2D Slices Research Articles

Radiomics nomogram for prediction of glypican-3 positive hepatocellular carcinoma based on hepatobiliary phase imaging.

The hepatobiliary-specific phase can help in early detection of changes in lesion tissue density, internal structure, and microcirculatory perfusion at the microscopic level and has important clinical value in hepatocellular carcinoma (HCC). Therefore, this study aimed to construct a preoperative nomogram for predicting the positive expression of glypican-3 (GPC3) based on gadoxetic acid-enhanced (Gd-EOB-DTPA) MRI hepatobiliary phase (HBP) radiomics, imaging and clinical feature. We retrospectively included 137 patients with HCC who underwent Gd-EOB-DTPA-enhanced MRI and subsequent liver resection or puncture biopsy at our hospital from January 2017 to December 2021 as training cohort. Subsequently collected from January 2022 to June 2023 as a validation cohort of 49 patients, Radiomic features were extracted from the entire tumor region during the HBP using 3D Slicer software and screened using a t-test and least absolute shrinkage selection operator algorithm (LASSO). Then, these features were used to construct a radiomics score (Radscore) for each patient, which was combined with clinical factors and imaging features of the HBP to construct a logistic regression model and subsequent nomogram model. The clinicoradiologic, radiomics and nomogram models performance was assessed by the area under the curve (AUC), calibration, and decision curve analysis (DCA). In the validation cohort,the nomogram performance was assessed by the area under the curve (AUC). In the training cohort, a total of 1688 radiomics features were extracted from each patient. Next, radiomics with ICCs<0.75 were excluded, 1587 features were judged as stable using intra- and inter-class correlation coefficients (ICCs), 26 features were subsequently screened using the t-test, and 11 radiomics features were finally screened using LASSO. The nomogram combining Radscore, age, serum alpha-fetoprotein (AFP) >400ng/mL, and non-smooth tumor margin (AUC=0.888, sensitivity 77.7%, specificity 91.2%) was superior to the radiomics (AUC=0.822, sensitivity 81.6%, specificity 70.6%) and clinicoradiologic (AUC=0.746, sensitivity 76.7%, specificity 64.7%) models, with good consistency in calibration curves. DCA also showed that the nomogram had the highest net clinical benefit for predicting GPC3 expression.In the validation cohort, the ROC curve results showed predicted GPC3-positive expression nomogram model AUC, sensitivity, and specificity of 0.800, 58.5%, and 100.0%, respectively. HBP radiomics features are closely associated with GPC3-positive expression, and combined clinicoradiologic factors and radiomics features nomogram may provide an effective way to non-invasively and individually screen patients with GPC3-positive HCC.

Adjuvant Chemoradiotherapy within One Year of Resection for Molecularly Defined Astrocytoma

Treatments for diffuse low-grade gliomas (LGG) are controversial. Level I evidence supports the use of adjuvant radiotherapy (RT) and PCV chemotherapy for histologic LGG, but integration of molecular biomarkers in recent WHO classification and the emergence of temozolomide chemotherapy for gliomas necessitates additional investigation of the optimal treatment and timing of postoperative interventions. We hypothesized molecularly-defined LGG (IDH-mutant astrocytoma (astro) and IDH-mutant, 1p/19q-codeleted oligodendroglioma (oligo)) may have different clinical outcomes following adjuvant RT (aRT) with chemotherapy (aRT+chemo) vs observation or chemo alone. A retrospective analysis of consecutive adult patients diagnosed with WHO Grade 2 astrocytoma or oligodendroglioma who underwent initial resection at a single institution from January 1998 to November 2017 was performed. Wilcoxon rank sum and Chi-squared tests were used to compare continuous and categorical variables, respectively. Survival analyses were performed using the Kaplan-Meier method and Cox proportional hazards models. Patients without clinical progression or death were censored at the date of last follow-up. Pre-operative and post-operative T2 FLAIR hyperintense tumor volumes were quantified using 3D Slicer to calculate extent of resection (EOR). A total of 342 patients with molecularly-defined LGG (178 astro, 164 oligo) were identified with a median follow up of 9.1 yr. 171 (50%) patients received RT during their treatment course, of which 31 (18%) were treated with aRT within 1 year of diagnosis. The median aRT dose was 54 Gy (range: 40-60 Gy). aRT was more likely for astro (58%) vs oligo (41%, p = 0.001) and for patients who had resections with lower median EOR (88% vs 95%, p = 0.014). 53 patients (15%) were treated with chemo alone, and 136 patients (40%) were treated with aRT+chemo. Temozolomide was used for 161 patients (85%). For astro, aRT+chemo was associated with longer PFS (median 14.9 yr) compared to observation (4.8 yr, p = 0.05), aRT without chemo (5.2 yr, p = 0.01), or chemo alone (4.7 yr, p = 0.02). For oligo, aRT+chemo was associated with longer PFS (median not reached) compared to aRT without chemo (1.6 yr, p = 0.03), but not when compared to observation (median not reached, p = 0.47), or chemo alone (7.9 yr, p = 0.45). Multivariate analysis showed preoperative tumor volume, EOR, and aRT+chemo (but not aRT or chemo alone) were independently associated with astro PFS compared to observation. Propensity matching based on pre-operative tumor volume, EOR, and age demonstrated longer astro PFS after aRT+chemo (14.9 yr) compared to observation or chemo alone (4.5 yr, p = 0.015), without significant difference in OS (18.2 vs. 11.5 yr, p = 0.40). Retrospective data from a single institution support the use of adjuvant radiotherapy with chemotherapy for patients with molecular astrocytomas, while the role of this approach for oligodendrogliomas is unclear in this cohort.

