CT-based deep learning radiomics biomarker for programmed cell death ligand 1 expression in non-small cell lung cancer.

BACKGROUND: Programmed cell death ligand 1 (PD-L1), as a reliable predictive biomarker, plays an important role in guiding immunotherapy of lung cancer. To investigate the value of CT-based deep learning radiomics signature to predict PD-L1 expression in non-small cell lung cancers(NSCLCs). METHODS: 259 consecutive patients with pathological confirmed NSCLCs were retrospectively collected and divided into the training cohort and validation cohort according to the chronological order. The univariate and multivariate analyses were used to build the clinical model. Radiomics and deep learning features were extracted from preoperative non-contrast CT images. After feature selection, Radiomics score (Rad-score) and deep learning radiomics score (DLR-score) were calculated through a linear combination of the selected features and their coefficients. Predictive performance for PD-L1 expression was evaluated via the area under the curve (AUC) of receiver operating characteristic, the calibration curves, and the decision curve analysis. RESULTS: The clinical model based on Cytokeratin 19 fragment and lobulated shape obtained an AUC of 0.767(95% CI: 0.673-0.860) in the training cohort and 0.604 (95% CI:0.477-0.731) in the validation cohort. 11 radiomics features and 15 deep learning features were selected by LASSO regression. AUCs of the Rad-score were 0.849 (95%CI: 0.783-0.914) and 0.717 (95%CI: 0.607-0.826) in the training cohort and validation cohort, respectively. AUCs of DLR-score were 0.938 (95%CI: 0.899-0.977) and 0.818(95%CI:0.727-0.910) in the training cohort and validation cohort, respectively. AUCs of the DLR-score were significantly higher than those of the Rad-score and the clinical model. CONCLUSION: The CT-based deep learning radiomics signature could achieve clinically acceptable predictive performance for PD-L1 expression, which showed potential to be a surrogate imaging biomarker or a complement of immunohistochemistry assessment.

CT-based radiomics signature to predict CD8+ tumor infiltrating lymphocytes in non-small-cell lung cancer.

BACKGROUND: An abundance of CD8+ tumor infiltrating lymphocytes (TILs) in the center of solid tumors is a reliable predictive biomarker for patients eligible for immunotherapy. PURPOSE: To develop a computed tomography (CT)-based radiomics signature for a preoperative prediction of an abundance of CD8+ TILs in non-small-cell lung cancer (NSCLC). MATERIAL AND METHODS: In this retrospective study, 117 consecutive patients with pathologically confirmed NSCLC were included and randomly divided into training (n = 77) and test sets (n = 40). A total of 107 radiomics features were extracted from the three-dimensional volumes of interest of each patient. Least absolute shrinkage and selection operator (LASSO) regression was used to select the strongest features for abundance of CD8+ TILs in NSCLC, and the radiomics score was constructed through a linear combination of these selected features. Receiver operating characteristic (ROC) curve analysis was used to evaluate the predictive performance of the radiomics score. RESULTS: The radiomics score was associated with an abundance of CD8+ TILs in NSCLC, which achieved an area under the curve (AUC) of 0.83 (95% CI=0.73-0.92) and 0.68 (95% CI=0.54-0.87) in the training and test sets, respectively. The difference was not statistically significant (P = 0.20). The tumors with high CD8+ TILs tended to have heterogeneous dependences (high value of Dependence Non-Uniformity Normalized) and complicated texture (high value of Informational Measure of Correlation 1). CONCLUSION: CT-based radiomics features have the ability to predict CD8+ TILs expression levels of an abundance of CD8+ TILs in NSCLC, which was shown to be a potential imaging biomarker for stratifying patients who may benefit from immunotherapy.

Multiparametric MRI-Based Radiomics Nomogram for Predicting Lymphovascular Space Invasion in Endometrial Carcinoma.

