Computed Tomography Imaging Research Articles

Prediction of Malignancy and Pathological Types of Solid Lung Nodules on CT Scans Using a Volumetric SWIN Transformer.

Lung adenocarcinoma and squamous cell carcinoma are the two most common pathological lung cancer subtypes. Accurate diagnosis and pathological subtyping are crucial for lung cancer treatment. Solitary solid lung nodules with lobulation and spiculation signs are often indicative of lung cancer; however, in some cases, postoperative pathology finds benign solid lung nodules. It is critical to accurately identify solid lung nodules with lobulation and spiculation signs before surgery; however, traditional diagnostic imaging is prone to misdiagnosis, and studies on artificial intelligence-assisted diagnosis are few. Therefore, we introduce a volumetric SWIN Transformer-based method. It is a multi-scale, multi-task, and highly interpretable model for distinguishing between benign solid lung nodules with lobulation and spiculation signs, lung adenocarcinomas, and lung squamous cell carcinoma. The technique's effectiveness was improved by using 3-dimensional (3D) computed tomography (CT) images instead of conventional 2-dimensional (2D) images to combine as much information as possible. The model was trained using 352 of the 441 CT image sequences and validated using the rest. The experimental results showed that our model could accurately differentiate between benign lung nodules with lobulation and spiculation signs, lung adenocarcinoma, and squamous cell carcinoma. On the test set, our model achieves an accuracy of 0.9888, precision of 0.9892, recall of 0.9888, and an F1-score of 0.9888, along with a class activation mapping (CAM) visualization of the 3D model. Consequently, our method could be used as a preoperative tool to assist in diagnosing solitary solid lung nodules with lobulation and spiculation signs accurately and provide a theoretical basis for developing appropriate clinical diagnosis and treatment plans for the patients.

Delta-CT radiomics based model for predicting postoperative anastomotic leakage following radical resection of esophageal squamous cell carcinoma.

To predict postoperative anastomotic leakage (AL) following radical resection of esophageal squamous cell carcinoma (ESCC) based on clinical data and preoperative enhanced Computed tomography(CT) radiomics of the esophagus. We retrospectively analyzed the clinicopathological and radiological data of 213 patients with ESCC who received radical resection at Xiangyang No.1 People's Hospital from July 2011 to February 2024. 3D slicer software was used in combination with Lasso extraction and 10-fold cross-validation to extract texture parameters from contrast-enhanced CT images and generate Delta-Radscores. Several models were built using logistic regression to predict postoperative AL in ESCC. In the training set, the univariate analysis confirmed that duration of surgery, surgical method, delta radscore 1, delta radscore 2, contrast enhancement patterns, peripheral lymph node metastasis, post thoracotomy pulmonary infection(PTPI), and hot pot were risk factors for ESCC-AL (P<0.05 for both). The multivariate analysis showed that delta radscore 1, delta radscore 2, PTPI, and hot pot were independent risk factors for AL (P<0.05 for all). These results were verified by the XGboost machine learning model. The combinational model based on all of the above risk factors [AUC 0.900, OR 0.0282, 95%CI 0.841-0.943] outperformed either the clinical model[AUC 0.759, OR 0.0392, 95%0.683-0.825,P<0.05] or the imaging model[AUC 0.869, OR 0.0335, 95%0.804-0.918,P=0.1277] alone in predictive efficacy. The decision curve proved that the combinational model had a higher clinical net benefit. The nomogram generated via the combinational model simplified the predictive process. The same predictions were verified in the testing set. Delta radscore 1, delta radscore 2, PTPI, and hot pot were related to ESCC-AL. The novel nomogram created using enhanced CT radiomics informed perioperative management and improved the survival quality of ESCC patients.