Multi-Specialty Physician Performance in Predicting Radiographic Extranodal Extension in Nodal Metastases of Oropharyngeal Squamous Carcinoma

The presence of extranodal extension (ENE) in oropharyngeal (OPC) cancer patients is an important prognostic factor and can be used to determine the optimal course of treatment; however, currently, the gold standard method for ENE assessment is performed pathologically, which can only be done in a post-hoc fashion after surgical treatment has already been performed. Anatomic imaging features are being explored as a possible method for the pre-therapeutic determination of ENE, but there is currently no objective standard for the assessment of ENE from radiographic images. In this study we recruited expert clinicians, including surgeons, radiation oncologists, and radiologists, across multiple institutions to individually evaluate the presence of ENE from CT scans in order to assess the performance of radiographic ENE evaluation in human experts across different specialties. Pre-therapy contrast-enhanced CT scans were collected from 25 OPC patients with lymph node metastasis that were pathologically evaluated for ENE after surgical resection. 5 scans were randomly chosen to be duplicated and left/right inverted, resulting in a total of 30 scans of which 21 had pathologically-confirmed ENE. To hide the inversion, all images were cropped to only show the oropharynx region. 34 expert head and neck cancer physicians, comprised of 12 surgeons, 11 radiation oncologists, and 11 radiologists, then separately evaluated the 30 CT scans using 3D Slicer for ENE presence or absence with their prediction confidence. For each physician, discriminative performance metrics were measured by calculating the accuracy, sensitivity, specificity, area under the receiver-operating characteristic curve (AUC), and Brier score, a measure of the probabilistic prediction accuracy calculated from their confidence where a lower Brier score is better. Statistical tests were performed using the Mann Whitney U test. The median (interquartile) study results are shown in Table 1. There was no statistically significant difference among groups for accuracy or AUC, but significant differences among groups for Brier score, sensitivity, and specificity. In this study we provide evidence that expert physicians, regardless of specialty, show poor performance in assessing the presence of ENE from CT scans in OPC patients. These results agree with conclusions from previous literature, and suggest the need for further research in the automated analysis of radiographic ENE.

Investigation of Ensemble Learning Method in Predicting Early Xerostomia Using Radiomics Features from CT and MRI

There are different side effects in radiotherapy of head and neck cancer including xerostomia. Xerostomia is due to irradiation of salivary glands during treatment. In the present study, the main aim is evaluation and validation of individual and ensemble classifiers for prediction of early xerostomia in radiotherapy of head and neck cancer. Totally 62 patients diagnosed with head and neck cancer were evaluated prospectively. For this purpose, patients' demographic information, dosimetric data, and the extracted features from CT and T1 and T2 MR images before radiotherapy treatments were used as input for the predictive model. There were two main steps in this study: 1- Comparison of the features from T1 and T2 MR images for prediction of xerostomia, and 2- Evaluation of the performance of the radiomics models using individual classifiers and combination of the classifiers using ensemble learning. Feature extraction was performed using 3D Slicer software and for both parotids, 642 features were extracted from both CT, and MRI and Pearson statistical test was used for feature selection. For model development, four models were used consist of Random Tree (RT), Neural Network (NN), Linear Support Vector Machine (LSVM) and Bayesian Network (BS) classifiers. Different algorithms combined by using the ensemble learning method. For evaluation of the performance of the constructed models, sensitivity, specificity, area under curve (AUC), and receiver operating characteristic (ROC) curve were utilized. The results of the present study show that the extracted features from T1 weighted images have superior prediction ability compared to the T2 weighted ones. The RT and LSVM models which were based on T1 weighted images have better performance than those with T2 weighted images. The AUC in the test group for these models are 0.95 and 0.76, respectively, while for the T2 weighted images the corresponding values are 0.75 and 0.63. The AUC for the test group for the RT-BN, RT-LSVM-BN and RT-NN-LSVM-BN models amount to 0.96, 0.91, and 0.88, respectively. The results show that the radiomics features from images before radiotherapy can be used as personalized and unique biomarkers for prediction of xerostomia. Ensemble classifiers can be more efficient than individual classifiers in prediction of early xerostomia.

