like lobules, invaginations, large nuclear size, and prominent nucleoli. While the importance of nuclear atypia has been known for centuries, its cause remains unclear. Nuclear shaping has been studied extensively in vitro, leading to the liquid drop model of nuclear shaping, which states that nuclear shape changes can occur due to excess area in the form of wrinkles or folds in the lamina. When the lamina becomes fully taut, the nucleus reaches a limiting steady shape. Wrinkles allow the nuclear shape to mirror the cell shape such as in invading cancer cells. The applicability of this model to nuclei in cancer tissues from human patients is untested. Here, tissues were stained for nuclear lamina proteins to test for the presence of excess area in the lamina. Diagnostic relevance of the drop model was assessed by quantitative Fourier analysis of high-resolution images of lamin-stained nuclei. Methods: Deidentified patient tissues from breast, colon, head and neck, and skin were immunostained for lamin B1 and imaged at high resolution (60X, 1.5 NA). Nuclei were then roughly segmented by the deep learning program Cellpose, followed by segmentation with a custom approach that sensitively detects micron-scale variations in the laminar contour. An Elliptical Fourier Analysis was performed to quantify nuclear irregularity by describing nuclear shape with a series of elliptical harmonics. More irregular nuclei are described by larger Fourier coefficients at higher frequencies. Results: Folds and wrinkles in the nuclear lamina were observed in healthy tissue from all four tissue types studied, indicating the presence of excess area in the lamina. The presence of irregular nuclear shape including wrinkling/folds in the lamina was quantified with the Elliptical Fourier Coefficient (EFC) ratio. Posterior probability analysis related EFC ratio to tumor grade to assess the clinical value of the excess area found in the nuclear lamina. This revealed cancer-type dependent trends between EFC ratio and tumor grade. Conclusions: Observations of wrinkling/folds in the nuclear lamina in healthy and cancerous tissue are consistent with the liquid drop model of nuclear deformation. Quantification of nuclear lamina wrinkling with the EFC ratio, along with other nuclear shape parameters, can be a powerful approach for computer-based, explainable, clinical grading. Our findings reveal a fundamental principle of nuclear deformation in cancer tissues and suggest a new approach for improving the accuracy in cancer grading. Citation Format: Christina R. Dubell, Ting-Ching Wang, Ishita Singh, Daniel G. Rosen, Saptarshi Chakraborty, Tanmay P. Lele. Drop-like deformation of nuclei in cancer patient tissue [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 6224.

Vulnerabilities of Prostate Glandular Concretions Considered as Artefacts: A Cross-sectional Biopsy Slide Study.

There is no clear consensus regarding the contour of prostatic glandular intraluminal concretions. This study enlightens the rational approach toward deciphering the true nature of these concretions and evaluates their role in normal routine histology of the prostate gland. Fifty hematoxylin and eosin slides that were prepared from procured transrectal biopsy specimens of normal prostate glands from asymptomatic patients suspected of having a prostatic disease but later found to be normal were retrospectively observed for the staining, contour, and positioning of the aggregated masses or concretions within their prostatic lumina and were then compared with the blood prostate specific antigen (PSA) levels. Although significant associations highlighting the utility of these masses in presumed pathological states of the gland were drawn by comparing their staining parameters and contours to those of their respective PSA levels, their interluminal contour variations and vivid staining appearances did not necessarily rule out the possibility of some of them being artefacts, provided they were assessed in totality with the surrounding acini. Intensely eosinophilic concretions were found in patients with a high mean age and those with high PSA levels. Prostatic intraluminal masses that were rounded tended to indicate pathological shifts within the gland; however, the possibility of them changing to artefacts during slide preparations could not be ruled out.

