Automatic grading of knee osteoarthritis with a plain radiograph radiomics model: combining anteroposterior and lateral images.

OBJECTIVES: To establish a radiomics-based automatic grading model for knee osteoarthritis (OA) and evaluate the influence of different body positions on the model's effectiveness. MATERIALS AND METHODS: Plain radiographs of a total of 473 pairs of knee joints from 473 patients (May 2020 to July 2021) were retrospectively analyzed. Each knee joint included anteroposterior (AP) and lateral (LAT) images which were randomly assigned to the training cohort and the testing cohort at a ratio of 7:3. First, an assessment of knee OA severity was done by two independent radiologists with Kallgren-Lawrence grading scale. Then, another two radiologists independently delineated the region of interest for radiomic feature extraction and selection. The radiomic classification features were dimensionally reduced and a machine model was conducted using logistic regression (LR). Finally, the classification efficiency of the model was evaluated using receiver operating characteristic curves and the area under the curve (AUC). RESULTS: The AUC (macro/micro) of the model using a combination of AP and LAT (AP&LAT) images were 0.772/0.778, 0.818/0.799, and 0.864/0.879, respectively. The radiomic features from the combined images achieved better classification performance than the individual position image (p < 0.05). The overall accuracy of the radiomic model with AP&LAT images was 0.727 compared to 0.712 and 0.417 for radiologists with 4 years and 2 years of musculoskeletal diagnostic experience. CONCLUSIONS: A radiomic model constructed by combining the AP&LAT images of the knee joint can better grade knee OA and assist clinicians in accurate diagnosis and treatment. CRITICAL RELEVANCE STATEMENT: A radiomic model based on plain radiographs accurately grades knee OA severity. By utilizing the LR classifier and combining AP&LAT images, it improves accuracy and consistency in grading, aiding clinical decision-making, and treatment planning. KEY POINTS: Radiomic model performed more accurately in K/L grading of knee OA than junior radiologists. Radiomic features from the combined images achieved better classification performance than the individual position image. A radiomic model can improve the grading of knee OA and assist in diagnosis and treatment.

Haplotype-based phylogenetic analysis and population genomics uncover the origin and domestication of sweetpotato.

The hexaploid sweetpotato (Ipomoea batatas) is one of the most important root crops worldwide. However, its genetic origin remains controversial, and its domestication history remains unknown. In this study, we used a range of genetic evidence and a newly developed haplotype-based phylogenetic analysis to identify two probable progenitors of sweetpotato. The diploid progenitor was likely closely related to Ipomoea aequatoriensis and contributed the B1 subgenome, IbT-DNA2, and the lineage 1 type of chloroplast genome to sweetpotato. The tetraploid progenitor of sweetpotato was most likely I. batatas 4x, which donated the B2 subgenome, IbT-DNA1, and the lineage 2 type of chloroplast genome. Sweetpotato most likely originated from reciprocal crosses between the diploid and tetraploid progenitors, followed by a subsequent whole-genome duplication. In addition, we detected biased gene exchanges between the subgenomes; the rate of B1 to B2 subgenome conversions was nearly three times higher than that of B2 to B1 subgenome conversions. Our analyses revealed that genes involved in storage root formation, maintenance of genome stability, biotic resistance, sugar transport, and potassium uptake were selected during the speciation and domestication of sweetpotato. This study sheds light on the evolution of sweetpotato and paves the way for improvement of this crop.

Association between social support and mutual-support needs among the rural adults in China: a cross-sectional study.

