METHOD: Two large datasets, including 1110 3D CT images, were split into five segments of 20% each. Each dataset's first 20% segment was separated as a holdout test set. 3D-CNN training was performed with the remaining 80% from each dataset. Two small external datasets were also used to independently evaluate the trained models.

RESULTS: The total combination of 80% of each dataset has an accuracy of 91% on Iranmehr and 83% on Moscow holdout test datasets. Results indicated that 80% of the primary datasets are adequate for fully training a model. The additional fine-tuning using 40% of a secondary dataset helps the model generalize to a third, unseen dataset. The highest accuracy achieved through transfer learning was 85% on LDCT dataset and 83% on Iranmehr holdout test sets when retrained on 80% of Iranmehr dataset.

METHOD: Prevalence of the anterior ethmoid genu, its morphology and its relationship with the frontal sinus drainage pathway was assessed. Computed tomography scans with multiplanar reconstruction were used to study non-diseased sinonasal complexes.

RESULTS: The anterior ethmoidal genu was present in all 102 anatomical sides studied, independent of age, gender and race. Its position was within the frontal sinus drainage pathway, and the drainage pathway was medial to it in 98 of 102 cases. The anterior ethmoidal genu sometimes extended laterally and formed a recess bounded by the lamina papyracea laterally, by the uncinate process anteriorly and by the bulla ethmoidalis posteriorly. Distance of the anterior ethmoidal genu to frontal ostia can be determined by the height of the posterior wall of the agger nasi cell rather than its volume or other dimensions.

METHODS: In total, 154 patients (wild-type EGFR, 72 patients; Del19 mutation, 45 patients; and L858R mutation, 37 patients) were retrospectively enrolled and randomly divided into 92 training and 62 test cases. Two support vector machine (SVM) models to distinguish between wild-type and mutant EGFR (mutation [M] classification) as well as between the Del19 and L858R subtypes (subtype [S] classification) were trained using 3DBN features. These features were computed from 3DBN maps by using histogram and texture analyses. The 3DBN maps were generated using computed tomography (CT) images based on the Čech complex constructed on sets of points in the images. These points were defined by coordinates of voxels with CT values higher than several threshold values. The M classification model was built using image features and demographic parameters of sex and smoking status. The SVM models were evaluated by determining their classification accuracies. The feasibility of the 3DBN model was compared with those of conventional radiomic models based on pseudo-3D BN (p3DBN), two-dimensional BN (2DBN), and CT and wavelet-decomposition (WD) images. The validation of the model was repeated with 100 times random sampling.

RESULTS: The mean test accuracies for M classification with 3DBN, p3DBN, 2DBN, CT, and WD images were 0.810, 0.733, 0.838, 0.782, and 0.799, respectively. The mean test accuracies for S classification with 3DBN, p3DBN, 2DBN, CT, and WD images were 0.773, 0.694, 0.657, 0.581, and 0.696, respectively.

METHODS: Medline and Embase databases were searched without date restriction on May 2022 for articles that examined EAT and cardiovascular outcomes. The inclusion criteria were (1) studies measuring EAT of adult patients at baseline and (2) reporting follow-up data on study outcomes of interest. The primary study outcome was major adverse cardiovascular events. Secondary study outcomes included cardiac death, myocardial infarction, coronary revascularization, and atrial fibrillation.

RESULTS: Twenty-nine articles published between 2012 and 2022, comprising 19 709 patients, were included in our analysis. Increased EAT thickness and volume were associated with higher risks of cardiac death (odds ratio, 2.53 [95% CI, 1.17-5.44]; P=0.020; n=4), myocardial infarction (odds ratio, 2.63 [95% CI, 1.39-4.96]; P=0.003; n=5), coronary revascularization (odds ratio, 2.99 [95% CI, 1.64-5.44]; P<0.001; n=5), and atrial fibrillation (adjusted odds ratio, 4.04 [95% CI, 3.06-5.32]; P<0.001; n=3). For 1 unit increment in the continuous measure of EAT, computed tomography volumetric quantification (adjusted hazard ratio, 1.74 [95% CI, 1.42-2.13]; P<0.001) and echocardiographic thickness quantification (adjusted hazard ratio, 1.20 [95% CI, 1.09-1.32]; P<0.001) conferred an increased risk of major adverse cardiovascular events.

CONCLUSIONS: The utility of EAT as an imaging biomarker for predicting and prognosticating cardiovascular disease is promising, with increased EAT thickness and volume being identified as independent predictors of major adverse cardiovascular events.

OBJECTIVE: The aim of this study is to evaluate the ICV in primary craniosynostosis patients after the cranial vault reshaping with or without FOA and to compare between syndromic and nonsyndromic synostosis group, to determine factors that associated with significant changes in the ICV postoperative, and to evaluate the resolution of copper beaten sign and improvement in neurodevelopmental delay after the surgery.

METHODS: This is a prospective observational study of all primary craniosynostosis patients who underwent operation cranial vault reshaping with or without FOA in Hospital Kuala Lumpur from January 2017 until Jun 2018. The ICV preoperative and postoperative was measured using the 3D computed tomography (CT) imaging and analyzed. The demographic data, clinical and radiological findings were identified and analyzed.

RESULTS: A total of 14 cases (6 males and 8 females) with 28 3D CT scans were identified. The mean age of patients was 23 months. Seven patients were having syndromic synostosis (4 Crouzon syndromes and 3 Apert syndromes) and 7 nonsyndromic synostosis. The mean preoperative ICV was 880 mL (range, 641-1234 mL), whereas the mean postoperative ICV was 1081 mL (range, 811-1385 mL). The difference was 201 mL which was statistically significant (P  1.0). However, there was 100% (n = 13) improvement of this copper beaten sign. However, the neurodevelopmental delay showed no improvement which was statistically not significant (P > 1.0).

METHODS: We propose to use Residual Blocks with a 3 × 3 kernel size for local feature extraction and Non-Local Blocks to extract the global features. The Non-Local Block has the ability to extract global features without using a huge number of parameters. The key idea behind the Non-Local Block is to apply matrix multiplications between features on the same feature maps.

METHODS: In this study, we propose a multi-view secondary input residual (MV-SIR) convolutional neural network model for 3D lung nodule segmentation using the Lung Image Database Consortium and Image Database Resource Initiative (LIDC-IDRI) dataset of chest computed tomography (CT) images. Lung nodule cubes are prepared from the sample CT images. Further, from the axial, coronal, and sagittal perspectives, multi-view patches are generated with randomly selected voxels in the lung nodule cubes as centers. Our model consists of six submodels, which enable learning of 3D lung nodules sliced into three views of features; each submodel extracts voxel heterogeneity and shape heterogeneity features. We convert the segmentation of 3D lung nodules into voxel classification by inputting the multi-view patches into the model and determine whether the voxel points belong to the nodule. The structure of the secondary input residual submodel comprises a residual block followed by a secondary input module. We integrate the six submodels to classify whether voxel points belong to nodules, and then reconstruct the segmentation image.

RESULTS: The results of tests conducted using our model and comparison with other existing CNN models indicate that the MV-SIR model achieves excellent results in the 3D segmentation of pulmonary nodules, with a Dice coefficient of 0.926 and an average surface distance of 0.072.