Computed Tomography Image Data Research Articles

Triquetrum osteotomy to approach an otherwise inaccessible coronal malunion of the hamate

Fractures of the hamate are rare, and a clear treatment algorithm does not exist. Nonetheless, surgical treatment is generally recommended for displaced fractures using a dorsal approach. There is also a lack of data on hamate malunion. We present a case of a 28-year-old female with a coronal malunion of the hamate and hamate hook fracture. Because the triquetrum prevented direct access to the fracture, we planned and undertook a transtriquetral coronal osteotomy based on three-dimensional computed tomography imaging data. After removing the bone callus, reduction was possible with subsequent fixation. We recommend performing a transtriquetral osteotomy to treat an otherwise inaccessible fracture or malunion of select hamate fractures.

Study on the lateral pterygoid muscle status after artificial temporomandibular joint replacement

The purpose of this study was to explore the status of the lateral pterygoid muscle (LPM) after detachment in artificial temporomandibular joint replacement (TJR) surgery. Patient clinical and computed tomography imaging data were collected before and after unilateral artificial TJR with LPM detachment. The volume of the LPM on the operated and unoperated sides was measured before and after surgery (at 1, 3, 6, 12 months) using ProPlan CMF 3.0 software. The volumes of the LPM on both sides, the patient’s mandibular movements, quality of life (QoL), and pain and diet scores (visual analogue scales) were evaluated and compared at the different follow-up stages. Ten patients were included in the study. After surgery, the volume of the operated LPM was significantly reduced to 60.78% at 3 months (P=0.007), and gradually stabilized to 51.58% at 6 months (P=0.025) and 54.68% at 1 year postoperative (P=0.002). There were no significant LPM volume changes on the unoperated side (P=0.67). Lateral movement of the operated joint was significantly reduced (P=0.021) and correlated with the LPM volume change after surgical detachment (P=0.042). The LPM shrank after detachment in the artificial TJR surgery and the muscle detachment affected the movement of the replaced joint.

An Uncertainty-Aware Transfer Learning-Based Framework for COVID-19 Diagnosis.

The early and reliable detection of COVID-19 infected patients is essential to prevent and limit its outbreak. The PCR tests for COVID-19 detection are not available in many countries, and also, there are genuine concerns about their reliability and performance. Motivated by these shortcomings, this article proposes a deep uncertainty-aware transfer learning framework for COVID-19 detection using medical images. Four popular convolutional neural networks (CNNs), including VGG16, ResNet50, DenseNet121, and InceptionResNetV2, are first applied to extract deep features from chest X-ray and computed tomography (CT) images. Extracted features are then processed by different machine learning and statistical modeling techniques to identify COVID-19 cases. We also calculate and report the epistemic uncertainty of classification results to identify regions where the trained models are not confident about their decisions (out of distribution problem). Comprehensive simulation results for X-ray and CT image data sets indicate that linear support vector machine and neural network models achieve the best results as measured by accuracy, sensitivity, specificity, and area under the receiver operating characteristic (ROC) curve (AUC). Also, it is found that predictive uncertainty estimates are much higher for CT images compared to X-ray images.

Abstract PO-030: Radiomics for head and neck cancer prognostication: results from the RADCURE machine learning challenge