BACKGROUND: Lymphovascular space invasion (LVSI) of endometrial carcinoma (EMC) is one of the important prognostic factors, which is not usually visible subjectively. Therefore, clinicians lack imaging-based evidence about LVSI for preoperative treatment decision-making. PURPOSE: To develop a multiparametric MRI (mpMRI)-based radiomics nomogram for predicting LVSI in EMC and provide decision-making support to clinicians. STUDY TYPE: Retrospective. POPULATION: In all, 144 patients with histologically confirmed EMC, 101 patients in a training cohort, and 43 patients in a test cohort. FIELD STRENGTH/SEQUENCE: T2 WI, contrast enhanced-T1 WI, and diffusion-weighted imaging (DWI) at 3.0T MRI. ASSESSMENT: Tumors were independently segmented images by two radiologists. Two pathologists reviewed the tissue specimens of the tumors to identify the existence of LVSI in consensus. STATISTICAL TESTS: The intraclass correlation coefficient was used to test the reliability and least absolute shrinkage and selection operator (LASSO) regression for features selection and then developed a radiomics signature named Rad-score. A nomogram was developed in the training cohort. The diagnostic performance of the nomogram model was assessed by area under the curve (AUC) of the receiver operator characteristic (ROC) in the training and test cohort, respectively. RESULTS: LVSI was identified in 32 patients (22.2%). Older age and high grade were correlated with LVSI at univariate analysis. There were five radiomics features that were identified as independent risk factors for LVSI by LASSO regression. Based on age, grade, and Rad-score, the AUC values of the nomogram model to predict LVSI in the training and test cohort were 0.820 (95% confidence interval [CI]: 0.725, 0.916; sensitivity: 82.6%, specificity: 72.9%), 0.807 (95% CI: 0.673, 0.941; sensitivity: 77.8%, specificity: 78.6%), respectively. DATA CONCLUSION: The radiomic-based machine-learning model using a nomogram algorithm achieved high diagnostic performance for predicting LVSI of EMC preoperatively, which could enhance risk stratification and provide support for therapeutic decision-making. LEVEL OF EVIDENCE: 2. TECHNICAL EFFICACY STAGE: 3. J. Magn. Reson. Imaging 2020;52:1257-1262.

CT-based radiomics and machine learning to predict spread through air space in lung adenocarcinoma.

PURPOSE: Spread through air space (STAS) is a novel invasive pattern of lung adenocarcinoma and is also a risk factor for recurrence and worse prognosis of lung adenocarcinoma. The aims of this study are to develop and validate a computed tomography (CT)­based radiomics model for preoperative prediction of STAS in lung adenocarcinoma. METHODS AND MATERIALS: This retrospective study was approved by an institutional review board and included 462 (mean age, 58.06 years) patients with pathologically confirmed lung adenocarcinoma. STAS was identified in 90 patients (19.5%). Two experienced radiologists segmented and extracted radiomics features on preoperative thin-slice CT images with radiomics extension independently. Intraclass correlation coefficients (ICC) and Pearson's correlation were used to rule out those low reliable (ICC < 0.75) and redundant (r > 0.9) features. Univariate logistic regression was applied to select radiomics features which were associated with STAS. A radiomics-based machine learning predictive model using a random forest (RF) was developed and calibrated with fivefold cross-validation. The diagnostic performance of the model was measured by the area under the curve (AUC) of receiver operating characteristic (ROC). RESULTS: With univariate analysis, 12 radiomics features and age were found to be associated with STAS significantly. The RF model achieved an AUC of 0.754 (a sensitivity of 0.880 and a specificity of 0.588) for predicting STAS. CONCLUSION: CT-based radiomics model can preoperatively predict STAS in lung adenocarcinoma with good diagnosis performance. KEY POINTS: • CT-based radiomics and machine learning model can predict spread through air space (STAS) in lung adenocarcinoma with high accuracy. • The random forest (RF) model achieved an AUC of 0.754 (a sensitivity of 0.880 and a specificity of 0.588) for predicting STAS.

[Logistic regression analysis on risk factors of neonates T1WI hyperintensity at globus pallidus and subthalamic nucleus].