Background: In rural China, there is now a huge gap between the supply and demand for old-age care. To close the gap, developing rural mutual old-age services is extremely important. The purpose of this study is to clarify the relationship among social support, mutual support need, and mutual support willingness. Methods: We conducted an online questionnaire survey using a Chinese Internet research company; 2,102 valid responses were received. The measures comprised the Social Support Rating Scale, the Mutual Support Willingness Questionnaire, and the Mutual Support Needs Scale. We calculated Pearson correlations to explore the association of social support with mutual-support need and mutual-support-need willingness. Multivariate analyses were also conducted using these factors as dependent variables. Results: The total score for the mutual support need for the adults in rural areas was 58.0 ± 12.1 and 36.96 ± 6.40 for social support, approximately 86.8% of the participants were willing to participate in mutual support. Furthermore, mutual support needs were positively correlated with subjective support (p < 0.01) and support utilization (p < 0.01), but negatively correlated with willingness to support each other (p < 0.05). The need for mutual support was also associated with age, sex, education level, dissatisfaction with the current economic situation, health status, and so on. Conclusion: It is necessary for government and health care providers to assess the different needs of rural older people and encourage individuals and organizations to provide mutual support for older people, especially to enhance emotional care for older people and improve their use of support. This is of great significance for developing mutual support services in rural China.

Deep learning-assisted knee osteoarthritis automatic grading on plain radiographs: the value of multiview X-ray images and prior knowledge.

Background: Knee osteoarthritis (OA) is harmful to people's health. Effective treatment depends on accurate diagnosis and grading. This study aimed to assess the performance of a deep learning (DL) algorithm based on plain radiographs in detecting knee OA and to investigate the effect of multiview images and prior knowledge on diagnostic performance. Methods: In total, 4,200 paired knee joint X-ray images from 1,846 patients (July 2017 to July 2020) were retrospectively analyzed. Kellgren-Lawrence (K-L) grading was used as the gold standard for knee OA evaluation by expert radiologists. The DL method was used to analyze the performance of anteroposterior and lateral plain radiographs combined with prior zonal segmentation to diagnose knee OA. Four groups of DL models were established according to whether they adopted multiview images and automatic zonal segmentation as the DL prior knowledge. Receiver operating curve analysis was used to assess the diagnostic performance of 4 different DL models. Results: The DL model with multiview images and prior knowledge obtained the best classification performance among the 4 DL models in the testing cohort, with a microaverage area under the receiver operating curve (AUC) and macroaverage AUC of 0.96 and 0.95, respectively. The overall accuracy of the DL model with multiview images and prior knowledge was 0.96 compared to 0.86 for an experienced radiologist. The combined use of anteroposterior and lateral images and prior zonal segmentation affected diagnostic performance. Conclusions: The DL model accurately detected and classified the K-L grading of knee OA. Additionally, multiview X-ray images and prior knowledge improved classification efficacy.

Sweet potato ADP-glucose pyrophosphorylase small subunit affects vegetative growth, starch content and storage root yield.

The development of storage roots is a key factor determining the yields of crop plants, including sweet potato. Here, using combined bioinformatic and genomic approaches, we identified a sweet potato yield-related gene, ADP-glucose pyrophosphorylase (AGP) small subunit (IbAPS). We found that IbAPS positively affects AGP activity, transitory starch biosynthesis, leaf development, chlorophyll metabolism, and photosynthesis, ultimately affecting the source strength. IbAPS overexpression in sweet potato led to increased vegetative biomass and storage root yield. RNAi of IbAPS resulted in reduced vegetative biomass, accompanied with a slender stature and stunted root development. In addition to the effects on root starch metabolism, we found that IbAPS affects other storage root development-associated events, including lignification, cell expansion, transcriptional regulation, and production of the storage protein sporamins. A combinatorial analysis based on transcriptomes, as well as morphological and physiological data, revealed that IbAPS affects several pathways that determine development of vegetative tissues and storage roots. Our work establishes an important role of IbAPS in concurrent control of carbohydrate metabolism, plant growth, and storage root yield. We showed that upregulation of IbAPS results in superior sweet potato with increased green biomass, starch content, and storage root yield. The findings expand our understanding of the functions of AGP enzymes and advances our ability to increase the yield of sweet potato and, perhaps, other crop plants.

Comparative Effectiveness of MRI, 4D-CT and Ultrasonography in Patients with Secondary Hyperparathyroidism.