Abstract Recently, there has been significant interest in using rich, multi-modal data routinely collected in the clinic, including imaging, demographic and clinical information for prognostic factor discovery in cancer. In particular, the emerging field of radiomics makes use of computational tools to extract quantitative features from radiological images, with the aim of capturing the morphological and biological characteristics of tumors. Previous studies have demonstrated the potential of computed tomography (CT) imaging features as independent prognostic factors for overall survival in multiple types of cancer, including head and neck (HNC). However, poor reproducibility and lack of large, rigorous validation studies have hindered widespread clinical use of radiomics so far. We conducted a HNC survival prediction challenge with the aim of 1) developing an accurate prognostic model for HNC survival using clinical, demographic and routinely collected CT imaging data and 2) evaluating the true added value of CT radiomics compared to other prognostic factors. Using a large, retrospective cohort of 2552 patients, we assessed prognostic performance of 12 different approaches developed by several research groups at University Health Network in Toronto, making use of engineered radiomics, deep learning, clinical information and combinations of those. To allow for unbiased comparison between different approaches, all participants had access to a public training dataset of 1802 patients, while 750 were held out for evaluation. The best challenge submission used a deep multi-task learning framework on clinical data and tumour volume, achieving area under the ROC curve (AUROC) of 0.812 [95% CI 0.763–0.858] for 2-year survival prediction and concordance (C) index for lifetime risk prediction of 0.795 [0.751–0.838], outperforming the best clinical-only model (AUROC=0.800 [0.749–0.848], C=0.708 [0.661–0.754]), best radiomics-only model (AUROC=0.766 [0.719–0.811], C=0.748 [0.704–0.790]), as well as the best model combining deep radiomics with clinical features (AUROC=0.786 [0.733–0.836], C=0.774 [0.726–0.820]). We also used a ‘wisdom of the crowds’ ensemble approach to combine the predictions of all challenge submissions to determine whether The ensemble achieved stronger performance than any individual model (AUROC=0.823 [0.778–0.864], C=0.810 [0.772–0.845]) indicating that there might be complementary information between the different data modalities. Our rigorous challenge framework allowed us to evaluate a diverse collection of prognostic models in a large multi-modal dataset, demonstrating the value of machine learning in HNC prognostication, as well as the advantages of simple imaging features over several hand-engineered and deep radiomics approaches. Furthemore, our ensemble approach achieves excellent performance for both 2-year and lifetime risk prediction, establishing new state-of-the-art in HNC prognostic modelling. Citation Format: Michal Kazmierski, Mattea Welch, Benjamin Haibe-Kains. Radiomics for head and neck cancer prognostication: results from the RADCURE machine learning challenge [abstract]. In: Proceedings of the AACR Virtual Special Conference on Artificial Intelligence, Diagnosis, and Imaging; 2021 Jan 13-14. Philadelphia (PA): AACR; Clin Cancer Res 2021;27(5_Suppl):Abstract nr PO-030.

Data traffic compression in spectral photon-counting CT imaging based on human visual characteristics

Spectral computed tomography (CT) with photon-counting detectors (PCDs) can provide a variety of cross-sectional images, showing improved diagnostic capability at low radiation doses. PCDs acquire and divide signals of each energy intensity into separate energy bins and improve the count rate by pixel size miniaturization, which can simultaneously increase spatial resolution. Nevertheless, PCDs do show some limitations in data processing. Acquisition of large-spectrum data requires large spaces of an analog-to-digital converter (ADC) and a memory in each pixel, and this process generates massive amounts of data traffic. The conventional compression techniques used in CT imaging are not suitable for such traffic because they are applied to the CT image data, not the projection data stored by PCDs. This study aimed to secure ADC and memory space and compress the data traffic while maintaining high image resolution by using a number of energy bin's information for computing. This approach was based on a composite method using CT images from high spatial resolution non-spectral CT and low spatial resolution spectral CT. The resolution in non-spectral CT was recognized by assigning grayscale values, while the resolution for spectral CT was maintained with a pseudocolor scale based on visual characteristics. In this paper, we evaluated the image quality against the compression rate to verify our concept.

Deep Learning Models for Predicting Severe Progression in COVID-19-Infected Patients: Retrospective Study.