OBJECTIVE: To determine risk factors of T1WI high signal intensity at globus pallidus and subthalamic nucleus (GP and STN) of neonates. METHODS: Brain MR images of 186 neonates with intact clinical files were retrospectively reviewed to identify whether there were T1WI high signal intensity at GP and STN. Among them, 15 neonates received followed-up MR imaging in 1-5 months after first MR examination. Statistic comparison of clinical features between neonates with and without T1WI high signal intensity at GP and STN were performed using univariate analyses. Then, multiple Logistic regression analysis was used to identify the risk factors of T1WI high signal intensity at GP and STN among those factors which were statistical significant at univariate analyses. ROC curve was employed to determine the cut-off value of the risk factors. RESULTS: T1WI high signal intensity at GP and STN was identified in 74.2% neonates (138/186). At univariate analyses, the following factors were found with statistical difference between neonates with and without T1WI hyperintensity at GP and STN: transcutaneous bilirubin (TCB), (132±62) vs (91±55) µmol/L (t=3.935, P<0.01); gestational age, 36.0±2.6 weeks vs 34.8±3.4 weeks (t=2.263, P=0.027); age, 9±5 days vs 19±7 days (t=8.992, P<0.01). Multiple Logistic regression analysis revealed a significant negative contribution of age to T1WI hyperintensity at GP and STN (OR=0.795, 95% CI 0.739-0.856, P<0.01). T1WI hyperintensity at GP and STN was seen in the fifteen neonates with followed-up MRI at the first MR imaging. It was disappeared in the followed-up MR imaging and all the neonates developed well without any remarkable abnormalities at physical examinations. ROC curve determined the cut-off value of age was 20 days (The incidence of T1WI high signal intensity at GP and STN was 16.0% in neonates>20 days and 83.2% in those ≤20 days, respectively χ2=51.084, P<0.01). CONCLUSIONS: T1WI high signal intensity at GP and STN of neonates' brain MR imaging is common, and is related to their age at examination. It should be regarded as a transient phenomenon instead of a sign of kernicterus and asphyxia.

Prevalence and risk factors of lumbar spondylolisthesis in elderly Chinese men and women.

OBJECTIVE: A screening survey for osteoporotic fractures in men and women in Hong Kong represents the first large-scale prospective population-based study on bone health in elderly (≥65 years) Chinese men and women. This study aims to identify the prevalence and potential risk factors of lumbar spondylolisthesis in these subjects. METHODS: The lateral lumbar radiographs of 1,994 male and 1,996 female patients were analysed using the Meyerding classification. RESULTS: Amongst the men, 380 (19.1%) had at least one spondylolisthesis and 43 (11.3%) had slips at two or more levels; 283 had anterolisthesis, 85 had retrolisthesis, whereas 12 subjects had both anterolisthesis and retrolisthesis. Amongst the women, 499 (25.0%) had at least one spondylolisthesis and 69 (13.8%) had slips at two or more levels; 459 had anterolisthesis, 34 had retrolisthesis, whereas 6 subjects had both anterolisthesis and retrolisthesis. Advanced age, short height, higher body mass index (BMI), higher bone mineral density (BMD) and degenerative arthritis are associated with spondylolisthesis. Lower Physical Activity Scale for the Elderly (PASE) score was associated with spondylolisthesis in men; higher body weight, angina and lower grip strength were associated with spondylolisthesis in women. CONCLUSION: The male/female ratio of lumbar spondylolisthesis prevalence was 1:1.3 in elderly Chinese. Men are more likely to have retrolisthesis. KEY POINTS: • The prevalence of spondylolisthesis is 19.1% in elderly Chinese men. • The prevalence of spondylolisthesis is 25.0% in elderly Chinese women. • Men are more likely to have retrolisthesis. • Anterolisthesis is most commonly seen at the L4/L5 level. • Retrolisthesis is most commonly seen at the L3/L4 level.

CT-Based Super-Resolution Deep Learning Models with Attention Mechanisms for Predicting Spread Through Air Spaces of Solid or Part-Solid Lung Adenocarcinoma.