Objective: Accurate preoperative localization of abnormal parathyroid glands is crucial for successful surgical management of secondary hyperparathyroidism (SHPT). This study was conducted to compare the effectiveness of preoperative MRI, 4D-CT, and ultrasonography (US) in localizing parathyroid lesions in patients with SHPT. Methods: We performed a retrospective review of prospectively collected data from a tertiary-care hospital and identified 52 patients who received preoperative MRI and/or 4D-CT and/or US and/or 99mTc-MIBI and subsequently underwent surgery for SHPT between May 2013 and March 2020. The sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of each imaging modality to accurately detect enlarged parathyroid glands were determined using histopathology as the criterion standard with confirmation using the postoperative biochemical response. Results: A total of 198 lesions were identified intraoperatively among the 52 patients included in this investigation. MRI outperformed 4D-CT and US in terms of sensitivity (P < 0.01), specificity (P = 0.455), PPV (P = 0.753), and NPV (P = 0.185). The sensitivity and specificity for MRI, 4D-CT, and US were 90.91%, 88.95%, and 66.23% and 58.33%, 63.64%, and 50.00%, respectively. The PPV of combined MRI and 4D-CT (96.52%) was the highest among the combined 2 modalities. The smallest diameter of the parathyroid gland precisely localized by MRI was 8×3 mm, 5×5 mm by 4D-CT, and 5×3 mm by US. Conclusion: MRI has superior diagnostic performance compared with other modalities as a first-line imaging study for patients undergoing renal hyperparathyroidism, especially for ectopic or small parathyroid lesions. We suggest performing US first for diagnosis and then MRI to make a precise localization, and MRI proved to be very helpful in achieving a high success rate in the surgical treatment of renal hyperparathyroidism in our own experience.

Ultrashort echo time magnetization transfer imaging of knee cartilage and meniscus after long-distance running.

OBJECTIVE: To assess the detection of changes in knee cartilage and meniscus of amateur marathon runners before and after long-distance running using a 3D ultrashort echo time MRI sequence with magnetization transfer preparation (UTE-MT). METHODS: We recruited 23 amateur marathon runners (46 knees) in this prospective cohort study. MRI scans using UTE-MT and UTE-T2* sequences were performed pre-race, 2 days post-race, and 4 weeks post-race. UTE-MT ratio (UTE-MTR) and UTE-T2* were measured for knee cartilage (eight subregions) and meniscus (four subregions). The sequence reproducibility and inter-rater reliability were also investigated. RESULTS: Both the UTE-MTR and UTE-T2* measurements showed good reproducibility and inter-rater reliability. For most subregions of cartilage and meniscus, the UTE-MTR values decreased 2 days post-race and increased after 4 weeks of rest. Conversely, the UTE-T2* values increased 2 days post-race and decreased after 4 weeks. The UTE-MTR values in lateral tibial plateau, central medial femoral condyle, and medial tibial plateau showed a significant decrease at 2 days post-race compared to the other two time points (p < 0.05). By comparison, no significant UTE-T2* changes were found for any cartilage subregions. For meniscus, the UTE-MTR values in medial posterior horn and lateral posterior horn regions at 2 days post-race were significantly lower than those at pre-race and 4 weeks post-race (p < 0.05). By comparison, only the UTE-T2* values in medial posterior horn showed a significant difference. CONCLUSIONS: UTE-MTR is a promising method for the detection of dynamic changes in knee cartilage and meniscus after long-distance running. KEY POINTS: • Long-distance running causes changes in the knee cartilage and meniscus. • UTE-MT monitors dynamic changes of knee cartilage and meniscal non-invasively. • UTE-MT is superior to UTE-T2* in monitoring dynamic changes in knee cartilage and meniscus.

Dual-energy CT Virtual Non-Calcium Image is Beneficial for the Detection of Non-displaced Knee Fractures.