Quantitative CT imaging and radiation-absorbed dose estimations of 166Ho microspheres: paving the way for clinical application

BackgroundMicrobrachytherapy enables high local tumor doses sparing surrounding tissues by intratumoral injection of radioactive holmium-166 microspheres (166Ho-MS). Magnetic resonance imaging (MRI) cannot properly detect high local Ho-MS concentrations and single-photon emission computed tomography has insufficient resolution. Computed tomography (CT) is quicker and cheaper with high resolution and previously enabled Ho quantification. We aimed to optimize Ho quantification on CT and to implement corresponding dosimetry.MethodsTwo scanners were calibrated for Ho detection using phantoms and multiple settings. Quantification was evaluated in five phantoms and seven canine patients using subtraction and thresholding including influences of the target tissue, injected amounts, acquisition parameters, and quantification volumes. Radiation-absorbed dose estimation was implemented using a three-dimensional 166Ho specific dose point kernel generated with Monte Carlo simulations.ResultsCT calibration showed a near-perfect linear relation between radiodensity (HU) and Ho concentrations for all conditions, with differences between scanners. Ho detection during calibration was higher using lower tube voltages, soft-tissue kernels, and without a scanner detection limit. The most accurate Ho recovery in phantoms was 102 ± 11% using a threshold of mean tissue HU + (2 × standard deviation) and in patients 98 ± 31% using a 100 HU threshold. Thresholding allowed better recovery with less variation and dependency on the volume of interest compared to the subtraction of a single HU reference value. Corresponding doses and histograms were successfully generated.ConclusionCT quantification and dosimetry of 166Ho should be considered for further clinical application with on-site validation using radioactive measurements and intra-operative Ho-MS and dose visualizations.Relevance statementImage-guided holmium-166 microbrachytherapy currently lacks reliable quantification and dosimetry on CT to ensure treatment safety and efficacy, while it is the only imaging modality capable of quantifying high in vivo holmium concentrations.Key PointsLocal injection of 166Ho-MS enables high local tumor doses while sparing surrounding tissue.CT enables imaging-based quantification and radiation-absorbed dose estimation of concentrated Ho in vivo, essential for treatment safety and efficacy.Two different CT scanners and multiple acquisition and reconstruction parameters showed near-perfect linearity between radiodensity and Ho concentration.The most accurate Ho recoveries on CT were 102 ± 11% in five phantoms and 98 ± 31% in seven canine patients using thresholding methods.Dose estimations and volume histograms were successfully implemented for clinical application using a dose point kernel based on Monte Carlo simulations.Graphical

Comparison of the diagnostic efficacy between imaging features and iodine density values for predicting microvascular invasion in hepatocellular carcinoma.

In recent years, studies have confirmed the predictive capability of spectral computer tomography (CT) in determining microvascular invasion (MVI) in patients with hepatocellular carcinoma (HCC). Discrepancies in the predicted MVI values between conventional CT imaging features and spectral CT parameters necessitate additional validation. In this retrospective study, 105 cases of small HCC were reviewed, and participants were divided into MVI-negative (n=53, Male:48 (90.57%); mean age, 59.40 ± 11.79 years) and MVI-positive (n=52, Male:50(96.15%); mean age, 58.74 ± 9.21 years) groups using pathological results. Imaging features and iodine density (ID) obtained from three-phase enhancement spectral CT scans were gathered from all participants. The study evaluated differences in imaging features and ID values of HCC between two groups, assessing their diagnostic accuracy in predicting MVI occurrence in HCC patients. Furthermore, the diagnostic efficacy of imaging features and ID in predicting MVI was compared. Significant differences were noted in the presence of mosaic architecture, nodule-in-nodule architecture, and corona enhancement between the groups, all with p-values < 0.001. There were also notable disparities in IDs between the two groups across the arterial phase, portal phase, and delayed phases, all with p-values < 0.001. The imaging features of nodule-in-nodule architecture, corona enhancement, and nonsmooth tumor margin demonstrate significant diagnostic efficacy, with area under the curve (AUC) of 0.761, 0.742, and 0.752, respectively. In spectral CT analysis, the arterial, portal, and delayed phase IDs exhibit remarkable diagnostic accuracy in detecting MVI, with AUCs of 0.821, 0.832, and 0.802, respectively. Furthermore, the combined models of imaging features, ID, and imaging features with ID reveal substantial predictive capabilities, with AUCs of 0.846, 0.872, and 0.904, respectively. DeLong test results indicated no statistically significant differences between imaging features and IDs. Substantial differences were noted in imaging features and ID between the MVI-negative and MVI-positive groups in this study. The ID and imaging features exhibited a robust diagnostic precision in predicting MVI. Additionally, our results suggest that both imaging features and ID showed similar predictive efficacy for MVI.

Age-dependent changes in the hyoid bone morphology in children.