Overview of modern digital diagnostic image markup tools

Background. In modern medicine, artificial intelligence algorithms are being actively introduced, for testing and training of which a large amount of labeled datasets is required. Software for labeling (annotation) of digital diagnostic images is a necessary element when creating datasets.&#x0D; Aim. To review the capabilities and comparative analysis of the functionality of the most common available software for annotating digital diagnostic images.&#x0D; Material and methods. Five free and one commercial software product for annotation of digital diagnostic images participated in the comparative analysis. When testing the marking process on medical images for several target types of pathology, the usability of the graphical user interface and functionality was evaluated. The functionality of the software products has been tested by radiologists with over 5 years of experience. In addition, a review of semi-automatic segmentation methods implemented in the studied software products was carried out. As initial medical images, datasets of computed tomography studies obtained from open sources, were used.&#x0D; Results. Comparison of software functionality for annotation of digital diagnostic images was made: supported formats; loading, presenting and saving original images and annotation data; the possibility of visualization of medical images; annotation tools. The algorithms underlying semi-automatic segmentation methods were studied and systematized. The requirements for the basic functionality of software for labeling digital diagnostic images have been formulated. The results obtained create a systematic basis for developing recommendations for radiologists on the choice and use of digital diagnostic image marking tools.&#x0D; Conclusion. The most complete functionality in the field of segmentation of digital diagnostic images among the considered free software has 3D Slicer; in the case of annotation for detection tasks, it is convenient to use the Supervisely, CVAT platforms; for automatic segmentation of some types of pathology and organs, 3D Slicer extensions and ready-made models in Medseg can be used.

Application of Glasses-Free Augmented Reality Localization in Neurosurgery

The accurate localization of intracranial lesions is critical in neurosurgery. Most surgeons locate the vast majority of neurosurgical sites through skull surface markers, combined with neuroimaging examination and marking lines. This project's primary purpose was to develop an augmented reality (AR) technology or tool that can be used for surgical positioning using the naked eye. Brain models were predesigned with intracranial lesions using computerized tomography scan, and Digital Imaging and Communications in Medicine data were segmented and modeled by 3D slicer software. The processed data were imported into a smartphone 3D viewing software application (Persp 3D) and were used by a Remebot surgical robot. The localization of intracranial lesions was performed, and the AR localization error was calculated compared with standard robot localization. After mastering the AR localization registration method, surgeons achieved an average localization error of 1.39±0.82mm. The error of AR positioning technology in surgical simulation tests based on brain modeling was millimeter level, which has verified the feasibility of clinical application. More efficient registration remains a need that should be addressed.

Three-dimensional evaluation of the association between tongue position and upper airway morphology in adults: A cross-sectional study.

This study aimed to evaluate the association between low tongue position (LTP) and the volume and dimensions of the nasopharyngeal, retropalatal, retroglossal, and hypopharyngeal segments of the upper airway. A total of 194 subjects, including 91 males and 103 females were divided into a resting tongue position (RTP) group and a LTP group according to their tongue position. Subjects in the LTP group were divided into four subgroups (Q1, Q2, Q3, and Q4) according to the intraoral space volume. The 3D slicer software was used to measure the volume and minimum and average cross-sectional areas of each group. Airway differences between the RTP and LTP groups were analyzed to explore the association between tongue position and the upper airway. No significant differences were found in the airway dimensions between the RTP and LTP groups. For both retropalatal and retroglossal segments, the volume and average cross-sectional area were significantly greater in the patients with extremely low tongue position. Regression analysis showed that the retroglossal airway dimensions were positively correlated with the intraoral space volume and negatively correlated with A point-nasion-B point and palatal plane to mandibular plane. Males generally had larger retroglossal and hypopharyngeal airways than females. Tongue position did not significantly influence upper airway volume or dimensions, except in the extremely LTP subgroup.