Using L1 tone and intonation categories to investigate patterns in nonnative tone identification

Many languages utilise lexical tone, a feature which assigns meaning to words through variations in pitch contours. In some (but not all) cases, prior tone language experience can facilitate the learning of a nonnative tone language. Regardless, multiple sessions of tone training can improve nonnative tone perception for both tone and nontone L1 listeners. However, the influence of L1 tone and intonation categories on nonnative tone learning is underexplored. Across five sessions of tone training, listeners of nontonal English (n = 25), tonal Mandarin (n = 23), and tonal Vietnamese (n = 25) learned tones from an artificial language. While all groups improved in tone identification and tone word learning by session five (p &amp;lt; .001), Mandarin listeners outperformed the other groups overall (identification, p &amp;lt; .001; word learning, p &amp;lt; .05). Tone experience was a main effect only in word learning (p &amp;lt; .05). Tone- and language-specific identification patterns were also observed. The low level tone was easier for all listeners to perceive, while the tonal groups also showed greater accuracy for the mid level tone. Contrastingly, Mandarin listeners were hindered by their single L1 falling tone category when attempting to identify two distinct nonnative falling tones.

Investigating the Influence of Cardiac Substructure Dose on Survival of Lung Cancer Patients

Since RTOG 0617, the relationship between cardiac dose and overall survival for lung cancer patients (pts) is increasingly scrutinized. However, clinical cardiac substructure (CS) contours are typically lacking. In this study, we use deep learning (DL) to create accurate CS contours for a large lung cohort, and then investigate the influence of CS dose and delineation-induced dose uncertainty on overall survival. Retrospectively included are 2075 lung cancer pts (T stage 1-4 with median survival times of 49, 36, 19, and 16 months respectively, median [range] age 70 [36-95], 58% male), treated with radiotherapy (RT) between 2009 and 2019 in a single academic institution. Re-irradiated pts were excluded. Input for DL were 70 randomly selected pts, which had manual ground-truth CS contours created on their RT planning scan. The 70 pts were split between training, validation, and test set (50, 10, 10). The trained network was then used to create 13 CS contours for the remaining pts. The influence of delineation-induced dose uncertainty on survival was investigated, first, by dilating, and second, by eroding the DL-created contours using the mean surface distance (MSD) ± 2SD of MSD, as estimated from the test set. Significant variables from univariable cox regression (CR) were included in multivariable CR (MCR). Separate MCR's include dose to one CS or the whole heart and age, gender, and log(V95%) while stratifying for T-stage, and palliative treatment. V95% is the volume receiving 95% prescribed dose. Tumor laterality was insignificant in the univariable analysis. DL test output was deemed sufficient (median dice score > 0.80 and/or MSD < 2mm) for all CS but the ascending aorta, aortic valve, coronary sinus, and pulmonary veins. Max and mean dose to inferior vena cava were not significant. Table 1 shows significant CS dose parameters in MCR's for 1465 pts with complete data. Hazard ratios (HR) for max and mean dose parameters (per Gy [EQD2]) are only shown when significant for all three contour variations. Confidence intervals (CI) are based on the interval over the three contour variations. *, **, *** indicating p-values <0.05, <0.01, <0.001 respectively. Nearly all CS dose parameters were significantly associated with reduced overall survival in MCR even after accounting for delineation uncertainty. Additional analysis is required to find the most crucial CS by solving multicollinearity, whereafter dosimetric constraints can be investigated.

Results From a Multi-Institutional Pilot Study of iContour, an Interactive Online Platform with Real-Time Feedback to Improve Contouring Education for Radiation Oncology Residents