BACKGROUND: Early and accurate diagnosis is vital for avoiding the development of nondisplaced fractures to displaced fractures. Dual-energy CT (Computed Tomography) can detect bone marrow edema (BME), which may help to detect non-displaced fractures. AIM: To evaluate the value of DECT (Dual-Energy Computed Tomography) VNCa (Virtual noncalcium) images for improving diagnostic performance and confidence in acute non-displaced knee fractures. METHODS: 125 patients with clinical suspicion of knee fractures underwent both DECT and MR. Conventional linear-blended CT and VNCa images were obtained from DECT. First, five readers with varying levels of experience evaluated the presence of fractures on conventional linear-blended CT and graded their diagnostic confidence on a scale of 1 to 10. Then BME with VNCa images was evaluated and compared with MR. Finally, the VNCa images combined with conventional linear-blended CT images were used to reassess the presence of fractures and diagnostic confidence. Diagnostic performance and matched pair analyses were performed. RESULTS: 20 non-displaced knee fractures were detected. The consistency test of VNCa images and MR by five radiologists showed Kappa values are 0.76, 0.79, 0.81,0.85,and 0.90,respectively. The diagnostic performance of all readers was improved when using VNCa images combined with conventional linear-blended CT compared with that with conventional linear-blended CT alone. Diagnostic confidence was improved with combined conventional linear-blended CT and VNCa images (median score:8,8,9,9, and 10, respectively) compared with conventional linear-blended CT alone (median score:7,7,8,9, and 9). CONCLUSION: DECT VNCa images could improve the radiologists' diagnostic performance and confidence with varying levels of experience in the detection of non-displaced knee fractures.

Nomogram model based on radiomics signatures and age to assist in the diagnosis of knee osteoarthritis.

BACKGROUND: Knee osteoarthritis (KOA) is a common disease in the elderly. An effective method for accurate diagnosis could affect the management and prognosis of patients. OBJECTIVES: To develop a nomogram model based on X-ray imaging data and age, and to evaluate its effectiveness in the diagnosis of KOA. METHODS: A total of 4403 knee X-rays from 1174 patients (July 2017 to November 2018) were retrospectively analyzed. Radiomics features were extracted and selected from the X-ray image data to quantify the phenotypic characteristics of the lesion region. Feature selection was performed in three steps to enable the derivation of robust and effective radiomics signatures. Then, logistic regression (LR), support vector machine (SVM) AdaBoost, gradient boosting decision tree (GBDT), and multi-layer perceptron (MLP) was adopted to verify the performance of radiomics signatures. In addition, a nomogram model combining age with radiomics signatures was constructed. At last, receiver operating characteristic (ROC) curve, calibration and decision curves were used to evaluate the discriminative performance. RESULTS: The LR model has the best classification performance among the four radiomics models in testing cohort (LR AUC vs. SVM AUC: 0.843 vs. 0.818, DeLong test P = 0.0024; LR AUC vs. GBDT AUC: 0.843 vs. 0.821, P = 0.0028; LR AUC vs. MLP AUC: 0.843 vs. 0.822, P = 0.0019). The nomogram model achieved better predictive efficacy than the radiomics model in testing cohort compared to radiomics models although the statistical difference was not significant (Nomogram AUC vs. Radiomics AUC: 0.847 vs. 0.843, P = 0.06). The decision curve analysis revealed that the constructed nomogram had clinical usefulness. CONCLUSION: The nomogram model combining radiomics signatures with age has good performance for the accurate diagnosis of KOA and may help to improve clinical decision-making.

Assessment of Achilles Tendon Changes After Long-Distance Running Using Ultrashort Echo Time Magnetization Transfer MR Imaging.