CAPNet: Context attention pyramid network for computer-aided detection of microcalcification clusters in digital breast tomosynthesis

Background and objectiveComputer-aided detection (CADe) of microcalcification clusters (MCs) in digital breast tomosynthesis (DBT) is crucial in the early diagnosis of breast cancer. Although convolutional neural network (CNN)-based detection models have achieved excellent performance in medical lesion detection, they are subject to some limitations in MC detection: 1) Most existing models employ the feature pyramid network (FPN) for multi-scale object detection; however, the rough feature sharing between adjacent layers in the FPN may limit the detection ability for small and low-contrast MCs; and 2) the MCs region only accounts for a small part of the annotation box, so the features extracted indiscriminately within the whole box may easily be affected by the background. In this paper, we develop a novel CNN-based CADe method to alleviate the impacts of the above limitations for the accurate and rapid detection of MCs in DBT. MethodsThe proposed method has two parts: a novel context attention pyramid network (CAPNet) for intra-layer MC detection in two-dimensional (2D) slices and a three-dimensional (3D) aggregation procedure for aggregating 2D intra-layer MCs into a 3D result according to their connectivity in 3D space. The proposed CAPNet is based on an anchor-free and one-stage detection architecture and contains a context feature selection fusion (CFSF) module and a microcalcification response (MCR) branch. The CFSF module can efficiently enrich shallow layers’ features by the complementary selection of local context features, aiming to reduce the missed detection of small and low-contrast MCs. The MCR branch is a one-layer branch parallel to the classification branch, which can alleviate the influence of the background region within the annotation box on feature extraction and enhance the ability of the model to distinguish MCs from normal breast tissue. ResultsWe performed a comparison experiment on an in-house clinical dataset with 648 DBT volumes, and the proposed method achieved impressive performance with a sensitivity of 91.56 % at 1 false positive per DBT volume (FPs/volume) and 93.51 % at 2 FPs/volume, outperforming other representative detection models. ConclusionsThe experimental results indicate that the proposed method is effective in the detection of MCs in DBT. This method can provide objective, accurate, and quick diagnostic suggestions for radiologists, presenting potential clinical value for early breast cancer screening.

Automatic segmentation of brain glioma based on XY-Net.

Glioma is a malignant primary brain tumor, which can easily lead to death if it is not detected in time. Magnetic resonance imaging is the most commonly used technique to diagnose gliomas, and precise outlining of tumor areas from magnetic resonance images (MRIs) is an important aid to physicians in understanding the patient's condition and formulating treatment plans. However, relying on radiologists to manually depict tumors is a tedious and laborious task, so it is clinically important to investigate an automated method for outlining glioma regions in MRIs. To liberate radiologists from the heavy task of outlining tumors, we propose a fully convolutional network, XY-Net, based on the most popular U-Net symmetric encoder-decoder structure to perform automatic segmentation of gliomas. We construct two symmetric sub-encoders for XY-Net and build interconnected X-shaped feature map transmission paths between the sub-encoders, while maintaining the feature map concatenation between each sub-encoder and the decoder. Moreover, a loss function composed of the balanced cross-entropy loss function and the dice loss function is used in the training task of XY-Net to solve the class unevenness problem of the medical image segmentation task. The experimental results show that the proposed XY-Net has a 2.16% improvement in dice coefficient (DC) compared to the network model with a single encoder structure, and compare with some state-of-the-art image segmentation methods, XY-Net achieves the best performance. The DC, HD, recall, and precision of our method on the test set are 74.49%, 10.89mm, 78.06%, and 76.30%, respectively. The combination of sub-encoders and cross-transmission paths enables the model to perform better; based on this combination, the XY-Net achieves an end-to-end automatic segmentation of gliomas on 2D slices of MRIs, which can play a certain auxiliary role for doctors in grasping the state of illness.

Clinical correlation research of 3D reconstruction of retrolabyrinthine space based on HRCT of temporal bone