This radiologic work aimed to display the alteration in the hyoid bone (HB) morphology in the pediatric population with advancing age. This pediatric examination consisted of computed tomography images of 129 subjects (49 males / 80 females) aged 1-17 years. The anterior-posterior length of HB, the lengths of right and left greater horns, the width and height of HB's body, and the distance between the midpoints of the posterior ends of the greater horns increased with advancing age (p < 0.001), but the angle of the right and left greater horns (p = 0.022) decreased. Four configurations regarding HB shape were observed: Type A (U-shaped HB) in 8.5% (11 HBs) out of 129 children, Type B (B-shaped HB) in 33.3% (43 HBs), Type C (D-shaped HB) in 45% (58 HBs), and Type D (V-shaped HB) in 13.2% (17 HBs). HB shape types correlated with the pediatric age (p < 0.001), but not gender (p = 0.153). Most of the parameters increased until the postpubescent period, but the angle of the right and left greater horns decreased after the late childhood. Our linear functions representing the growth pattern of HB in children may be useful to estimate HB size.

Comparative Analysis of Ocular Morphology in Dromedary Camels: Insights from Ultrasound and Computed Tomography Imaging Techniques

Comparative Analysis of Ocular Morphology in Dromedary Camels: Insights from Ultrasound and Computed Tomography Imaging Techniques

Key Factors for Predicting Appendicitis from Contrast-Enhanced Computed Tomography Images Obtained Through Multiplanar Reconstruction

Background: The diameter of the appendix is a key parameter in diagnosing appendicitis. The diagnostic threshold for this parameter is 6 mm, originally established through graded compression sonography of the right lower quadrant (RLQ) of the abdomen. However, without corroborative findings from computed tomography (CT), this threshold may not be a reliable indicator of appendicitis. To ensure accurate diagnosis, clinicians should perform a comprehensive, multiparameter imaging assessment of the appendix, rather than relying solely on appendix diameter. Objectives: This study aimed to identify key factors for predicting appendicitis using contrast-enhanced coronal and sagittal CT images obtained through multiplanar reconstruction. Patients and Methods: This single-center, retrospective, cross-sectional study included patients who presented to our emergency department (ED) with RLQ abdominal pain and subsequently underwent contrast-enhanced CT between July 2019 and September 2020. The primary study outcome was pathologically confirmed appendicitis. Two experienced radiologists assessed parameters such as appendix diameter, wall thickness, abnormal appendix enhancement, abnormal appendix content, appendix erection, and periappendiceal fat stranding. Multivariate logistic regression was performed to identify significant predictive factors for appendicitis. Results: The study included 173 patients (median age: 37 years; women: 86). They were divided into appendicitis (n = 102) and alternative diagnosis (n = 71) groups. Significant differences were observed between the groups in terms of appendix diameter, wall thickness, wall enhancement, luminal content, appendix erection, and periappendiceal fat stranding (P &lt; 0.001). The diagnostic sensitivity and specificity values for an appendix diameter threshold of 7.7 mm were 91% and 82%, respectively. An appendix diameter of &gt; 7.7 mm (OR: 15.3; P &lt; 0.001), abnormal appendix enhancement (OR: 12.5; P &lt; 0.001), and appendix erection (OR: 6.1; P = 0.004) emerged as significant independent predictors of appendicitis. Conclusion: An appendix diameter of 7.7 mm appears to be the optimal threshold for diagnosing appendicitis. Additionally, the detection of abnormal appendix enhancement and appendix erection on contrast-enhanced CT images holds considerable diagnostic value.

Surgeon perspectives on a virtual reality platform for preoperative planning in complex bone sarcomas