BackgroundMany COVID-19 patients rapidly progress to respiratory failure with a broad range of severities. Identification of high-risk cases is critical for early intervention.ObjectiveThe aim of this study is to develop deep learning models that can rapidly identify high-risk COVID-19 patients based on computed tomography (CT) images and clinical data.MethodsWe analyzed 297 COVID-19 patients from five hospitals in Daegu, South Korea. A mixed artificial convolutional neural network (ACNN) model, combining an artificial neural network for clinical data and a convolutional neural network for 3D CT imaging data, was developed to classify these cases as either high risk of severe progression (ie, event) or low risk (ie, event-free).ResultsUsing the mixed ACNN model, we were able to obtain high classification performance using novel coronavirus pneumonia lesion images (ie, 93.9% accuracy, 80.8% sensitivity, 96.9% specificity, and 0.916 area under the curve [AUC] score) and lung segmentation images (ie, 94.3% accuracy, 74.7% sensitivity, 95.9% specificity, and 0.928 AUC score) for event versus event-free groups.ConclusionsOur study successfully differentiated high-risk cases among COVID-19 patients using imaging and clinical features. The developed model can be used as a predictive tool for interventions in aggressive therapies.

Morphometric Analysis and Three-Dimensional Computed Tomography Reconstruction of Thai Distal Femur

This study evaluates the distal femur morphology of the Thai population using a three-dimensional (3D) measurement method, measuring the distance between the triangular point of the femoral 3D model. The 3D model of 360 Thai femoral obtained from 180 volunteers (90 males, 90 females; range 20–50 years, average 32.8 years) was created using reverse engineering techniques from computed tomography imaging data. Using the 3D identified landmark method, the morphometric parameters evaluated included transepicondylar axis length (TEA), mediolateral length (ML), anteroposterior width (AP), medial anteroposterior width (MAP), lateral anteroposterior width (LAP), medial condyle width (MCW), lateral condyle width (LCW), intercondylar notch width (WIN), intercondylar notch depth (DIN), medial posterior condyle height (MPC), lateral posterior condyle height (LPC), femoral aspect ratio (ML/AP), lateral femoral aspect ratio (ML/LAP), and medial femoral aspect ratio (ML/MAP). The measured data were summarized for the analysis of an average value and standard deviation. Statistical analysis was performed using the independent samples t-test, unequal variances t-test, and linear regression. A p-value less than 0.05 (<0.05) was regarded as statistically significant and indicates strong evidence of the hypothesis. Additionally, the K-means clustering analysis of Thai distal femoral to the optimum size of the prosthesis with the correlation between ML length and AP width was performed. The results found that the morphometric parameters of the Thai male distal femur were significantly different and higher than those of Thai females, except for the ratio of ML/AP and ML/MAP. Comparatively, there was a significant difference between the specific size of Thai distal femur and that of the Korean population, which was also smaller than that for Caucasians. In addition, there was a mismatch between the distal femoral component sizing of knee prosthesis and what is available and commonly used in Thailand. At least six sizes of ML and/or AP should be recommended for the reasonable design of distal femoral prosthesis for covering the anatomy of Thais. These data are useful for predicting the morphometric parameters in forensic anthropology and provide basic data for the design of knee prostheses suitable for the Thai population.

Fabrication and maturation of integrated biphasic anatomic mesenchymal stromal cell‐laden composite scaffolds for osteochondral repair and joint resurfacing

Articular cartilage injury can lead to joint-wide erosion and the early onset of osteoarthritis. To address this, we recently developed a rapid fabrication method to produce patient-specific engineered cartilage tissues to replace an entire articular surface. Here, we extended that work by coupling a mesenchymal stromal cell-laden hydrogel (methacrylated hyaluronic acid) with the porous polycaprolactone (PCL) bone integrating phase and assessed the composition and mechanical performance of these constructs over time. To improve initial construct stability, PCL/hydrogel interface parameters were first optimized by varying PCL pretreatment (with sodium hydroxide before ethanol) before hydrogel infusion. Next, cylindrical osteochondral constructs were formed and cultured in media containing transforming growth factor β3 for up to 8 weeks, with constructs evaluated for viability, histological features, and biochemical content. Mechanical properties were also assessed in axial compression and via an interface shear strength assay. Results showed that the fabrication process was compatible with cell viability, and that construct biochemical content and mechanical properties increased with time. Interestingly, compressive properties peaked at 5 weeks, while interfacial shear properties continued to improve beyond this time point. Finally, these fabrication methods were combined with a custom mold developed from limb-specific computed tomographyimaging data to create an anatomic implantable cell-seeded biologic joint surface, which showedmaturation similar to the osteochondral cylinders. Future work will apply these advances in large animal models of critically sized osteochondral defects to study repair and whole joint resurfacing.