BACKGROUND: Long-distance running is a common cause of Achilles tendinopathy. A reliable magnetic resonance imaging (MRI) technique to track early changes in the tendon caused by running could facilitate more effective interventions to combat progression. PURPOSE: To evaluate an ultrashort echo time sequence with magnetization transfer preparation (UTE-MT) in the detection of changes in Achilles tendons of amateur marathon runners before and after long-distance running. STUDY TYPE: Prospective. POPULATION: Thirty-two runners (19 enrolled for full marathons and 13 enrolled for half-marathons) and 5 healthy non-runners. FIELD STRENGTH/SEQUENCE: 3.0 T; UTE-MT and dual-echo UTE for T2* assessment (UTE-T2*). ASSESSMENT: MRI was performed 1-week pre-race, 2-days post-race, and 4-weeks post-race. UTE-MT ratio (UTE-MTR) and UTE-T2* of tendon were measured by two independent radiologists who were blinded to the scan time point and participant data. The Achilles tendon was divided into six regions of interest (ROIs) for data analysis, namely the insertion part (INS), middle part (MID), muscle-tendon junction (MTJ), tendon-bone insertion (TBI), tendon-muscle insertion (TMI), and whole tendon (bulk). STATISTICAL TESTS: Analysis of variance and Friedman's rank tests were used to evaluate changes in UTE-MTR and UTE-T2* between time points. Tukey test and Bonferroni method were used for further comparisons. P < 0.05 was considered significant. RESULTS: The UTE-MTR values of most tendon ROIs changed significantly between the measured time points, except for the INS region (P = 0.1977). Conversely, the UTE-T2* values only showed significant changes in the MID and TBI regions. Paired comparisons showed that the UTE-MTR decreases in the MTJ, MID, TMI, and bulk regions at 2-days post-race were significant compared to measures taken pre-race and 4-weeks post-race. For UTE-T2* measurements, significant differences were observed only for the MID region between pre-race and 2-days post-race (P = 0.0408, 95% CI: 0.0061, 0.1973), and for the TBI region between pre-race and 4-weeks post-race (P = 0.0473, 95% CI: 0.0013, 0.1766). DATA CONCLUSION: The UTE-MT sequence is able to detect biochemical changes in the Achilles tendon after long-distance running. LEVEL OF EVIDENCE: 2 TECHNICAL EFFICACY: Stage 1.

Molecular Imaging of Collagen Destruction of the Spine.

As the leading cause of disability worldwide, low back pain is commonly caused by biomechanical and catabolic disruptions to key structures of the spine, such as intervertebral discs and facet joints. To date, accurate, noninvasive detection of microdestruction within these tissues remains an elusive goal. Here, we report an in vivo imaging approach based on a collagen hybridizing peptide (CHP) that specifically targets disruption to the extracellular matrix architecture at the molecular scale─the denatured collagen molecules. Utilizing fluorescently labeled CHPs, live animal imaging, and light sheet fluorescence microscopy, we mapped collagen destruction in the lumbar spines in 3D, revealing that under normal conditions collagen destruction was localized to load-bearing anatomical structures including annulus fibrosus of the disc and the facet joints, where aging, tensile force (hindlimb suspension), and disc degeneration (needle puncture) escalated the CHP-binding in specific mouse models. We showed that targeting denatured collagen molecules allowed for an accurate, quantifiable interrogation of the structural integrity of these spinal matrixes with a greater sensitivity than anatomical imaging and histology. Finally, we demonstrated CHP's binding to degenerated human discs, suggesting exciting potentials for applying CHP for diagnosing, monitoring, and treating various spinal disorders, including intervertebral disc degeneration, facet joint osteoarthritis, and ankylosing spondylitis.

Early prediction of severity in coronavirus disease (COVID-19) using quantitative CT imaging.

PURPOSE: To evaluate whether the extent of COVID-19 pneumonia on CT scans using quantitative CT imaging obtained early in the illness can predict its future severity. METHODS: We conducted a retrospective single-center study on confirmed COVID-19 patients between January 18, 2020 and March 5, 2020. A quantitative AI algorithm was used to evaluate each patient's CT scan to determine the proportion of the lungs with pneumonia (VR) and the rate of change (RAR) in VR from scan to scan. Patients were classified as being in the severe or non-severe group based on their final symptoms. Penalized B-splines regression modeling was used to examine the relationship between mean VR and days from onset of symptoms in the two groups, with 95% and 99% confidence intervals. RESULTS: Median VR max was 18.6% (IQR 9.1-32.7%) in 21 patients in the severe group, significantly higher (P < 0.0001) than in the 53 patients in non-severe group (1.8% (IQR 0.4-5.7%)). RAR was increasing with a median RAR of 2.1% (IQR 0.4-5.5%) in severe and 0.4% (IQR 0.1-0.9%) in non-severe group, which was significantly different (P < 0.0001). Penalized B-spline analyses showed positive relationships between VR and days from onset of symptom. The 95% confidence limits of the predicted means for the two groups diverged 5 days after the onset of initial symptoms with a threshold of 11.9%. CONCLUSION: Five days after the initial onset of symptoms, CT could predict the patients who later developed severe symptoms with 95% confidence.