Numerous studies have shown that variability in contouring by radiation oncologists is common and associated with poor clinical outcomes. Contouring is taught via an apprenticeship model during residency with inconsistent results. Currently, there is no standardized contouring curriculum. We hypothesized that an interactive online educational platform for learners to practice contouring and receive real-time visual feedback would be a useful curricular tool. The iContour platform displays anonymized DICOM data and allows for input and analysis of user contours in a web-based interface. Nine cases are available from the Head/Neck (H+N), Gynecologic (Gyn), and Gastrointestinal (GI) disease sites. The system presents users with a case and asks them to contour representative slices from 2-3 target volumes or OARs. Upon submission, users are shown several forms of feedback. These include immediate visual comparison with expert contours on the same dataset (all cases), customized feedback based on overlap with prespecified "avoidance" and "inclusion" structures that highlight common mistakes (Gyn cases), and tumor control probability models to estimate the clinical impact of a user's contour variations on patient outcomes (H+N cases). Some cases include short videos outlining anatomy and contouring principles which are shown before the user contours. A pilot study was performed to evaluate technical performance and educational utility. Pre- and post-surveys with Likert-type questions (1-5 scale from strongly disagree to strongly agree) were used to assess user satisfaction and preferences regarding feedback. A total of 9 residents participated (median PGY3, range PGY2-5) from 5 institutions in 2 countries. Each participant completed 2 cases from a single disease site (n = 3 each for H+N, Gyn and GI), one with educational videos and one without. 67% of users had completed a prior clinical rotation in their disease site. Overall, residents felt the system was a useful educational tool (mean Likert score 4.67 +/- 0.47) and were interested in using it during clinical rotations (4.89 +/- 0.31). Most participants (7/9) felt iContour was more useful than existing resources for contouring education. Residents unanimously (9/9) found direct visual comparison with expert contours the most useful type of feedback, and that cases with videos before contouring were more educational. The iContour platform is a useful educational tool for radiation oncology residents. Participants felt receiving immediate visual feedback on contours was a valuable learning experience. Short instructional videos before contouring can be utilized to provide "just in time" teaching. A randomized study to formally assess the platform's impact on contouring skills is planned.

Contouring variation affects estimates of normal tissue complication probability for breast fibrosis after radiotherapy

BackgroundNormal tissue complication probability (NTCP) models can be useful to estimate the risk of fibrosis after breast-conserving surgery (BCS) and radiotherapy (RT) to the breast. However, they are subject to uncertainties. We present the impact of contouring variation on the prediction of fibrosis. Materials and methods280 breast cancer patients treated BCS-RT were included. Nine Clinical Target Volume (CTV) contours were created for each patient: i) CTV_crop (reference), cropped 5 mm from the skin and ii) CTV_skin, uncropped and including the skin, iii) segmenting the 95% isodose (Iso95%) and iv) 3 different auto-contouring atlases generating uncropped and cropped contours (Atlas_skin/Atlas_crop). To illustrate the impact of contour variation on NTCP estimates, we applied two equations predicting fibrosis grade ≥ 2 at 5 years, based on Lyman-Kutcher-Burman (LKB) and Relative Seriality (RS) models, respectively, to each contour. Differences were evaluated using repeated-measures ANOVA. For completeness, the association between observed fibrosis events and NTCP estimates was also evaluated using logistic regression. ResultsThere were minimal differences between contours when the same contouring approach was followed (cropped and uncropped). CTV_skin and Atlas_skin contours had lower NTCP estimates (−3.92%, IQR 4.00, p < 0.05) compared to CTV_crop. No significant difference was observed for Atlas_crop and Iso95% contours compared to CTV_crop. For the whole cohort, NTCP estimates varied between 5.3% and 49.5% (LKB) or 2.2% and 49.6% (RS) depending on the choice of contours. NTCP estimates for individual patients varied by up to a factor of 4. Estimates from “skin” contours showed higher agreement with observed events. ConclusionContour variations can lead to significantly different NTCP estimates for breast fibrosis, highlighting the importance of standardising breast contours before developing and/or applying NTCP models.

Sensitivity of standardised radiomics algorithms to mask generation across different software platforms

The field of radiomics continues to converge on a standardised approach to image processing and feature extraction. Conventional radiomics requires a segmentation. Certain features can be sensitive to small contour variations. The industry standard for medical image communication stores contours as coordinate points that must be converted to a binary mask before image processing can take place. This study investigates the impact that the process of converting contours to mask can have on radiomic features calculation. To this end we used a popular open dataset for radiomics standardisation and we compared the impact of masks generated by importing the dataset into 4 medical imaging software. We interfaced our previously standardised radiomics platform with these software using their published application programming interface to access image volume, masks and other data needed to calculate features. Additionally, we used super-sampling strategies to systematically evaluate the impact of contour data pre processing methods on radiomic features calculation. Finally, we evaluated the effect that using different mask generation approaches could have on patient clustering in a multi-center radiomics study. The study shows that even when working on the same dataset, mask and feature discrepancy occurs depending on the contour to mask conversion technique implemented in various medical imaging software. We show that this also affects patient clustering and potentially radiomic-based modelling in multi-centre studies where a mix of mask generation software is used. We provide recommendations to negate this issue and facilitate reproducible and reliable radiomics.