Background: The retrolabyrinthine approach helps clinicians perform complex surgeries such as vestibular neurectomy, resection of petrous apex cholesteatoma, or use this space to complete endoscopic combined with microscope surgical operations in a relatively safe buffer space. Some of our current studies using 3D reconstruction in the clinic have also helped us perform some complex surgical procedures.Objective: This study aims to reveal the relationship between important structures in retrolabyrinthine space through objective parameters. These measurement data help clinicians locate intraoperatively and provide a reference for clinical surgery. Also, we are intended to help improve surgical techniques and expand the operating space to increase reachable anatomic structure.Material and Methods: The inner structures of the temporal bone from HRCT (High-resolution computed tomography) images which were taken at the Eye & ENT Hospital of Fudan University were reconstructed. Precise measurement of the structures was accomplished by using the software 3D-Slicer (3D Slicer, https://www.slicer.org/; version 4.8.0, Massachusetts, USA).Results: 3D model of temporal bone structures, including the cochlea, semicircular canals (SCCs), the internal auditory canal (IAC), facial nerve (FN), jugular bulb(JB), and carotid artery was reconstructed. The combination of HRCT and 3D models is utilized to analyze the Quantitative data of the retrolabyrinthine space and its adjacent structures.Conclusions and Significance: 3D reconstruction of CT images clearly displayed the detailed structures of the temporal bone. Surgical adaptability of the retrolabyrinthine approach can be assessed preoperatively by image and other methods, and anatomical parameters play an important role in the retrolabyrinthine space. Therefore, this study helps to skeleton the bone as much as possible to expand the surgical space, so that the surgeon can contact the anatomical structure more diversified to expand the surgical indications.

A conditional GAN-based approach to build 3D facies models sequentially upwards

This paper presents an alternative way of simulating 3D facies models using conditional generative adversarial networks (GANs) by reconstructing the 3D volume upward with customised thickness from generated 2D slices. To mimic natural fluvial reservoirs, GANs need to learn the complicated facies’ lateral and vertical sequences, connectivity patterns and spatial correlation, e.g. the distribution of channel deposits and lateral accretion packages, including channel lags and point bars in a meandering fluvial system. 3D GAN modelling remains challenging, despite some success in 3D, due to the increased data and GAN model complexity caused by the 3D training data. For example, a 3D facies model with a larger number of voxels introduces more complex geobody shapes and requires 3D convolutions that increases the computational burden compared to 2D convolutions.Our proposed conditional GAN-based framework predicts 2D upper layers slice by slice to build 3D facies models, mimicking the sedimentary process in a purely aggrading system, which is less computationally costly than generating the facies models used as training data and generating facies models directly in 3D from a GAN. Each upper layer is predicted conditionally to a lower layer that contains the facies distributions in all layers below the target using several unique training enhancements. All those methods help prevent the simulated 3D models from gradually losing realism and the preset sedimentary settings vertically while inferring the plausible evolution of the channel system. The results demonstrate that the trained conditional generator, a SPADE generator, can work as a simulator to produce unconditional 3D meandering fluvial facies models by inputting a user-defined spatial correlation strength between layers and the number of layers needed (thickness in metres) without post-processing or retraining. This framework’s performance was evaluated using a connectivity metric towards the range of avulsion scenarios that represent a range of depositional uncertainty.

Application effect of head-mounted mixed reality device combined with 3D printing model in neurosurgery ventricular and hematoma puncture training

BackgroundThe purpose of this study was to explore the applicability of application effect of head-mounted mixed reality (MR) equipment combined with a three-dimensional (3D) printed model in neurosurgical ventricular and haematoma puncture training.MethodsDigital Imaging and Communications in Medicine (DICOM) format image data of two patients with common neurosurgical diseases (hydrocephalus and basal ganglia haemorrhage) were imported into 3D Slicer software for 3D reconstruction, saved, and printed using 3D printing to produce a 1:1-sized head model with real person characteristics. The required model (brain ventricle, haematoma, puncture path, etc.) was constructed and imported into the head-mounted MR device, HoloLens, and a risk-free, visual, and repeatable system was designed for the training of junior physicians. A total of 16 junior physicians who studied under this specialty from September 2020 to March 2022 were selected as the research participants, and the applicability of the equipment and model during training was evaluated with assessment score sheets and questionnaires after training.ResultsAccording to results of the assessment and questionnaire, the doctors trained by this system are more familiar with the localization of the lateral anterior ventricle horn puncture and the common endoscopic surgery for basal ganglia haemorrhage, as well as more confident in the mastery of these two operations than the traditional training methods.ConclusionsThe use of head-mounted MR equipment combined with 3D printing models can provide an ideal platform for the operation training of young doctors. Through holographic images created from the combination of virtual and real images, operators can be better immersed in the operation process and deepen their understanding of the operation and related anatomical structures. The 3D printed model can be repeatedly reproduced so that doctors can master the technology, learn from mistakes, better achieve the purpose of teaching and training, and improve the effect of training.