COVID-19 Pandemisinin Acil Bilgisayarlı Tomografi Kullanımına Etkileri: Kesitsel Çalışma

Objective: To investigate the number of performed computed tomography (CT) scans in emergency department at our institution during the coronavirus disease-2019 (COVID-19) pandemic in comparison to the same period of the previous year. Material and Methods: In this retrospective study, we evaluated CT imaging data in the emergency department in April to June timespan in the years of 2019 and 2020. Monthly admitted patient numbers, CT imaging case volumes, age and gender of the patients underwent CT scans at emergency department were noted and compared for each year. For the year 2020, among the CT performed patients who had suspicious and nonsuspicious symptoms for COVID-19 and who had positive reverse transcryptase polymerase chain reaction test were noted for each month. The Student t-test was used for comparison of variables in two groups. Results: In 2020, despite the decrease in total number of patients admitted to emergency department, CT utilization was increased more than 12 times compared to 2019. In comparison to 2019, the mean age of CT performed patients was significantly younger in 2020 (p<0.001). Among the CT performed patients, brain (68.5%) and chest (88.6%) were the most scanned anatomical sites in 2019 and 2020, respectively. Based on couple comparisons for all anatomical sites, there was a significant increase in chest CT utilization in 2020 (p<0.001). Conclusion: During the COVID-19 pandemic CT usages increased, a higher rate of chest CT was performed and younger patients were scanned in emergency department compared with 2019.

Clinical implication of psoas muscle mass measurement for predicting death within 3 months in patients with incurable uterine cervical or corpus malignancy

The aim of this study was to find a clinical marker for identifying refractory cancer cachexia. We analyzed computed tomography imaging data, which included the third lumbar vertebra, from 94 patients who died of uterine cervix or corpus malignancy. The time between the date of examination and date of death was the most important attribute for this study, and the computed tomography images were classified into >3months before death and ≤3months before death. Psoas muscle mass index was defined as the left-right sum of the psoas muscle areas (cm2) at the level of third lumbar vertebra, divided by height squared (m2). A data set of 94 computed tomography images was obtained at baseline hospital visit, and a data set of 603 images was obtained at other times. One hundred (16.6%) of the 603 non-baseline images were scanned ≤3months before death. Mean psoas muscle mass index change rates at >3months before death and ≤3months before death were -1.3 and -20.1%, respectively (P<0.001). Receiver operating characteristic curve analysis yielded a cutoff value of -13.0%. The area under the curve reached a moderate accuracy level (0.777, 95% confidence interval 0.715-0.838). When we used the cutoff value to predict death within 3months, sensitivity and specificity were 74.0 and 82.1%, respectively. Measuring change in psoas muscle mass index might be useful for predicting cancer mortality within 3months. It could become a potential tool for identifying refractory cancer cachexia.

Clinical importance of muscle volume in lenvatinib treatment for hepatocellular carcinoma: Analysis adjusted with inverse probability weighting.