Numerical study of entropy generation in magneto-convective flow of nanofluid in porous enclosure using fractional order non-Darcian model

The present numerical work examines the effect of fractional order parameter on heat transfer and entropy generation for a thermo-magnetic convective flow of nanofluid (Cu-water) in a square porous enclosure that contains semi-circular bottom wall. The Darcy–Brinkmann–Forchheimer model is utilized to evaluate the momentum transfer in porous media, and the Caputo-time fractional derivative term is introduced in momentum as well as in the energy equation. Further, non-dimensional governing equations are simulated through the penalty finite element method, and the Caputo time derivative is approximated by L1-scheme. The study is carried out for various parameters, including Rayleigh number (Ra), Darcy number (Da), radius of the semicircle (r), fractional order (α), and Hartmann number (Ha). The comprehensive results are presented by the contour variation of isotherms, streamlines, and total entropy generation at the selected range of parameters. In addition, thermal transport and irreversibilities due to heat transfer, fluid friction, and magnetic field have been accounted through the numerical variation of mean Nusselt number (Num) and Bejan number due to heat transfer (Beht), fluid friction (Beff), and magnetic field (Bemf), respectively. The key findings of the present study reveal that during the initial evolution period, the Num value increases as α→1. Additionally, time taken to achieve the steady state condition varies and depends on fractional order α. Furthermore, in the absence of Ha, the heat transfer and entropy generation intensifies with augmentation of Ra and Da for all α, while, the increasing value of Ha shows an adverse impact on the heat transfer rate.

Deep-Learning-Based Dose Predictor for Glioblastoma-Assessing the Sensitivity and Robustness for Dose Awareness in Contouring.

External beam radiation therapy requires a sophisticated and laborious planning procedure. To improve the efficiency and quality of this procedure, machine-learning models that predict these dose distributions were introduced. The most recent dose prediction models are based on deep-learning architectures called 3D U-Nets that give good approximations of the dose in 3D almost instantly. Our purpose was to train such a 3D dose prediction model for glioblastoma VMAT treatment and test its robustness and sensitivity for the purpose of quality assurance of automatic contouring. From a cohort of 125 glioblastoma (GBM) patients, VMAT plans were created according to a clinical protocol. The initial model was trained on a cascaded 3D U-Net. A total of 60 cases were used for training, 15 for validation and 20 for testing. The prediction model was tested for sensitivity to dose changes when subject to realistic contour variations. Additionally, the model was tested for robustness by exposing it to a worst-case test set containing out-of-distribution cases. The initially trained prediction model had a dose score of 0.94 Gy and a mean DVH (dose volume histograms) score for all structures of 1.95 Gy. In terms of sensitivity, the model was able to predict the dose changes that occurred due to the contour variations with a mean error of 1.38 Gy. We obtained a 3D VMAT dose prediction model for GBM with limited data, providing good sensitivity to realistic contour variations. We tested and improved the model's robustness by targeted updates to the training set, making it a useful technique for introducing dose awareness in the contouring evaluation and quality assurance process.

Variations of MHD double-diffusive mixed convection and entropy generation in various nanofluids and hybrid nanofluids due to the deviation of the spinning of double rotating cylinders

The current numerical study aims to investigate the effects of various rotating speeds of rotating cylinders on mixed convection and entropy generation in a lid-driven trapezoidal enclosure for fixed magnetic field inclination angle. Several water-based hybrid nanofluids are used as governing fluids. In this work, the impacts of SWCNT-water, Cu-water, and Al2O3-water nanofluids individually, as well as the effects of three different types of SWCNT-Cu-Al2O3-water hybrid nanofluids are examined. The average Nusselt number, average Sherwood number, average temperature, and average entropy generation because of heat transfer, diffusion, fluid friction, and the applied magnetic field as well as total entropy generation and Bejan number are evaluated as output parameters for different values of governing parameters. The Galerkin weighted residual finite element method is used to numerically solve the governing Navier-Stokes, thermal energy, and mass conservation equations. The optimal values of each governing parameter are established in order to produce the best results for the output parameters. To inspect the thermal and flow field within the trapezoidal enclosure as well as the variations in mass concentrations, isotherm contours, streamlines and isoconcentration contours are studied. Better performance is found for higher rotating speed regardless of other parameters.