Quantitative Assessment of the Pharyngeal Recess Morphometry in Anatolian Population Using 3D Models Generated from Multidetector Computed Tomography Images

Aim: It was aimed to analyze the detailed morphometry of the pharyngeal recess (PR) using three-dimensional (3D) models reconstructed from multidetector computed tomography (MDCT) images.&#x0D; Material and Methods: This study was a retrospective analysis and performed on MDCT images of 97 patients (43 males, 54 females). 3D models of the PR were reconstructed using 3D Slicer software, enabling morphometric measurements according to established protocols. Measurements included PR depths, distances between the posterior nasal spine and torus levatorius (PNS-TL), distances between right and left TL (RTL-LTL) distances between the PNS and posterior wall of the nasopharynx (PNS-PWN), the angle (α) between the centerline of the PR and the sagittal plane. The morphologies of the PR classified into three types.&#x0D; Results: The average measurements for the parameters were as follows: PR depth - 10.42 mm, distance between PNS and TL - 10.40 mm, distance between RTL and LTL - 19.13 mm, distance between PNS and PWN - 19.92 mm, and the angle (α) - 53.65°. The prevalence of PR types was 20.62%, 47.42% and 31.96% for type 1, type 2 and type 3, respectively.&#x0D; Conclusion: Variations in reported measurements of the pharyngeal recess can be attributed to imaging techniques, patient positioning, anatomical differences, and sample sizes. The use of 3D models generated from MDCT datasets offers a high-resolution and comprehensive approach to understanding the PR's morphometry and spatial relationships, enabling accurate measurements and advancing our knowledge of this anatomical region.

Preoperative prediction of pancreatic neuroendocrine tumor grade based on 68Ga-DOTATATE PET/CT

ObjectiveTo establish a prediction model for preoperatively predicting grade 1 and grade 2/3 tumors in patients with pancreatic neuroendocrine tumors (PNETs) based on 68Ga-DOTATATE PET/CT.MethodsClinical data of 41 patients with PNETs were included in this study. According to the pathological results, they were divided into grade 1 and grade 2/3. 68Ga-DOTATATE PET/CT images were collected within one month before surgery. The clinical risk factors and significant radiological features were filtered, and a clinical predictive model based on these clinical and radiological features was established. 3D slicer was used to extracted 107 radiomic features from the region of interest (ROI) of 68Ga-dotata PET/CT images. The Pearson correlation coefficient (PCC), recursive feature elimination (REF) based five-fold cross validation were adopted for the radiomic feature selection, and a radiomic score was computed subsequently. The comprehensive model combining the clinical risk factors and the rad-score was established as well as the nomogram. The performance of above clinical model and comprehensive model were evaluated and compared.ResultsAdjacent organ invasion, N staging, and M staging were the risk factors for PNET grading (p < 0.05). 12 optimal radiomic features (3 PET radiomic features, 9 CT radiomic features) were screen out. The clinical predictive model achieved an area under the curve (AUC) of 0.785. The comprehensive model has better predictive performance (AUC = 0.953).ConclusionWe proposed a comprehensive nomogram model based on 68Ga-DOTATATE PET/CT to predict grade 1 and grade 2/3 of PNETs and assist personalized clinical diagnosis and treatment plans for patients with PNETs.

Three-Dimensional (3D) Analysis of Orbital Morphometry in Healthy Anatolian Adults: Sex, Side Discrepancies, and Clinical Relevance.

Background and objective The complex structural integrity of the orbit is crucial for visual functions. Three-dimensional (3D) reconstructions from radiological images have revolutionized anatomical understanding, eliminating the limitations of two-dimensional (2D) imaging and offering intricate spatial details for complex structures. The aim of this study was to analyze the detailed morphometry of the orbit in healthy Anatolian adults, considering sex, side discrepancies, and clinical relevance using 3D models reconstructed from multidetector computed tomography (MDCT) images. Materials and methods Fifty-six (44.44%) males and 70 (55.56%) females (total: 126, mean age: 48.62) MDCT images were randomly selected and 3D skull models were reconstructed using 3D Slicer software. Measurements were conducted in millimeters (mm) for transverse and vertical diameters, circumference of the orbital opening, orbital wall lengths, as well as intraorbital and extraorbital distances. Results The method of measurements showed high reproducibility of results. The mean values for transverse and vertical diameters, circumference of the orbital opening, medial and lateral wall lengths, roof and floor lengths, and intraorbital and extraorbital distances were 40.23±2.12 mm (p<0.001), 34.94±2.16 mm (p=0.365), 44.74 ±3.02 mm (p<0.001), 46.30±2.69 mm (p<0.001), 51.26±2.91 mm (p<0.001), 49.01±3.22 mm (p<0.001), 126.10±5.71 mm (p<0.001), 19.63±2.35 mm (p=0.026), and 94.09±4.84 mm (p<0.001), respectively. Conclusion The study's high measurement reproducibility contributes significantly to the existing literature and clinical practice. These findings offer specific insights into Anatolian orbital morphometry, aiding in surgical planning, implant placement, and diagnostic assessments. The precise measurement values serve as a reliable reference for clinicians, facilitating the identification of normal and abnormal orbital anatomy and enhancing patient care. We believe this study provides valuable data for craniofacial and ophthalmological research, benefiting both clinical practice and future research endeavors in these fields.