RATIONALE AND OBJECTIVES: Spread through air space (STAS) is a novel invasive pattern of lung adenocarcinoma (LUAD), and preoperative knowledge of STAS status is helpful in choosing an appropriate surgical approach. MATERIALS AND METHODS: This retrospective study collected and analyzed 602 patients diagnosed with LUAD from two medical centers: center 1 was randomly partitioned into training (n = 358) and validation cohorts (n = 154) at a 7:3 ratio; and center 2 was the external test cohort (n = 90). Super resolution was performed on all images to acquire high-resolution images, which were used to train the SE-ResNet50 model, before creating an equivalent parameter ResNet50 model. Disparities were compared between the two models using receiver operating characteristic curves, area under the curve, accuracy, precision, sensitivity, and specificity. RESULTS: In this study, 512 and 90 patients with LUAD were enrolled from centers 1 and 2, respectively. The curve values of the SE-ResNet50 and ResNet50 models were compared for training, validation, and test cohorts, resulting in values of 0.933 vs 0.909, 0.783 vs 0.728, and 0.806 vs 0.695, respectively. In the external test cohort, the accuracy of the SE-ResNet50 model demonstrated a 10% improvement over the ResNet50 model (82.2% vs 72.2%). CONCLUSION: The SE-ResNet50 model based on computed tomography super-resolution has great potential for predicting STAS status in patients with solid or partially solid LUAD, with superior predictive performance compared to traditional deep learning models.

[Relationship between dynamic contrast-enhanced and perfusion magnetic resonance imaging and T-staging of nasopharyngeal carcinoma].

OBJECTIVE: To perform the dynamic contrast-enhanced and perfusion magnetic resonance imaging (MRI) of nasopharyngeal carcinoma (NPC) and analyze the correlation with T-staging. METHODS: A total of 46 naïve NPC patients underwent MRI. The parameters of dynamic contrast-enhanced and perfusion MRI included time to peak (TTP), Slopemax and area under the curve (AUC). RESULTS: The increasing period of signal intensity-time curve of all cases was steep. And the perfusion image of AUC could reflect the blood perfusion of tumor tissue. Parameters (TTP/Slopemax/AUC) in different T-staging were T1-staging (60.45/10.59/20 619.56), T2-staging (58.12/12.47/23 037.23), T3-staging (70.61/15.06/26 507.23) and T4-staging (41.72/19.87/30 092.27). Their statistical results were r = -0.247, P > 0.05 and r = 0.859, P < 0.050 and r = 0.963, P < 0.05 respectively. And statistical significance existed in Slopemax, AUC with T-staging. CONCLUSION: Dynamic contrast-enhanced and perfusion MRI can reflect angiogenesis of NPC. And there is a positive correlation between the parameters of dynamic contrast-enhanced and perfusion MRI (Slopemax, AUC) and T-staging.

CT-based radiomic phenotypes of lung adenocarcinoma: a preliminary comparative analysis with targeted next-generation sequencing.

Objectives: This study aimed to explore the relationship between computed tomography (CT)-based radiomic phenotypes and genomic profiles, including expression of programmed cell death-ligand 1 (PD-L1) and the 10 major genes, such as epidermal growth factor receptor (EGFR), tumor protein 53 (TP53), and Kirsten rat sarcoma viral oncogene (KRAS), in patients with lung adenocarcinoma (LUAD). Methods: In total, 288 consecutive patients with pathologically confirmed LUAD were enrolled in this retrospective study. Radiomic features were extracted from preoperative CT images, and targeted genomic data were profiled through next-generation sequencing. PD-L1 expression was assessed by immunohistochemistry staining (chi-square test or Fisher's exact test for categorical data and the Kruskal-Wallis test for continuous data). A total of 1,013 radiomic features were obtained from each patient's CT images. Consensus clustering was used to cluster patients on the basis of radiomic features. Results: The 288 patients were classified according to consensus clustering into four radiomic phenotypes: Cluster 1 (n = 11) involving mainly large solid masses with a maximum diameter of 5.1 ± 2.0 cm; Clusters 2 and 3 involving mainly part-solid and solid masses with maximum diameters of 2.1 ± 1.4 cm and 2.1 ± 0.9 cm, respectively; and Cluster 4 involving mostly small ground-glass opacity lesions with a maximum diameter of 1.0 ± 0.9 cm. Differences in maximum diameter, PD-L1 expression, and TP53, EGFR, BRAF, ROS1, and ERBB2 mutations among the four clusters were statistically significant. Regarding targeted therapy and immunotherapy, EGFR mutations were highest in Cluster 2 (73.1%); PD-L1 expression was highest in Cluster 1 (45.5%). Conclusion: Our findings provide evidence that CT-based radiomic phenotypes could non-invasively identify LUADs with different molecular characteristics, showing the potential to provide personalized treatment decision-making support for LUAD patients.