Background and objectivesTreatment of primary bone and soft tissue sarcomas typically includes complete surgical resection with or without adjunctive modalities. Despite best efforts, for the most challenging clinical scenarios such as axial or pelvic sarcoma, five-year survival rates are reported to be between 27 and 40 %. Since quality of resection is a key determinant of oncologic outcomes, it is critical to preoperatively plan the surgical approach to improve resection accuracy, ensure sufficient surgical margins, and reduce the risk of local or metastatic recurrence. The computer conversion of 2-dimensional (2D) computerized tomography (CT) and magnetic resonance imaging (MRI) to a three-dimensional (3D) virtual reality (VR) avatar image may allow improved preoperative estimation of tumor size, location, adjacent anatomy, and spatial understanding of the tumor without relying on surgeon experience, memory, and imagination. The purpose of this study is to investigate the utility of a virtual reality platform in preoperative planning and surgical approach in a retrospective cohort of pelvic bone sarcoma cases. MethodsThe histopathology database at our institution was queried for all historical cases of bone and soft tissue sarcoma with surgical resection failure, defined as positive gross or microscopic margins. Four cases of pelvic bone sarcoma were chosen for retrospective review by fellowship-trained orthopedic tumor specialists. For each case, participants first studied conventional 2D preoperative CT images and answered a questionnaire pertaining to objective case parameters. Participants then interacted with case-specific 3D models while wearing a VR headset and answered the same questionnaire. The VR ‘avatar’ was created with custom-developed software. After using both modalities, participants completed a Likert-scale survey aiming to evaluate the VR technology's subjective impact on understanding tumor environment, surgical plan confidence, and its ability to improve communication with colleagues and patients. Four attending orthopedic oncologists, one orthopedic oncology fellow, and one senior orthopedic oncology resident participated in the study. ResultsFour cases of failed resection were evaluated by a group of both attending surgeons and a group of trainees composed of both residents and fellows. Tumor borders were clearly delineated in 0 % and 66.6 % cases when evaluating with conventional 2D imaging and VR, respectively. Participants changed adjacent structure involvement grade 22.2 % of the time after assessing involvement grade on the VR technology, with adjacent ligamentous structure grading changed most frequently in 55.5 % of cases. Users reported they would change the surgical approach or margins 44.4 % of the time after reviewing with VR technology. Initial 6 plane resection plans were changed in every user case. Subjective responses indicated that surgeons expressed more confidence in their approach, confidence with obtaining negative margins, and provided more detail regarding structures to be resected in specific planes. ConclusionPelvic tumors present unique surgical challenges due to complex 3D anatomy, the proximity of vital structures, consistency of the tumor, and the need to alter patient position during resection procedures. Using examples of failed pelvic bone sarcoma resections, our study found that VR imaging increased understanding of the tumor environment, characteristics, and ability to communicate with patients and colleagues.

Application of Different Image Processing Methods for Measuring Rock Fracture Structures under Various Confining Stresses

Fractures within granite may become channels for fluid flow and have a significant impact on the safety of waste storage. However, internal aperture variation under coupled conditions are usually difficult to grasp, and the inevitable differences between the measured data and the real fracture structure will lead to erroneous permeability predictions. In this study, two different CT (Computed Tomography) image processing methods are adopted to grasp internal fractures. Several CT images are extracted from different positions of a rock sample under different confining stresses. Two critical factors, i.e., aperture and the contact area ratio value within a single granite fracture sample, are investigated. Results show that aperture difference occurs under these two image processing methods. The contact area ratio value under two image processing methods has less than 1% difference without confining stress. However, there is larger than ten times difference when the confining stress increases to 3.0 MPa. Moreover, the edge detection method has the capability to obtain a relatively accurate internal fracture structure when confining pressure is applied to the rock sample. The analysis results provide a better approach to understanding practical rock fracture variations under various conditions.

A hybrid of supervised and unsupervised deep learning models for multi-vendor kernel conversion of chest CT images

ObjectiveWhen reconstructing a computed tomography (CT) volume, different filter kernels can be used to highlight different structures depending on the medical purpose. The aim of this study was to perform CT conversion for intra-/inter-vendor kernel conversion while preserving image quality. Materials and methodsThis study used CT scans from 632 patients who underwent contrast-enhanced chest CT on either a GE or Siemens scanner. Raw data from each CT scan was reconstructed with Standard and Chest kernels of GE or B10f, B30f, B50f, and B70f kernels of Siemens. In intra-vendor, all images reconstructed with one kernel are paired with another kernel, so the U-Net based supervised method was applied. In the case of inter-vendor where the input and target kernels have each different vendor, Siemens' B30f and GE's Standard kernel were trained through unsupervised image-to-image translation using contrastive learning. ResultsIn the intra-vendor, quantitative evaluation of the image quality of our model showed reasonable performance on the internal test set (structural similarity index measure (SSIM) > 0.96, peak signal-to-noise ratio (PSNR) > 42.55) compared with the SR-block model (SSIM > 0.93, PSNR > 42.92). In the 6-class classification to evaluate the inter-vendor conversion performance, similar accuracy was shown in the converted image (0.977) compared to the original image (0.998). ConclusionsIn this study, we developed a network that can translate a given CT image into a target kernel among multi-vendors. Our model showed clinically acceptable quality in quantitative and qualitative evaluations, including image quality metrics.