The possibilities of machine learning in determining the variants of the course of viral pneumonia associated with COVID-19 based on computed tomography data

Introduction. In acute COVID-19 respiratory infection caused by SARS-CoV-2 coronavirus (2019-nCov), lung damage has a different course, which has not been studied so far. The purpose of the study. To study the variants of the course of COVID-19 viral pneumonia (VP) based on the analysis of the dynamics of lung damage, quantified by computed tomography.Material and methods. Quantitative analysis of computed tomography (CT) data of the chest of 144 patients with VP was performed using the 3D Slicer software application. Cluster and comparative nonparametric analyses of the severity of lung damage (CT1, CT2, СТЗ, CT4) and the total volume of affected lungs (%) obtained during primary and two repeated CT studies (pCT, 1dCT and 2dCT) in the program "Statistica 12" were carried out.Results. With a stable course, the total volume of the affected lungs is constant, within one degree of severity: CT1 - in 23.6 %, CT2 - in 14.6 %, CT3 - in 5.6 %, CT4 - in 8.3 %. With a progressive course, the volume of lung damage gradually increases from minimum to maximum CT1-CT2-CT3 in 3.5 %, or increases to the maximum level on the second CT, remaining stable on the third study (CT2-CT3-CT3) - in 4.2 %. With a regredient course, the volume of lung damage varies from the maximum at the primary examination (pCT) to the minimum at the third (2dCT): gradually from CT4 through CT3 to CT2 in 4.2 %, or by one degree of severity CT3-CT2-CT2 in 9.7 %, CT3-CT3-CT2 in 9.0 %. With a progressive-regredient course, the volume of lung damage first reaches a maximum on 1dCT, a minimum on 2dCT - CT2-CT3-CT2 - in 17.4 %. An intragroup comparison of three repeated, dependent indicators of the total volume of the affected lungs and an intergroup comparison on pCT, 1dCT and 2dCT showed a dynamic statistically significant difference between them for variants of the course of VP (p &lt; 0.05).Conclusions. Cluster analysis of the total volume of affected lungs on a series of three CT studies in dynamics allowed us to identify 5 variants of the course of COVID-19 - stable - light, stable-severe, progressive, regredient, progressive-regredient.

Brain-derived subgroups of bipolar II depression associate with inflammation and choroid plexus morphology.

Elevated inflammation and larger choroid plexus (ChP) volume has been previously identified in mood disorders. Connections between inflammation, ChP, and clinical symptoms in bipolar II depression (BDII-D) are unclear. Data-driven clustering based on neuroanatomical phenotypes may help to elucidate neurobiological associations in BDII-D. Inflammatory cytokines, clinical symptoms, and neuroanatomical features were assessed in 150 BDII-D patients. Sixty-eight cortical surface area (SA) and 19 subcortical volumes were extracted using FreeSurfer. The ChP volume was segmented manually using 3D Slicer. Regularized canonical correlation analysis was used to identify significantly correlated components between cortical SA and subcortical volumes (excluding the ChP), followed by k-means clustering to define brain-derived subgroups of BDII-D. Low-grade inflammation was derived by averaging the standardized z scores of interleukin (IL)-6, IL-1β, and tumor necrosis factor (TNF)-α), which were computed to create a composite z-value score. Partial Pearson correlations followed by multiple comparison correction was conducted to explore associations between inflammation, clinical symptoms, and ChP volume. Subgroup I demonstrated smaller subcortical volume and cortical SA, higher inflammation, and larger ChP volume compared with subgroup II. Greater ChP volume was associated with a higher low-grade inflammation (mean r = 0.289, q = 0.003), CRP (mean r = 0.249, q = 0.007), IL-6 (left r = 0.200, q = 0.03), and TNF-α (right r = 0.226, q = 0.01), while greater IL-1β was significantly associated with severe depressive symptoms in BDII-D (r = 0.218, q = 0.045). Neuroanatomically-derived subgroups of BDII-D differed in their inflammation levels and ChP volume. These findings suggest an important role of elevated peripheral inflammation and larger ChP in BDII-D. This article is protected by copyright. All rights reserved.