Background and AimThis study aimed to elucidate the clinical importance of muscle volume loss (pre‐sarcopenia) in patients receiving lenvatinib as treatment for unresectable hepatocellular carcinoma (u‐HCC).MethodsOf 437 u‐HCC patients treated with lenvatinib at specific institutions in Japan between March 2018 and May 2020, 151 with available computed tomography imaging data from the time of lenvatinib introduction were enrolled. Pre‐sarcopenia was diagnosed based on a previously reported cut‐off value calculation formula [psoas muscle area at level of middle of third lumbar vertebra (cm2)/height (m)2]. Clinical features and prognostic factors for overall survival (OS) with inverse probability weighting were investigated retrospectively for their relationship with pre‐sarcopenia.ResultsCox hazard multivariate analysis showed alpha‐fetoprotein (≥400 ng/mL) (hazard ratio [HR] 2.271, P < 0.001), Barcelona Clinic Liver Cancer stage (C and D) (HR 1.625, P = 0.018), and positive for pre‐sarcopenia (HR 1.652, P = 0.042) to be significant prognostic factors. OS rates for the pre‐sarcopenia group (n = 41) were worse than those for the non‐pre‐sarcopenia group (n = 110) (0.5‐, 1‐, and 1.5‐year OS: 72.5%, 27.9%, and 7.0% vs 80.7%, 56.7%, and 46.1%, respectively; P < 0.001), as was progression‐free survival (P = 0.025). Time to stopping lenvatinib or disease progression was better in the non‐pre‐sarcopenia group (0.5‐, 1‐, and 1.5‐year OS: 48.0%, 24.5%, and 8.4% vs 20.0%, 10.3%, and 4.2%, respectively; P < 0.001). Also, the frequency of the adverse event appetite loss (any grade) was greater in the pre‐sarcopenia group (43.9% vs 18.2%, P = 0.003).ConclusionPre‐sarcopenia was shown to be a significant prognostic factor in patients treated with lenvatinib for u‐HCC.

CT-Scan as an Opportunistic Osteoporosis Screening Tool in High-Risk Populations; A Diagnostic Study Conducted on a Sample of Urinary Stone Patients

Background: Osteoporosis is a globally common disorder with significant burdens on the quality of life, and therefore, presents the need for a viable screening tool. Objectives: To establish the diagnostic value of opportunistic computed tomography (CT) scan performed routinely on urinary stone patients for identifying and screening osteoporosis, through measuring vertebral trabecular bone attunement in Hounsfield Units. Patients and Methods: Consecutive subjects were enrolled in this cross-sectional study, who suffered from urinary stones and were undergoing percutaneous nephrolithotomy (PCNL) procedure for stone removal at a tertiary referral hospital. The CT imaging data for the patients were retrospectively extracted, and the patients then underwent a dual energy X-ray absorptiometry (DEXA) scan as the standard protocol in measuring bone mineral density (BMD). CTs were assessed, measuring the bone mineral attenuation at the L1 vertebral level. Diagnostic performance protocols and data comparison with DEXA results were measured. Results: Ninety-two patients were enrolled in the study, 31 (34%) of which were designated as osteoporotic according to standard definitions based on DEXA. Vertebral attunement measurements showed sufficient diagnostic performance, with an area under the curve (AUC) of 0.983 (95% confidence interval [CI], 0.961 – 1.000), 93.55% sensitivity (95% CI, 78.55 – 99.21) and 90.16% specificity (95% CI, 79.81 – 96.30) at the pre-designed 160 Hounsfield units (HU) discriminatory threshold. Conclusion: This study showed that in a high-risk population, vertebral trabecular bone attenuation values retrieved from routine CT imaging holds significant diagnostic value when compared to the gold-standard DEXA, and can be utilized as a viable screening method.

Long-term Stress Distribution Patterns Across the Ankle Joint in Soccer Players: A Computed Tomography Osteoabsorptiometry Study.

Background:The distribution pattern of subchondral bone density is considered to accurately reflect the stress distribution over a joint under long-term physiologic loading. The biomechanical characteristics of the surface of the ankle joint in soccer players can be determined by measuring this distribution pattern under long-term loading.Purpose:To evaluate the distribution of subchondral bone density across the ankle joint in soccer players and to determine the effects of soccer activities, including kicking motion, on the ankle joint surface under long-term loading conditions by computed tomography (CT) osteoabsorptiometry (CTOAM).Study Design:Cross-sectional study; Level of evidence, 3.Methods:CT imaging data were obtained from both ankles of 10 soccer players (soccer group) and 10 nonathletic volunteers (control group). The distribution patterns of subchondral bone density across the articular surface of the ankle joints were assessed by CTOAM. Quantitative analysis was performed of the locations and percentages of high-density areas on the articular surface.Results:Stress distribution patterns over the ankle joint differed between the soccer players and controls. In the soccer players, the high-density areas were found in the anterior part of the distal tibia and proximal talus as well as the distal fibula. The percentages of high-density areas were greater in the soccer players compared with controls (P < .0001).Conclusion:Stress distribution over the articular surface of the ankle joint was affected by soccer activities. A high stress concentration was seen in soccer players in the anterior part of the tibia and talus and in the fibula; such excessive stress may lead to anterior impingement.