Interstitial lung disease associated with anti-aminoacyl-tRNA synthetase syndrome: quantitative evaluation of CT after initial treatment and long-term follow-up.

Visual evaluation of interstitial lung disease (ILD)-related changes can generate intra- and inter-observer errors. However, recent deep learning (DL) algorithm advances have facilitated accurate lung segmentation, lesion characterization, and quantification. To evaluate the treatment response and long-term course in ILD associated with anti-aminoacyl-tRNA synthetase syndrome (anti-ARS ILD) using a DL algorithm. Patients with anti-ARS ILD who underwent both pre- and post-initial-treatment computed tomography (CT) (n = 68) were divided into two groups (responders and non-responders) according to forced vital capacity improvement after initial treatment. We also analyzed the CT images of patients for whom long-term follow-up CT (>5 years) was performed after post-treatment CT (n = 43). DL analysis was used to classify CT imaging features into five patterns: normal; ground-glass opacity (GGO); consolidation; fibrotic lesions; and emphysema. The initial responder group had a larger volume of consolidation. Consolidation and GGO volumes decreased after initial treatment in both groups. However, whole-lung and normal-area volumes increased in the responder group; conversely, there was no significant increase in the non-responder group. At the long-term follow-up, fibrotic lesions significantly increased in both groups. The emphysema pattern increased significantly in both groups after initial treatment and long-term follow-up. Six of 26 (23.1%) responders and 8 of 17 (47.1%) non-responders were judged as having progressive pulmonary fibrosis. DL-based analysis facilitated the chronological evaluation of anti-ARS ILD. During the long-term follow-up, anti-ARS ILD was associated with chronological progression, regardless of initial treatment efficacy.

Development, validation, and application of a generic image-based noise addition method for simulating reduced dose computed tomography images.

Major efforts in computed tomography (CT) have been focused on reducing radiation dose to patients while maintaining adequate diagnostic quality. To that end, research tools have been developed to simulate reduced-dose images via either image-based or projection-based methods. The former is limited to fully capturing realistic texture, streak, and non-stationary characteristics of reduced dose, while the latter is impractical clinically. To develop and validate an image-based noise addition method that accounts for such attributes while being practical in clinical settings. A noise addition method was developed to add realistic noise in the image domain. The method first estimates the noise power spectrum (NPS) of CT images, which are also forward-projected to form synthetic projections. The projection data are supplemented with random white noise proportional to their attenuation values. The noise sinogram is then back-projected onto the image, filtered by the NPS, and scaled according to the desired dose reduction level. The tool was evaluated using both phantom images and patient data. The phantom images were acquired using a multi-sized image quality phantom (Mercury Phantom 3.0, Duke University), and a thorax anthropomorphic phantom (Lungman Phantom, Kyoto Kagaku) at different dose levels and reconstruction settings. The patient images consisted of two dose levels of various CT examinations and reconstruction settings. The simulated and real reduced-dose images were compared in terms of the noise magnitude and texture (i.e., NPS average frequency, NPS-fav). The utility of this methodology was also assessed for routine clinical use for CT protocol review. For the phantom images, the percent errors in the noise magnitude between the simulated images and the actual images of the Mercury Phantom and anthropomorphic phantom images were 3.34% and 3.50%, respectively. The difference in fav was 0.07mm-1 for the Mercury Phantom and 0.06mm-1 for the anthropomorphic phantom between the simulated and actual images. The average noise magnitude percent error between the simulated and actual patient images was 4.61% with noise texture judged to be visually comparable with some kernel dependencies. When implemented clinically, the tool proved practical to simplify the process of estimating radiation dose reduction for CT protocols, resulting in a 50% dose reduction of our multiple myeloma protocol. The method generated simulated CT images with realistic noise properties similar to images acquired at the same radiation exposure without needing access to raw projection data.

Computed tomography of the equine caudal spine and pelvis: Technique, image quality and anatomical variation in 56 clinical cases (2018-2023).