Non-contrast cine cardiovascular magnetic resonance-based radiomics nomogram for predicting microvascular obstruction after reperfusion in ST-segment elevation myocardial infarction.

Purpose: This study aimed to develop and validate a cine cardiovascular magnetic resonance (CMR)-based radiomics nomogram model for predicting microvascular obstruction (MVO) following reperfusion in patients with ST-segment elevation myocardial infarction (STEMI). Methods: In total, 167 consecutive STEMI patients were retrospectively enrolled. The patients were randomly divided into training and validation cohorts with a ratio of 7:3. All patients were diagnosed with myocardial infarction with or without MVO based on late gadolinium enhancement imaging. Radiomics features were extracted from the cine CMR end-diastolic volume phase of the entire left ventricular myocardium (3D volume). The least absolute shrinkage and selection operator (LASSO) regression was employed to select the features that were most relevant to the MVO; these features were then used to calculate the radiomics score (Rad-score). A combined model was developed based on independent risk factors screened using multivariate regression analysis and visualized using a nomogram. Performance was assessed using receiver operating characteristic curve, calibration curve, and decision curve analysis (DCA). Results: The univariate analysis of clinical features demonstrated that only cardiac troponin I (cTNI) was significantly associated with MVO. LASSO regression revealed that 12 radiomics features were strongly associated with MVO. Multivariate regression analysis indicated that cTNI and Rad-score were independent risk factors for MVO. The nomogram based on these two features achieved an area under the curve of 0.86 and 0.78 in the training and validation cohorts, respectively. Calibration curves and DCA indicated the clinical feasibility and utility of the nomogram. Conclusions: A CMR-based radiomics nomogram offers an effective means of predicting MVO without contrast agents and radiation, which could facilitate risk stratification of patients with STEMI after PCI for reperfusion.

Habitat-based radiomics enhances the ability to predict lymphovascular space invasion in cervical cancer: a multi-center study.

Introduction: Lymphovascular space invasion (LVSI) is associated with lymph node metastasis and poor prognosis in cervical cancer. In this study, we investigated the potential of radiomics, derived from magnetic resonance (MR) images using habitat analysis, as a non-invasive surrogate biomarker for predicting LVSI in cervical cancer. Methods: This retrospective study included 300 patients with cervical cancer who underwent surgical treatment at two centres (centre 1 = 198 and centre 2 = 102). Using the k-means clustering method, contrast-enhanced T1-weighted imaging (CE-T1WI) images were segmented based on voxel and entropy values, creating sub-regions within the volume ofinterest. Radiomics features were extracted from these sub-regions. Pearson correlation coefficient and least absolute shrinkage and selection operator LASSO) regression methods were used to select features associated with LVSI in cervical cancer. Support vector machine (SVM) model was developed based on the radiomics features extracted from each sub-region in the training cohort. Results: The voxels and entropy values of the CE-T1WI images were clustered into three sub-regions. In the training cohort, the AUCs of the SVM models based on radiomics features derived from the whole tumour, habitat 1, habitat 2, and habitat 3 models were 0.805 (95% confidence interval [CI]: 0.745-0.864), 0.873(95% CI: 0.824-0.922), 0.869 (95% CI: 0.821-0.917), and 0.870 (95% CI: 0.821-0.920), respectively. Compared with whole tumour model, the predictive performances of habitat 3 model was the highest in the external test cohort (0.780 [95% CI: 0.692-0.869]). Conclusions: The radiomics model based on the tumour sub-regional habitat demonstrated superior predictive performance for an LVSI in cervical cancer than that of radiomics model derived from the whole tumour.

Association between cine CMR-based radiomics signature and microvascular obstruction in patients with ST-segment elevation myocardial infarction.