Clinical and Imaging Features of Novel Coronavirus Pneumonia in the Elderly Compared with Young and Middle-Aged Individuals

Background: A cluster of patients with coronavirus disease 2019 (COVID-19) pneumonia caused by infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) were reported worldwide, but there are few reports on age-related computed tomography (CT) signs and clinical characteristics of COVID-19 patients. Objectives: We aimed to analyze the CT imaging characteristics and clinical features of elderly patients with novel COVID-19 and compare them with young COVID-19 patients, in order to provide imaging basis for clinical diagnosis. Patients and Methods: The clinical and chest CT imaging data of elderly patients (≥ 60 years) and young patients (age of 20 - 60 years) with positive novel coronavirus nucleic acid admitted to our hospital from January 28, 2020 to February 16, 2020 were retrospectively analyzed, and the differences in clinical and imaging characteristics between the two groups were compared. Results: One hundred eighty three COVID-19 patients (96 elderly: mean age, 72.76 ± 5.65; 87 young: mean age, 42.69 ± 12.44) with positive novel coronavirus nucleic acid were retrospectively enrolled. The clinical symptoms of fever (79%), cough (64%) and weakness (59%) in elderly patients were higher than those of young patients (P &lt; 0.01). Asymptomatic young patients (15%) were more common than the elderly (2%) (P &lt; 0.01). Twenty-four patients (13%) had lesions that involved both the periphery and the center of the lung, while the lesions were located in the periphery of the lung in 159 patients (87%). Ground-glass opacification in young patients (71%) was significantly higher than that of the elderly (26%) (P &lt; 0.001). Compared with the young group (37%), the proportion of extensive involvement of lung lobes was higher in the elderly group (72%) (P &lt; 0.001). Conclusion: Elderly and young patients with corona virus disease had some common CT and clinical features, but elderly patients may have relatively severe CT findings, and these differences may be related to the severity and prognosis of COVID-19 pneumonia.

Producing three-dimensional printed models of the hepatobiliary system from computed tomography imaging data.

Macroscopic anatomy has traditionally been taught using cadaveric material, lectures and a variety of additional resources including online modules and anatomical models. Traditional plastic models are effective educational tools yet they have significant drawbacks such as a lack of anatomical detail, a lack of texturisation and cost. Three-dimensional printed models stand to solve these problems and widen access to high-quality anatomical teaching. This paper outlines the use of three-dimensional multiplanar imaging (computed tomography) in the development of an accurate model of the hepatobiliary system. Computed tomography scans were used to construct a virtual three-dimensional model of the hepatobiliary system. This was printed locally as a full-size colour model. We give a complete account of the process and software used. This study is among the first of a series in which we will document the newly formed Oxford Library of Anatomy. This series will provide the methodology for the production of three-dimensional models from computed tomography and magnetic resonance imaging scans, and the library will provide a complete collection of the most complex anatomical areas. We hope that these models will form an important adjunct in teaching anatomy to medical students and surgical trainees.