Cross-sectional imaging improves the diagnostic accuracy of complex anatomical regions. Computed tomography (CT) of the pelvis and caudal spine in a large group of live horses and ponies has not been previously reported. To describe the procedure for acquiring CT images of horses' caudal spine/pelvis under general anaesthesia (GA) and to detail the image quality, artefacts and anatomical variations in this region. Retrospective case series. Horses with CT of the caudal spine/pelvis were included. Horses under 6 months and CT examination performed post-mortem were excluded. Protocols, image quality, region of interest, anatomical features and morbidities were analysed. Fifty-six horses (8 months to 20 years, 85-680 kg) met the inclusion criteria. GA ranged from 10 to 60 min (mean: 30, median: 32). There were no adverse events recorded in any of the horses associated with the procedure. Images of all horses were considered of diagnostic quality. Anatomical variations were common and included the location of diverging (widest) interspinous space, the presence of spina bifida in the lumbar and sacral spine, the shape of the last lumbar vertebra and the location of intertransverse joints in terms of where they were present and the degree of fusion/modelling. Not all horses underwent CT examination of the same regions, the upper size limit of horses is unknown and will vary depending on bore size and table infrastructure. Image noise, particularly in large horses and beam hardening artefacts from hardware and pelvis degraded image quality. Images were of insufficient quality in large horses for soft tissue interpretation. CT of the caudal spine and pelvis in live horses with wide-bore CT machines and modified patient infrastructure was safe and produced diagnostic images.

Computed tomography and structured light imaging guided orthopedic navigation puncture system: effective reduction of intraoperative image drift and mismatch.

Image-guided surgical navigation systems are widely regarded as the benchmark for computer-assisted surgical robotic platforms, yet a persistent challenge remains in addressing intraoperative image drift and mismatch. It can significantly impact the accuracy and precision of surgical procedures. Therefore, further research and development are necessary to mitigate this issue and enhance the overall performance of these advanced surgical platforms. The primary objective is to improve the precision of image guided puncture navigation systems by developing a computed tomography (CT) and structured light imaging (SLI) based navigation system. Furthermore, we also aim to quantifying and visualize intraoperative image drift and mismatch in real time and provide feedback to surgeons, ensuring that surgical procedures are executed with accuracy and reliability. A CT-SLI guided orthopedic navigation puncture system was developed. Polymer bandages are employed to pressurize, plasticize, immobilize and toughen the surface of a specimen for surgical operations. Preoperative CT images of the specimen are acquired, a 3D navigation map is reconstructed and a puncture path planned accordingly. During surgery, an SLI module captures and reconstructs the 3D surfaces of both the specimen and a guiding tube for the puncture needle. The SLI reconstructed 3D surface of the specimen is matched to the CT navigation map via two-step point cloud registrations, while the SLI reconstructed 3D surface of the guiding tube is fitted by a cylindrical model, which is in turn aligned with the planned puncture path. The proposed system has been tested and evaluated using 20 formalin-soaked lower limb cadaver specimens preserved at a local hospital. The proposed method achieved image registration RMS errors of 0.576 ± 0.146 mm and 0.407 ± 0.234 mm between preoperative CT and intraoperative SLI surface models and between preoperative and postoperative CT surface models. In addition, preoperative and postoperative specimen surface and skeletal drifts were 0.033 ± 0.272 mm and 0.235 ± 0.197 mm respectively. The results indicate that the proposed method is effective in reducing intraoperative image drift and mismatch. The system also visualizes intraoperative image drift and mismatch, and provides real time visual feedback to surgeons.

Node Reporting and Data System Combined With Computed Tomography Radiomics Can Improve the Prediction of Nonenlarged Lymph Node Metastasis in Gastric Cancer.