Background: In the process of percutaneous coronary intervention (PCI), patients with ST-segment elevation myocardial infarction (STEMI) may receive large doses of the iodine contrast agent. Some adverse events may be aroused if the patients receive the gadolinium agents. We investigate the association between cine cardiac magnetic resonance (CMR)-based radiomics signature and microvascular obstruction (MVO) in patients with STEMI. Methods: A total of 116 STEMI patients who received continuous CMR within 6 days after PCI were retrospectively included in this study. According to the late gadolinium enhancement (LGE) of CMR, the myocardial infarction (MI) was divided into with and without MVO. Radiomic features were extracted from cine CMR images and the least absolute shrinkage and selectionator operator (LASSO) algorithm was used for features selection and radiomic signatures construction. Binary logistic regression was used to assess association between radiomic signatures and MVO with adjusted for baseline clinical characteristics. Results: Of 116 patients with STEMI, MI with MVO was found in 50 patients and MI without MVO was found in 66 patients. LASSO regression selected five radiomics features for radiomics signature construction. Logistic regression revealed that radiomics score, high sensitivity C-reactive protein (hs-CRP) and creatine phosphokinases (CPK) were independent risk factors for MVO with odds ratio (OR) of 4.41 (95% CI: 2.26-9.93), 1.018 (95% CI: 1.006-1.034) and 1.0007 (95% CI: 1.0004-1.0012), respectively. Area under curve (AUC) of receiver operating characteristic (ROC) of radiomics score to predict MVO was 0.75 (95% CI: 0.68-0.85). Conclusions: Cine CMR-based radiomics signature was an independent predictive factor of MVO in patients with STEMI, which showed the potential of this contrast free radiomics signature to be an imaging biomarker for MVO.

Efficacy Evaluation of Neoadjuvant Chemotherapy in Breast Cancer by MRI.

Breast cancer is a highly harmful malignancy, which often causes great distress to patients and seriously affects their physical and mental health. Breast cancer causes patients to experience decreased appetite, decreased eating, and indigestion, which in turn leads to malnutrition, body wasting, resistance, immune compromise, progressive anemia, cachexia, and, as a result, severe secondary infections. To investigate the efficacy evaluation of neoadjuvant chemotherapy in breast cancer by MRI, forty-eight subjects treated at the hospital from June 2014 to August 2019 were recruited. After the neoadjuvant chemotherapy, the patients were divided into two groups based on the results of histopathological examination, namely, the ineffective group (n = 14) and the effective group (n = 34). Changes in MRI indicators were compared between the two groups before and after the neoadjuvant chemotherapy. The maximum diameter of lesions decreased significantly after the neoadjuvant chemotherapy than before. The apparent diffusion coefficient (ADC) increased considerably, and the time-intensity curve (TIC) showed a transition from type III to type II/I and from type II to type I. MRI can indicate the maximum diameter of the breast cancer lesion, ADC, and TIC type. Therefore, it can be used to evaluate the efficacy of neoadjuvant chemotherapy for breast cancer and be widely applied in clinical practice.

Development and validation of a CT-based nomogram to predict spread through air space (STAS) in peripheral stage IA lung adenocarcinoma.

INTRODUCTION: To develop and validate a simple-to-use nomogram based on preoperative CT to predict spread through air space (STAS) status of stage IA lung adenocarcinoma (ADC). METHODS: In this retrospective study, 434 patients with pathological proven periphery stage IA lung adenocarcinoma were included, which consisted of 349 patients from center I for training group and 85 patients from Center II for test group. STAS was identified in 53 patients (40 patient in the training group and 13 patients in the test group). On the basis of preoperative CT images, 19 morphological characteristics were analyzed. Univariable analysis was used to explore the association between clinical and CT characteristics and STAS status in the training group (P < 0.002). Independent risk factors for STAS were identified using multivariable logistic regression analysis and then used to build a nomogram for preoperative predicting STAS status. RESULTS: Type of nodules, diameter of solid component, lobulation and percentage of the solid component (PSC) were associated with STAS status of peripheral stage IA lung ADCs statistical significantly. Multivariate logistics regression analysis revealed that PSC and lobulation were independent risk factors for STAS. The nomogram based on these factors achieved good predictive performance for STAS with a C-index of 0.803 in the training group and a well-fitted calibration curve. Using a cut-off value which was obtained from Youden index of the receiver operating characteristic (ROC) curve, a diagnosis accuracy of 70.6% was obtained in the test group with sensitivity, specificity, positive prediction value (PPV) and negative prediction value (NPV) of 92.3%, 66.7%, 33.3% and 98.0%, respectively. CONCLUSION: The nomogram based on preoperative CT images could achieve good predictive performance for STAS status of lung adenocarcinomas. This simple-to-used model can facilitate surgeons for a rational operation pattern choice at bedside.