Comprehensive vector analysis for electro-optical, opto-electronic, and optical devices.

High-performance electro-optical (E-O), opto-electronic (O-E), and optical (O-O) devices are widely used in optical communications, microwave photonics, fiber sensors, and so on. Measurement of the amplitude and phase responses are essential for the development and fabrication of these devices. However, the previous methods can hardly characterize the E-O, O-E, and O-O devices with arbitrary responses. Here we propose a comprehensive vector analyzer based on optical asymmetrical double-sideband (ADSB) modulation to overcome this difficulty. The ADSB solves the problem of frequency aliasing and can extract information from both the +1st- and -1st-order sidebands. Thus, most devices in photonic applications, including phase modulators, can be characterized. In the experiment, a commercial photodetector, a phase modulator, and a sampled FBG are used as the O-E, E-O, and O-O devices under test, respectively. A frequency resolution of 2 MHz, an electrical sweeping range of 40 GHz, and an optical sweeping range of 80 GHz are achieved.

Auxiliary Diagnosis for COVID-19 with Deep Transfer Learning.

To assist physicians identify COVID-19 and its manifestations through the automatic COVID-19 recognition and classification in chest CT images with deep transfer learning. In this retrospective study, the used chest CT image dataset covered 422 subjects, including 72 confirmed COVID-19 subjects (260 studies, 30,171 images), 252 other pneumonia subjects (252 studies, 26,534 images) that contained 158 viral pneumonia subjects and 94 pulmonary tuberculosis subjects, and 98 normal subjects (98 studies, 29,838 images). In the experiment, subjects were split into training (70%), validation (15%) and testing (15%) sets. We utilized the convolutional blocks of ResNets pretrained on the public social image collections and modified the top fully connected layer to suit our task (the COVID-19 recognition). In addition, we tested the proposed method on a finegrained classification task; that is, the images of COVID-19 were further split into 3 main manifestations (ground-glass opacity with 12,924 images, consolidation with 7418 images and fibrotic streaks with 7338 images). Similarly, the data partitioning strategy of 70%-15%-15% was adopted. The best performance obtained by the pretrained ResNet50 model is 94.87% sensitivity, 88.46% specificity, 91.21% accuracy for COVID-19 versus all other groups, and an overall accuracy of 89.01% for the three-category classification in the testing set. Consistent performance was observed from the COVID-19 manifestation classification task on images basis, where the best overall accuracy of 94.08% and AUC of 0.993 were obtained by the pretrained ResNet18 (P < 0.05). All the proposed models have achieved much satisfying performance and were thus very promising in both the practical application and statistics. Transfer learning is worth for exploring to be applied in recognition and classification of COVID-19 on CT images with limited training data. It not only achieved higher sensitivity (COVID-19 vs the rest) but also took far less time than radiologists, which is expected to give the auxiliary diagnosis and reduce the workload for the radiologists.

Optical vector analyzer with time-domain analysis capability.

Time-domain analysis (TDA) is useful for measuring optical devices along with a link and for diagnosing a long device. In this Letter, an optical vector analyzer with TDA capability is proposed and experimentally demonstrated. The key to realizing TDA is a low-coherence optical carrier, which is achieved by modulating an electrical broadband signal on a continuous-wave light via acousto-optic modulation. Then, optical single-sideband modulation and vector balanced detection are used to measure the total frequency response of multiple devices under test (DUTs). Through an inverse Fourier transform, the obtained DUT impulses are distinguished in the time domain. Finally, time-domain gating and Fourier transform are applied to extract the frequency response of each DUT. An experiment is performed in which a fiber link comprising three DUTs and an H 13 C 14 N gas cell with a breakpoint inserted is characterized. The frequency setting resolution is 5 MHz, and a time-domain resolution of 30.84 ns is proved, which can reach 14.881 ns in theory.