To investigate the diagnostic performance of Node Reporting and Data System (Node-RADS) combined with computed tomography (CT) radiomics for assessing nonenlargement regional lymph nodes in gastric cancer (GC). Preoperative CT images were retrospectively collected from 376 pathologically confirmed of gastric adenocarcinoma from January 2019 to December 2023, with 605 lymph nodes included for analysis. They were divided into training (n = 362) and validation (n = 243) sets. Radiomics features were extracted from venous-phase, and the radiomics score was obtained. Clinical information, CT parameters, and Node-RADS classification were collected. A combined model was built using machine-learning approach and tested in validation set using receiver operating characteristic curve analysis. Further validation was conducted in different subgroups of lymph node short-axis diameter (SD) range. Node-RADS score, SD, maximum diameter of thickness of tumor, and radiomics were identified as the most predictive factors. The results demonstrated that the integrated model combining SD, maximum diameter of thickness of tumor, Node-RADS, and radiomics outperformed the model excluding radiomics, yielding an area under the receiver operating characteristic curve of 0.82 compared with 0.79, with a statistically significant difference (P < 0.001). Subgroup analysis based on different SDs of lymph nodes also revealed enhanced diagnostic accuracy when incorporating the radiomics score for the 4- to 7.9-mm subgroups, all P < 0.05. However, for the 8- to 9.9-mm subgroup, the combination of the radiomics did not significantly improve the prediction, with an area under the receiver operating characteristic curve of 0.85 versus 0.85, P = 0.877. The integration of radiomics scores with Node-RADS assessments significantly enhances the accuracy of lymph node metastasis evaluation for GC. This combined model is particularly effective for lymph nodes with smaller standard deviations, yielding a marked improvement in diagnostic precision. The findings of this study indicate that a composite model, which incorporates Node-RADS, radiomics features, and conventional parameters, may serve as an effective method for the assessment of nonenlarged lymph nodes in GC.

The Association between Fatty Liver Index and Lower Limb Arterial Calcification in Patients with Type 2 Diabetes Mellitus.

Peripheral arterial calcification is a prevalent condition in patients with type 2 diabetes mellitus (T2DM), resulting in lower-limb amputation and reduced life quality. Non-alcoholic fatty liver disease (NAFLD), which can be simply evaluated using the fatty liver index (FLI), is closely associated with T2DM development. In this study, we aimed to explore the relationship between FLI and lower limb arterial calcification (LLAC) in T2DM patients and to reveal the value of T2DM patients with NAFLD in predicting the occurrence of LLAC. A total of 77 T2DM patients with LLAC who underwent comprehensive physical and health examinations, serological examinations, as well as lower limb computed tomography imaging at Sun Yat-sen Memorial Hospital of Sun Yat-sen University between January 2018 and January 2019 were enrolled in this study. The FLI was calculated using body mass index, waist circumference, triglycerides, and γ-glutamyl transferase. Additionally, LLAC was evaluated using computed tomography with the Agatston scoring algorithm. The patients were divided into three groups based on their FLI values: Non-liver disease group (FLI <30, n = 29), borderline-liver disease group (30 ≤ FLI < 60, n = 32), and NAFLD group (FLI ≥60, n = 16). Univariate and multivariate binary logistic regression analyses were employed to investigate the association between FLI and LLAC in T2DM patients. Furthermore, differences in LLAC among groups were analyzed using post-hoc multiple comparisons and ordinal logistic regression model analysis. Univariate and multivariate analyses showed that age and FLI influenced LLAC severity in T2DM patients. Moreover, T2DM patients in the NAFLD group had significantly lower LLAC scores than those in the Non-liver disease group. The correlation analysis showed that FLI was negatively associated with LLAC scores (R = -0.31, p = 0.006), while age was positively associated (R = 0.361, p = 0.001). Our study revealed an inverse relationship between FLI and the degree of LLAC. This indicates that, based on evidence in the current research, NAFLD may not be reliable as a predictor of LLAC in T2DM patients.

Partially Ablative Body Radiation Therapy: A Widely Applicable Planning Technique for Palliation of Locally Advanced Unresectable Tumors

Patients with locally advanced, bulky, and unresectable tumors frequently exhibit frailty and endure symptomatic burdens arising from the mass effect of their tumors. Conservative approaches may often fail to provide symptomatic benefits in relatively radioresistant, slower-growing tumors such as sarcomas. A novel technique termed partially ablative body radiation therapy (PABR) administers a highly centralized ablative dose through the utilization of a simultaneous integrated boost while delivering a low and safe palliative dose to the peripheral regions of tumors. The purpose of this paper was to describe a widely applicable radiation therapy protocol in detail for the PABR technique, of which clinical results are available in previous work.7 In summary, a PABR prescription of 20 Gy in 5 fractions is applied to the planning target volume and is planned for 95% of the volume to be covered by 95% of the prescribed dose. A dose of 50 Gy is planned to the boost target volume, with an allowed maximum dose of up to 65 to 70 Gy, using volumetric modulated arc therapy. Daily Cone-Beam Computed Tomography images are used for delivery verification and imaging study. The centrally located volume exceeding 50 Gy effectively achieved the desired outcomes of symptom relief and tumor size reduction. The PABR approach is widely accessible and can be readily implemented in a routine clinical setting to address a pressing need for the challenging palliative patient cohort.