Pancreatic morphology/contour variations should be recognized and remembered.

Pancreatic contour variations can be detected incidentally on computed tomography (CT). Recognition and remembering of these variations are important in volumetric measurements and surgery as well as in preventing misdiagnosis. This study aims to evaluate the morphology/contour variations in the pancreas head-neck, body-tail, and uncinate process with multi-detector CT (MDCT) examinations (triple phase CT abdomen). Around 1662 adult age (>18 years old) patients were evaluated retrospectively, and after exclusion criteria, 945 patients were included in the study. Aplasia and hypoplasia of the uncinate process were determined, and pancreatic contour variances were categorized according to the Ross et al. and Omeri et al. classifications. Pancreatic head-neck variants were categorized into Type I-anterior, Type II-posterior, and Type III-horizontal variations. Pancreatic body-tail variants were sectioned into Type Ia-anterior protrusion, Type Ib-posterior protrusion, and Types IIa-globular, IIb-lobulated, IIc-tapered, and IId-bifid pancreatic tail. Of the 945 patients, 481 (50.9%) were female. The mean age was 43.28 ± 10.49 (min. 20-max. 68). In the evaluations made according to the uncinate process morphology variant, hypoplasia was detected in 66 (7%) patients and aplasia in 12 (1.3%) patients. Pancreatic head-neck and body-tail contour variations were observed in 596 (63.1%) patients. The most common head-neck variation was Type II in 233 (24.6%) patients, followed by type III in 96 (10.2%). There were Type Ia in 83 (8.8%) patients and Type Ib in 14 (1.5%) patients. The pancreatic tail configuration was normal in 792 (83.8%) patients; it was Type IIa in 62 (6.6%) patients and IIb in 50 (5.3%) patients. The most common variation was head and tail in 33 (3.5%) patients. Pancreatic variations detected in CT examinations for distinct reasons are not rare; these variations should be recognized and remembered.

Image guidance and interfractional anatomical variation in paediatric abdominal radiotherapy.

To identify variables predicting interfractional anatomical variations measured with cone-beam CT (CBCT) throughout abdominal paediatric radiotherapy, and to assess the potential of surface-guided radiotherapy (SGRT) to monitor these changes. Metrics of variation in gastrointestinal (GI) gas volume and separation of the body contour and abdominal wall were calculated from 21 planning CTs and 77 weekly CBCTs for 21 abdominal neuroblastoma patients (median 4 years, range: 2 - 19 years). Age, sex, feeding tubes, and general anaesthesia (GA) were explored as predictive variables for anatomical variation. Furthermore, GI gas variation was correlated with changes in body and abdominal wall separation, as well as simulated SGRT metrics of translational and rotational corrections between CT/CBCT. GI gas volumes varied 74 ± 54 ml across all scans, while body and abdominal wall separation varied 2.0 ± 0.7 mm and 4.1 ± 1.5 mm from planning, respectively. Patients < 3.5 years (p = 0.04) and treated under GA (p < 0.01) experienced greater GI gas variation; GA was the strongest predictor in multivariate analysis (p < 0.01). Absence of feeding tubes was linked to greater body contour variation (p = 0.03). GI gas variation correlated with body (R = 0.53) and abdominal wall (R = 0.63) changes. The strongest correlations with SGRT metrics were found for anterior-posterior translation (R = 0.65) and rotation of the left-right axis (R = -0.36). Young age, GA, and absence of feeding tubes were linked to stronger interfractional anatomical variation and are likely indicative of patients benefiting from adaptive/robust planning pathways. Our data suggest a role for SGRT to inform the need for CBCT at each treatment fraction in this patient group. This is the first study to suggest the potential role of SGRT for the management of internal interfractional anatomical variation in paediatric abdominal radiotherapy.