Analysis and improvement of the three-column spinal theory

BackgroundDenis and Ferguson et al.’s three-column spinal theory has been widely accepted and applied. However, this three-column theory was proposed based solely on observation and experience without thorough documented data and analysis. The aim of this study was to analyze and improve Denis and Ferguson et al.’s three-column spinal theory to propose a novel three-column concept in epidemiology, morphology and biomechanics.MethodsA retrospective analysis of the computed tomography imaging data of patients with a diagnosis of T11-L5 vertebral fractures was conducted between February 2010 and December 2018. Three-dimensional (3D) distribution maps of fracture lines of all subjects were obtained based on 3D mapping techniques. In addition, a 25-year-old health male volunteer was recruited for the vertebral finite element force analysis.ResultsThe present study enrolled 459 patients (age: 48 ± 11.42 years), containing a total of 521 fractured vertebrae. The fracture lines peaked in the upper and the outer third sections of the vertebra, starting from the anterior part of the vertebral pedicles in 3-D maps. Regarding flexion and extension of the spine, the last third of the vertebral body in front of the spinal canal was one main stress center in the finite element analysis. The stress on the vertebral body was greater in front of the pedicles in the lateral bending.ConclusionThe study reveals that the posterior one-third of the vertebral body in front of the spinal canal and the posterior one-third of the vertebral body in front of the pedicle are very different in terms of fracture characteristics and risks to spinal canal (3D maps and stress distributing graphs), therefore, they should be classified as different columns. We provide strong evidence that Su’s three-column theory complies with the characteristics of vertebral physiological structure, vertebral fracture, and vertebral biomechanics.Graphical abstract

A typical computed tomography presentations of coronavirus disease 2019

ObjectiveThis study aimed to investigate the atypical computed tomography (CT) presentations of coronavirus disease 2019 (COVID-19) patients to comprehensively understand this highly infectious disease. MethodsThe clinical and chest CT imaging data of 16 patients diagnosed with COVID-19 were retrospectively analyzed, and patients with atypical CT presentations were selected for analysis and review. ResultsOf the 16 patients, 6 had atypical CT presentations, including 2 with faint ground glass opacities, 2 with single nodule, 1 with predominantly linear opacities, and 1 with predominantly reticular opacities. The dynamic changes of CT showed the faint ground glass opacities gradually became weak (2 cases). The scope of the single nodule was enlarged, and it developed into consolidation and residual fibrosis (2 cases). There was no obvious change of linear opacity (1 case). The reticular opacities were enlarged, then partially absorbed and new developed ground-glass opacities were found. Finally, the lesions were absorbed with residual fibrosis (1 case). ConclusionAtypical CT presentations of COVID-19 can be classified as faint ground glass opacities, single nodule, linear opacities, and reticular opacities. Understanding the atypical presentation of COVID-19 is beneficial in the assessment and epidemic prevention and control of this disease.

Quantitative Serial CT Imaging-Derived Features Improve Prediction of Malignant Cerebral Edema after Ischemic Stroke.

Malignant cerebral edema develops in a small subset of patients with hemispheric strokes, precipitating deterioration and death if decompressive hemicraniectomy (DHC) is not performed in a timely manner. Predicting which stroke patients will develop malignant edema is imprecise based on clinical data alone. Head computed tomography (CT) imaging is often performed at baseline and 24-h. We determined the incremental value of incorporating imaging-derived features from serial CTs to enhance prediction of malignant edema. We identified hemispheric stroke patients at three sites with NIHSS ≥ 7 who had baseline as well as 24-h clinical and CT imaging data. We extracted quantitative imaging features from baseline and follow-up CTs, including CSF volume, intracranial reserve (CSF/cranial volume), as well as midline shift (MLS) and infarct-related hypodensity volume. Potentially lethal malignant edema was defined as requiring DHC or dying with MLS over 5-mm. We built machine-learning models using logistic regression first with baseline data and then adding 24-h data including reduction in CSF volume (ΔCSF). Model performance was evaluated with cross-validation using metrics of recall (sensitivity), precision (predictive value), as well as area under receiver-operating-characteristic and precision-recall curves (AUROC, AUPRC). Twenty of 361 patients (6%) died or underwent DHC. Baseline clinical variables alone had recall of 60% with low precision (7%), AUROC 0.59, AUPRC 0.15. Adding baseline intracranial reserve improved recall to 80% and AUROC to 0.82 but precision remained only 16% (AUPRC 0.28). Incorporating ΔCSF improved AUPRC to 0.53 (AUROC 0.91) while all imaging features further improved prediction (recall 90%, precision 38%, AUROC 0.96, AUPRC 0.66). Incorporating quantitative CT-based imaging features from baseline and 24-h CT enhances identification of patients with malignant edema needing DHC. Further refinements and external validation of such imaging-based machine-learning models are required.