An efficient stacked bidirectionalGRU‐LSTMnetwork for intracranial hemorrhage detection

AbstractIntracranial hemorrhage (ICH) is a dangerous condition that needs prompt diagnosis and treatment. Computed tomography (CT) images are employed in examination of individuals with ICH, which produces better results and cost‐effective than MRI. The existing convolutional neural network (CNN) models are unable to consider inter‐pixel dependency, which leads to false predictions while considering the input CT Images. In this study, we implemented an efficient model of a stack of bidirectional gated recurrent unit (Bi‐GRU) with a bidirectional long short‐term memory (Bi‐LSTM) based CNN to improve detection accuracy in the case of 2D slices. The proposed model holds slice‐wise information by accessing the properties of both Bi‐LSTM and Bi‐GRU modules in a single unit. As a result, the model attained a testing and training accuracy of 96.2% and 93.4%, respectively, with a test loss score of 0.126. In addition, the proposed model could outperform the state‐of‐the‐art CNN in identifying brain hemorrhages.

Volumetric and functional outcomes at 1-year between percutaneous-ablation and partial-nephrectomy for T1b renal tumors.

Indication for percutaneous-ablation (PA) is gradually expanding to renal tumors T1b (4-7cm). Few data exist on the alteration of renal functional volume (RFV) post-PA. Yet, it is a surrogate marker of post partial-nephrectomy (PN) glomerular filtration rate (GFR) impairment. The objective was to compare RFV and GFR at 1-year post-PN or PA, in this T1b population. Patients with unifocal renal tumor≥4cm treated between 2014 and 2019 were included. Tumor, homolateral (RFVh), contralateral RFV, and total volumes were assessed by manual segmentation (3D Slicer) before and at 1 year of treatment, as was GFR. The loss of RFV, contralateral hypertrophy, and preservation of GFR were compared between both groups (PN vs. PA). 144 patients were included (87PN, 57PA). Preoperatively, PA group was older (74 vs. 59years; P<0.0001), had more impaired GFR (73 vs. 85mL/min; P=0.0026) and smaller tumor volume(31.1 vs. 55.9cm3; P=0.0007) compared to PN group. At 1year, the PN group had significantly more homolateral RFV loss (-19 vs. -14%; P=0.002), and contralateral compensatory hypertrophy (+4% vs. +1,8%; P=0.02, respectively). Total-RFV loss was similar between both (-21.7 vs. -19cm3; P=0.07). GFR preservation was better in the PN group (95.9 vs. 90.7%; P=0.03). In multivariate analysis, age and tumor size were associated with loss of RFVh. For renal tumors T1b, PN is associated with superior compensatory hypertrophy compared with PA, compensating for the higher RFVh loss, resulting in similar ΔRFV-total between both groups. The superior post-PN GFR preservation suggests that the preserved quantitative RFV factor is insufficient. Therefore, the underlying quality of the parenchyma would play a major role in postoperative GFR.

Standardized measurement of mid-surface shift of brain based on deep Hough transform.

The measurement of mid-surface shift (MSS), the geometric displacement between the actual mid-surface and the ideal midsagittal plane (iMSP), is of great significance for accurate diagnosis, treatment and prognosis of patients with intracranial hemorrhage (ICH). Most previous studies are subject to inherent inaccuracy on account of calculating midline shift (MLS) based on 2D slices and ignoring pathological conditions. In this study, we propose a novel standardized measurement model to quantify the distance and the overall volume of mid-surface shift (MSS-D, MSS-V). Our work has four highlights. First, we develop an end-to-end network architecture with multiple sub-tasks including the actual mid-surface segmentation, hematoma segmentation and iMSP detection, which significantly improves the efficiency and accuracy of MSS measurement by taking advantage of the common properties among tasks. Second, an efficient iMSP detection scheme is proposed based on the differentiable deep Hough transform (DHT), which converts and simplifies the plane detection problem in the image space into a keypoint detection problem in the Hough space. Third, we devise a sparse DHT strategy and a weighted least square (WLS) method to increase the sparsity of features, improving inference speed and greatly reducing computation cost. Fourth, we design a joint loss function to comprehensively consider the correlation of features between multi-tasks and multi-domains. Extensive validation on our large in-house dataset (519 patients) and the public CQ500 dataset (491 patients), demonstrates the superiority of our method over the state-of-the-art methods.