Use of oblique view in periacetabular osteotomy and delayed postoperative weight bearing reduce delayed union in the ischium after one year

This study aimed to determine the efficacy of using both the postero-anterior and oblique image intensifier views intra-operatively and late start of post-operative partial weight-bearing (PWB) in reducing the incidence of delayed union in the ischium at one year after curved periacetabular osteotomy. We evaluated computed tomography images to clarify the incidence of delayed union at the osteotomy sites at one year post-operatively for 117 hips in 104 patients. Fifty-eight hips in 54 patients with use of both the postero-anterior and oblique image intensifier views intra-operatively and late start of post-operative PWB were assigned to the oblique view and late PWB group (OL group) and 59 hips in 50 patients with use of only the postero-anterior image intensifier view intra-operatively and early start of post-operative PWB were assigned to the control group (C group). In univariate analyses, the incidence of delayed union in the ischium at one year post-operatively was significantly lower in the OL group (3.5%) than in the C group (22%). Use of both the postero-anterior and oblique image intensifier views during curved periacetabular osteotomy and late start of PWB were effective for reducing delayed union of the ischium at one year post-operatively.

Accelerated muscle mass estimation from CT images through transfer learning

BackgroundThe cost of labeling to collect training data sets using deep learning is especially high in medical applications compared to other fields. Furthermore, due to variances in images depending on the computed tomography (CT) devices, a deep learning based segmentation model trained with a certain device often does not work with images from a different device.MethodsIn this study, we propose an efficient learning strategy for deep learning models in medical image segmentation. We aim to overcome the difficulties of segmentation in CT images by training a VNet segmentation model which enables rapid labeling of organs in CT images with the model obtained by transfer learning using a small number of manually labeled images, called SEED images. We established a process for generating SEED images and conducting transfer learning a model. We evaluate the performance of various segmentation models such as vanilla UNet, UNETR, Swin-UNETR and VNet. Furthermore, assuming a scenario that a model is repeatedly trained with CT images collected from multiple devices, in which is catastrophic forgetting often occurs, we examine if the performance of our model degrades.ResultsWe show that transfer learning can train a model that does a good job of segmenting muscles with a small number of images. In addition, it was confirmed that VNet shows better performance when comparing the performance of existing semi-automated segmentation tools and other deep learning networks to muscle and liver segmentation tasks. Additionally, we confirmed that VNet is the most robust model to deal with catastrophic forgetting problems.ConclusionIn the 2D CT image segmentation task, we confirmed that the CNN-based network shows better performance than the existing semi-automatic segmentation tool or latest transformer-based networks.

Comparison of Endoscopic Ultrasound and CT Scan in the Diagnosis of Esophageal Duplication Cysts.

Esophageal duplication cysts (EDCs) are rare congenital malformations, often discovered incidentally during endoscopy or on computed tomography (CT) scans. The role of endoscopic ultrasound (EUS) and CT scan in the diagnosis of these lesions and indications for surgical treatment are underreported. The aim of this study was to investigate these topics in a cohort of patients. Between January 2001 and October 2020, 82 patients had a suspicion of esophageal duplication cyst on endoscopic ultrasound. Thirty four of these patients were referred for surgical enucleation of the lesion, but three patients were lost to follow-up. At the end, 31 patients, who underwent surgical treatment for their suspected EDC were included in this study. Clinical features, EUSfindings, CT images, surgical treatment, and outcome were collected from hospital health records. CT images were re-evaluated by a chest radiologist. Type of surgery, surgical complications, and final histological diagnosis were reported. The patients referred for surgery were younger (p = 0.0001) and had larger lesions (> 2cm; p = 0.005) than the patients who had non-operative follow-up. From thirty-one operated patients, eighteen (58%) had post-operative histological diagnosis of duplication cyst. On EUS the final histological diagnosis was correct in 58% (18/31) of all the operated cases and on CT scan 57% (17/30). CT scan misdiagnosed three of the EDCs but found two leiomyomas correctly. None of these patients developed malignancy. According to this study, neither EUS without fine-needle biopsy nor CT scan alone can differentiate EDCs from other mediastinal masses.