Opportunistic osteoporosis screening in multi-detector CT images using deep convolutional neural networks.

OBJECTIVE: To explore the application of deep learning in patients with primary osteoporosis, and to develop a fully automatic method based on deep convolutional neural network (DCNN) for vertebral body segmentation and bone mineral density (BMD) calculation in CT images. MATERIALS AND METHODS: A total of 1449 patients were used for experiments and analysis in this retrospective study, who underwent spinal or abdominal CT scans for other indications between March 2018 and May 2020. All data was gathered from three different CT vendors. Among them, 586 cases were used for training, and other 863 cases were used for testing. A fully convolutional neural network, called U-Net, was employed for automated vertebral body segmentation. The manually sketched region of vertebral body was used as the ground truth for comparison. A convolutional neural network, called DenseNet-121, was applied for BMD calculation. The values post-processed by quantitative computed tomography (QCT) were identified as the standards for analysis. RESULTS: Based on the diversity of CT vendors, all testing cases were split into three testing cohorts: Test set 1 (n = 463), test set 2 (n = 200), and test set 3 (n = 200). Automated segmentation correlated well with manual segmentation regarding four lumbar vertebral bodies (L1-L4): the minimum average dice coefficients for three testing sets were 0.823, 0.786, and 0.782, respectively. For testing sets from different vendors, the average BMDs calculated by automated regression showed high correlation (r > 0.98) and agreement with those derived from QCT. CONCLUSIONS: A deep learning-based method could achieve fully automatic identification of osteoporosis, osteopenia, and normal bone mineral density in CT images. KEY POINTS: • Deep learning can perform accurate fully automated segmentation of lumbar vertebral body in CT images. • The average BMDs obtained by deep learning highly correlates with ones derived from QCT. • The deep learning-based method could be helpful for clinicians in opportunistic osteoporosis screening in spinal or abdominal CT scans.

Diagnostic accuracy of dual-energy CT virtual non-calcium images with different related contrast material values for the detection of bone marrow edema in knee.

PURPOSE: The purpose of this study was to evaluate the diagnostic accuracy of different related contrast material (Rel.CM) values in dual-energy computed tomography (DECT) virtual non-calcium (VNCa) images for the detection of bone marrow edema (BME) in knee. METHOD: This prospective study was approved by the institutional research ethics board, and written informed consent was obtained from all participants. Twenty-three patients (24 knees) who underwent dual-energy CT and MRI within three weeks from July 2018 to June 2019 with a definite history of trauma were enrolled. Each knee was divided into 12 regions. First, MR images served as the reference standard, Receiver operating characteristic (ROC) curve was used and diagnostic accuracy of VNCa images corresponding to different Rel.CM values (1.25, 1.35, 1.45, 1.55, 1.65, 1.75) were analyzed, aimed to select an optimal Rel.CM value of VNCa images for detecting BME. Then, CT values of the normal areas and BME areas were measured on the VNCa images corresponding to the optimal Rel.CM value for preliminary quantitative analysis. The rank-sum test was used to compare the differences of CT values between BME areas and normal bone marrow areas on the VNCa images. RESULTS: The 24 knees were divided into 288 areas. MR Imaging showed BME in 121 areas. The areas under the ROC curve with different Rel.CM values (1.25, 1.35, 1.45, 1.55, 1.65, and 1.75) were 0.633, 0.674, 0.882, 0.684, 0.651, and 0.649, respectively. On the VNCa images of Rel.CM = 1.45, the diagnostic accuracy was the highest (up to 89.2 %), the CT values of the BME area and the normal area were -67.9 (1.7∼-100.1) HU and -94.5 (-69.7∼-144.9) HU, respectively, with statistical significance (Z=-9.804, P < 0.05). CONCLUSIONS: The VNCa images with a Rel.CM value of 1.45 is optimal for the detection of BME in knee.

[Radiomics models based on non-enhanced MRI can differentiate chondrosarcoma from enchondroma].