A software program for automated compressive vertebral fracture detection on elderly women's lateral chest radiograph: Ofeye 1.0.

Background: Because osteoporotic vertebral fracture (OVF) on chest radiographs is commonly missed in radiological reports, we aimed to develop a software program which offers automated detection of compressive vertebral fracture (CVF) on lateral chest radiographs, and which emphasizes CVF detection specificity with a low false positivity rate. Methods: For model training, we retrieved 3,991 spine radiograph cases and 1,979 chest radiograph cases from 16 sources, with among them in total 1,404 cases had OVF. For model testing, we retrieved 542 chest radiograph cases and 162 spine radiograph cases from four independent clinics, with among them 215 cases had OVF. All cases were female subjects, and except for 31 training data cases which were spine trauma cases, all the remaining cases were post-menopausal women. Image data included DICOM (Digital Imaging and Communications in Medicine) format, hard film scanned PNG (Portable Network Graphics) format, DICOM exported PNG format, and PACS (Picture Archiving and Communication System) downloaded resolution reduced DICOM format. OVF classification included: minimal and mild grades with <20% or ≥20-25% vertebral height loss respectively, moderate grade with ≥25-40% vertebral height loss, severe grade with ≥40%-2/3 vertebral height loss, and collapsed grade with ≥2/3 vertebral height loss. The CVF detection base model was mainly composed of convolution layers that include convolution kernels of different sizes, pooling layers, up-sampling layers, feature merging layers, and residual modules. When the model loss function could not be further decreased with additional training, the model was considered to be optimal and termed 'base-model 1.0'. A user-friendly interface was also developed, with the synthesized software termed 'Ofeye 1.0'. Results: Counting cases and with minimal and mild OVFs included, base-model 1.0 demonstrated a specificity of 97.1%, a sensitivity of 86%, and an accuracy of 93.9% for the 704 testing cases. In total, 33 OVFs in 30 cases had a false negative reading, which constituted a false negative rate of 14.0% (30/215) by counting all OVF cases. Eighteen OVFs in 15 cases had OVFs of ≥ moderate grades missed, which constituted a false negative rate of 7.0% (15/215, i.e., sensitivity 93%) if only counting cases with ≥ moderate grade OVFs missed. False positive reading was recorded in 13 vertebrae in 13 cases (one vertebra in each case), which constituted a false positivity rate of 2.7% (13/489). These vertebrae with false positivity labeling could be readily differentiated from a true OVF by a human reader. The software Ofeye 1.0 allows 'batch processing', for example, 100 radiographs can be processed in a single operation. This software can be integrated into hospital PACS, or installed in a standalone personal computer. Conclusions: A user-friendly software program was developed for CVF detection on elderly women's lateral chest radiographs. It has an overall low false positivity rate, and for moderate and severe CVFs an acceptably low false negativity rate. The integration of this software into radiological practice is expected to improve osteoporosis management for elderly women.

Primary malignant fibrous histiocytoma of the liver: CT findings in five histopathological proven patients.

BACKGROUND: To describe imaging features of primary hepatic malignant fibrous histiocytoma (MFH), especially its dynamic enhancement pattern. MATERIALS: The CT manifestations of five patients with histopathological proven primary hepatic MFH retrospectively reviewed. RESULTS: Six hepatic masses were detected in the five patients at CT examinations. All tumors were hypodense on plain CT scanning related to adjacent hepatic parenchyma. Five lesions were heterogeneous with predominantly cystic or necrosis areas and one lesion was homogeneous solid. After administration of contrast material, solid components of the four tumors with cystic or necrotic areas and the solid lesion showed a "fast rushing in and washing out" enhancement pattern with marked enhancement at the arterial dominated phase and fast decreasing at the portal vein dominated phase and the delay phase. The remaining one tumor showed slight enhancement of its solid components. Portal lymph node enlargement was observed in one patient and the inferior vena cava was invaded in one patient, respectively. CONCLUSION: The CT manifestations of primary hepatic MFH are various and nonspecific. A "fast rushing in and washing out" enhancement pattern of MFH in our series has not been reported before. These manifestations could help radiologists to include primary hepatic MFH in the differential diagnosis list.

Radiological findings of renal leiomyomas: a report of five surgically confirmed cases.