Accuracy and Early Clinical Outcome After 3-Dimensional Correction of Distal Radius Intra-Articular Malunions Using Patient-Specific Instruments

To investigate the residual articular incongruity on computed tomography image data and the early clinical outcome of 3-dimensional planned and navigated intra-articular osteotomies of the distal radius. We conducted a retrospective analysis of intra-articular osteotomies executed between 2008 and 2016. We identified 37 patients (aged 26-73 years) and performed a combined intra-articular and extra-articular osteotomy on 20 patients. A preoperative 3-dimensional plan with the superimposed bone of the contralateral healthy side was performed in each case to analyze and execute the osteotomy by intraoperative navigation. The residual articular incongruity was assessed by quantification of the maximal stepoff in the coronal or sagittal computed tomography scans. Clinical outcome, including range of motion, grip strength, and return to work, was assessed after a minimum follow-up of 12 months and compared with preoperative measurements. On average, the preoperative intra-articular stepoff was 2.5 mm (±0.6 mm; range, 1.4-4.2 mm) and was significantly reduced to 0.8 mm (±0.2 mm) after surgery. After surgery, 30 patients had a stepoff less than 1 mm; in 7, a stepoff of 1.1 to 1.4 mm was measured. After 1 year, 22 had no pain, 9 had slight pain during heavy work, and 5 had moderate pain with no improvement compared with their preoperative status, although wrist strength and range of motion improved in all 37 patients. One patient underwent a secondary radioscapholunate arthrodeses owing to persistent pain despite a congruent joint with a small residual intra-articular stepoff (0.6 mm). Intra-articular osteotomies of the distal radius treated by 3-dimensional preoperative planning and patient-specific guides are an accurate technique to reduce articular incongruity to an average stepoff of 0.8 mm (range, 0.3-1.4 mm). The early clinical outcomes demonstrated overall reduction in pain and improvement of range of motion and grip strength in 36 of 37 patients. Therapeutic IV.

A novel model for predicting the outcome of intracerebral hemorrhage: Based on 1186 Patients

ObjectiveTo establish a model for predicting the outcome according to the clinical and computed tomography(CT) image data of patients with intracerebral hemorrhage(ICH). MethodsThe clinical and CT image data of the patients with ICH in Qinghai Provincial People's Hospital and Xuzhou Central Hospital were collected. The risk factors related to the poor outcome of the patients were determined by univariate and multivariate logistic regression analysis. To determine the effect of factors related to poor outcome, the nomogram model was made by software of R 3.5.2 and the support vector machine operation was completed by software of SPSS Modelor. ResultsA total of 8265 patients were collected and 1186 patients met the criteria of the study. Age, hospitalization days, blend sign, intraventricular extension, subarachnoid hemorrhage, midline shift, diabetes and baseline hematoma volume were independent predictors of poor outcome. Among these factors, baseline hematoma volume๥20ml (odds ratio:13.706, 95% confidence interval:9.070-20.709, p < 0.001) was the most significant factor for poor outcome, followed by the volume among 10ml-20ml (odds ratio:11.834, 95% confidence interval:7.909-17.707, p < 0.001). It was concluded that the highest percentage of weight in outcome was baseline hematoma volume (25.0%), followed by intraventricular hemorrhage (23.0%). ConclusionThis predictive model might accurately predict the outcome of patients with ICH. It might have a wide range of application prospects in clinical.