OBJECTIVE: To develop and validate radiomics models based on non-enhanced magnetic resonance (MR) imaging for differentiating chondrosarcoma from enchondroma. METHODS: We retrospectively evaluated a total of 68 patients (including 27 with chondrosarcoma and 41 with enchondroma), who were randomly divided into training group (n=46) and validation group (n=22). Radiomics features were extracted from T1WI and T2WI-FS sequences of the whole tumor by two radiologists independently and selected by Low Variance, Univariate feature selection, and least absolute shrinkage and selection operator (LASSO). Radiomics models were constructed by multivariate logistic regression analysis based on the features from T1WI and T2WI-FS sequences. The receiver-operating characteristics (ROC) curve and intraclass correlation coefficient (ICC) analyses of the radiomics models and conventional MR imaging were performed to determine their diagnostic accuracy. RESULTS: The ICC value for interreader agreement of the radiomics features ranged from 0.779 to 0.923, which indicated good agreement. Ten and 11 features were selected from the T1WI and T2WI-FS sequences to construct radiomics models, respectively. The areas under the curve (AUCs) of T1WI and T2WI-FS models were 0.990 and 0.925 in training group and 0.915 and 0.855 in the validation group, respectively, showing no significant differences between the two sequence-based models (P>0.05). In all the cases, the AUCs of the two radiomics models based on T1WI and T2WI-FS sequences and conventional MR imaging were 0.955, 0.901 and 0.569, respectively, demonstrating a significantly higher diagnostic accuracy of the two sequence-based radiomics models than conventional MR imaging (P<0.01). CONCLUSIONS: The radiomics models based on T1WI and T2WI-FS non-enhanced MR imaging can be used for the differentiation of chondrosarcoma from enchondroma.

Clinical and CT features of the COVID-19 infection: comparison among four different age groups.

PURPOSE: To compare and analyze the clinical and CT features of coronavirus disease 2019 (COVID-19) among four different age groups. METHODS: 97 patients (45 males, 52 females, mean age, 66.2 ± 5.0) with chest CT examination and positive reverse transcriptase-polymerase chain reaction test (RT-PCR) from January 17, 2020 to February 21, 2020 were retrospectively studied. The patients were divided into four age groups (children [0-17 years], young adults [18-44 years], middle age [45-59 years], and senior [≥ 60 years]) according to their age after the diagnosis was made based on PCR test and clinical symptoms. RESULTS: Comorbidities such as hypertension, diabetes mellitus, and heart disease are more common in the senior group. Cluster onset (two or more confirmed cases in a small area) is more common in the children group and senior group. Older patients were found to have a higher incidence of the highest clinical classification (severe or critical) in these four groups. Senior patients have a higher incidence of large/multiple ground-glass opacity (GGO). Child patients are mostly negative for chest CT or with involvement of only one lobe of the lung; while in older patients, there was a higher incidence of involvement of four or five lung lobes. The frequency of lobe involvement was also found to have significant differences in the four age groups. CONCLUSION: The clinical and imaging features of patients in different age groups were found to be significantly different. A better understanding of the age differences in comorbidities, cluster onset, highest clinical classification, large/multiple GGO, numbers of lobes affected, and frequency of lobe involvement can be useful in the diagnosis of COVID-19 patients of different ages.

Chest computed tomography in children with COVID-19 respiratory infection.

BACKGROUND: Infection with COVID-19 is currently rare in children. OBJECTIVE: To describe chest CT findings in children with COVID-19. MATERIALS AND METHODS: We studied children at a large tertiary-care hospital in China, during the period from 28 January 2019 to 8 February 2020, who had positive reverse transcriptase polymerase chain reaction (RT-PCR) for COVID-19. We recorded findings at any chest CT performed in the included children, along with core clinical observations. RESULTS: We included five children from 10 months to 6 years of age (mean 3.4 years). All had had at least one CT scan after admission. Three of these five had CT abnormality on the first CT scan (at 2 days, 4 days and 9 days, respectively, after onset of symptoms) in the form of patchy ground-glass opacities; all normalised during treatment. CONCLUSION: Compared to reports in adults, we found similar but more modest lung abnormalities at CT in our small paediatric cohort.