BACKGROUND: To describe radiological features of renal leiomyomas. METHODS: We retrospectively reviewed radiological findings of surgically confirmed renal leiomyomas in five patients. Three patients only underwent CT examinations, while two patients received both CT and MR examinations. RESULTS: Six renal leiomyomas were found in the five patients, which located in the right kidney. Four tumors were at the periphery of the kidney, one tumor was in the renal parenchyma, and one tumor located in the renal pelvis. At plain CT scans, five lesions were homogeneous dense with three hyperdense and two isodense, while one lesion was heterogeneous dense. At MR imaging, one lesion was heterogeneous signal and two lesions were homogeneous signal. After i.v. contrast materials, all lesions demonstrated homogenous or heterogeneous enhancement which were lower than the renal cortex at the corticomedullary phase. A trend of continuous and homogeneous enhancement was observed in dynamic enhancement on CT and MR imaging. CONCLUSION: Renal leiomyomas have some characteristic radiological findings, such as homogeneous density or signal peripheral mass with well-defined margins, less heterogeneous or homogeneous enhancement than adjacent renal cortex at the corticomedullary phase, a trend of continuous and homogeneous on later phase of dynamic scanning.

Interpretation of osteoporotic vertebral deformity on frontal view radiographs of the chest and abdomen: a pictorial review.

Osteoporotic vertebral fracture (OVF) has high prevalence in the elderly population. It affects at least one-fourth of all postmenopausal women and is commonly seen among women approximately one decade after menopause. A vertebral fracture, after minor trauma, is a hallmark of osteoporosis. Many fractures and associated complications, including secondary fractures and mortality, can be prevented by routine osteoporosis screening in older people and timely treatment initiation in at-risk individuals. Depending on the technical condition of the radiographs, a substantial portion of moderate to severe grades OVFs in mid-thoracic and lower thoracic spine as well as lumbar spine can be detected on a frontal view digital radiograph of the chest or abdomen. Radiologists should pay attention to the potential existence of an OVF while reading chest and abdominal radiographs of elderly female subjects. In this pictorial review, we describe our experience in evaluating the normal shaped and deformed vertebrae on chest and abdominal radiographs.

A CT-based logistic regression model to predict spread through air space in lung adenocarcinoma.

BACKGROUND: Spread through air space (STAS) is a novel invasive pattern of lung adenocarcinoma and is also a risk factor for recurrence and worse prognosis of lung adenocarcinoma. This study aimed to develop and validate a computed tomography (CT)-based logistic regression model to predict STAS in lung adenocarcinoma. METHODS: This retrospective study was approved by the institutional review board of two centers and included 578 patients (462 from center I and 116 from center II) with pathologically confirmed lung adenocarcinoma. STAS was identified from 90 center I patients (19.5%) and 28 center II patients (24.1%) from. The maximum diameter, nodule area, and area of solid components in part-solid nodules were measured. Twenty-one semantic characteristics were assessed. Univariate analysis was used to select CT characteristics, which were associated with STAS in the patient cohort of center I. Multivariable logistic regression was used to develop a CT characteristics-based model on those variables with statistical significance. The model was validated in the validation cohort and then tested in the external test cohort (patients from center II). The diagnostic performance of the model was measured by area under the curve (AUC) of receiver operating characteristic (ROC). RESULTS: At univariate analysis, age and 11 CT characteristics, including the maximum diameter of the tumor, the maximum area of the tumor, the area and ratio of the solid component, nodule type, pleural thickening, pleural retraction, mediastinal lymph node enlargement, vascular cluster sign, and lobulation, specula were found to be significantly associated with STAS. The optimal logistic regression model included age, maximum diameter and ratio of solid component with odds ratio (OR) value of 0.967 (95% CI: 0.944-0.988), 1.027 (95% CI: 1.008-1.046) and 5.14 (95% CI: 2.180-13.321), respectively. This model achieved an AUC of 0.801 (95% CI: 0.709-0.892) and 0.692 (95% CI: 0.518-0.866) in the validation cohort and the external test cohort, respectively. The difference was not statistically significant (P=0.280). CONCLUSIONS: CT-based logistic regression machine learning model could preoperatively predict STAS in lung adenocarcinoma with excellent diagnosis performance, which could be supplementary to routine CT interpretation.