A radiograph-based deep learning model improves radiologists' performance for classification of histological types of primary bone tumors: A multicenter study.

PURPOSE: To develop a deep learning (DL) model for classifying histological types of primary bone tumors (PBTs) using radiographs and evaluate its clinical utility in assisting radiologists. METHODS: This retrospective study included 878 patients with pathologically confirmed PBTs from two centers (638, 77, 80, and 83 for the training, validation, internal test, and external test sets, respectively). We classified PBTs into five categories by histological types: chondrogenic tumors, osteogenic tumors, osteoclastic giant cell-rich tumors, other mesenchymal tumors of bone, or other histological types of PBTs. A DL model combining radiographs and clinical features based on the EfficientNet-B3 was developed for five-category classification. The area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and specificity were calculated to evaluate model performance. The clinical utility of the model was evaluated in an observer study with four radiologists. RESULTS: The combined model achieved a macro average AUC of 0.904/0.873, with an accuracy of 67.5 %/68.7 %, a macro average sensitivity of 66.9 %/57.2 %, and a macro average specificity of 92.1 %/91.6 % on the internal/external test set, respectively. Model-assisted analysis improved accuracy, interpretation time, and confidence for junior (50.6 % vs. 72.3 %, 53.07[s] vs. 18.55[s] and 3.10 vs. 3.73 on a 5-point Likert scale [P < 0.05 for each], respectively) and senior radiologists (68.7 % vs. 75.3 %, 32.50[s] vs. 21.42[s] and 4.19 vs. 4.37 [P < 0.05 for each], respectively). CONCLUSION: The combined DL model effectively classified histological types of PBTs and assisted radiologists in achieving better classification results than their independent visual assessment.

Evaluation of gadolinium deposition in cortical bone using three-dimensional ultrashort echo time quantitative susceptibility mapping: A preliminary study.

The aim of the current study was to investigate the feasibility of three-dimensional ultrashort echo time quantitative susceptibility mapping (3D UTE-QSM) for the assessment of gadolinium (Gd) deposition in cortical bone. To this end, 40 tibial bovine cortical bone specimens were divided into five groups then soaked in phosphate-buffered saline (PBS) solutions with five different Gd concentrations of 0, 0.4, 0.8, 1.2, and 1.6 mmol/L for 48 h. Additionally, eight rabbits were randomly allocated into three groups, consisting of a normal-dose macrocyclic gadolinium-based contrast agent (GBCA) group (n = 3), a high-dose macrocyclic GBCA group (n = 3), and a control group (n = 2). All bovine and rabbit tibial bone samples underwent magnetic resonance imaging (MRI) on a 3-T clinical MR system. A 3D UTE-Cones sequence was utilized to acquire images with five different echo times (i.e., 0.032, 0.2, 0.4, 0.8, and 1.2 ms). The UTE images were subsequently processed with the morphology-enabled dipole inversion algorithm to yield a susceptibility map. The average susceptibility was calculated in three regions of interest in the middle of each specimen, and the Pearson's correlation between the estimated susceptibility and Gd concentration was calculated. The bone samples soaked in PBS with higher Gd concentrations exhibited elevated susceptibility values. A mean susceptibility value of -2.47 ± 0.23 ppm was observed for bovine bone soaked in regular PBS, while the mean QSM value increased to -1.75 ± 0.24 ppm for bone soaked in PBS with the highest Gd concentration of 1.6 mmol/L. A strong positive correlation was observed between Gd concentrations and QSM values. The mean susceptibility values of rabbit tibial specimens in the control group, normal-dose GBCA group, and high-dose GBCA group were -4.11 ± 1.52, -3.85 ± 1.33, and -3.39 ± 1.35 ppm, respectively. In conclusion, a significant linear correlation between Gd in cortical bone and QSM values was observed. The preliminary results suggest that 3D UTE-QSM may provide sensitive noninvasive assessment of Gd deposition in cortical bone.

Construction and validation of risk prediction model for gastrointestinal bleeding in patients after coronary artery bypass grafting.

This study aimed to develop a risk prediction model for gastrointestinal bleeding in patients after coronary artery bypass grafting (CABG) and assessed its accuracy. A retrospective analysis was conducted on 232 patients who underwent CABG under general anesthesia in our hospital between January 2022 and December 2022. The patients were divided into gastrointestinal bleeding (GIB) group (n = 52) and group without gastrointestinal bleeding (non-GIB) (n = 180). The independent risk factors for gastrointestinal bleeding in post-CABG patients were analyzed using χ2 test, t test and logistic multivariate regression analysis. A prediction model was established based on the identified risk factors. To verify the accuracy of the prediction model, a verification group of 161 patients who met the criteria was selected between January to June 2023, and the Bootstrap method was used for internal validation. The discrimination of the prediction model was evaluated using the area under the curve (AUC), where a higher AUC indicates a stronger discrimination effect of the model. The study developed a risk prediction model for gastrointestinal bleeding after CABG surgery. The model identified four independent risk factors: duration of stay in the intensive care unit (ICU) (OR 0.761), cardiopulmonary bypass time (OR 1.019), prolonged aortic occlusion time (OR 0.981) and re-operation for bleeding (OR 0.180). Based on these factors, an individualized risk prediction model was constructed. The C-index values of the modeling group and the verification group were 0.805 [95% CI (0.7303-0.8793)] and 0.785 [95% CI (0.6932-0.8766)], respectively, which indicated a good accuracy and discrimination of this model. The calibration and standard curves showed similar results, which further supported the accuracy of the risk prediction model. In conclusion, ICU time, cardiopulmonary bypass time, aortic occlusion time and re-operation for bleeding are identified as independent risk factors for gastrointestinal bleeding in patients after CABG. The risk prediction model developed in this study demonstrates strong predictive performance and provides valuable insights for clinical medical professionals in evaluating gastrointestinal complications in CABG patients.

Do Oral Factor Xa Inhibitors have a Role in Patients with Mechanical Heart Valves?

Mechanical heart valves (MHVs) replacement is recommended for younger patients with valvular heart disease because of the durability of the mechanical valve, but these patients need antithrombotic therapy for a lifelong time. The vitamin K antagonist (VKA) warfarin is the only approved oral antithrombotic drug for patients with MHVs. The direct oral anticoagulants (DOACs), including direct thrombin inhibitors and factor Xa (FXa) inhibitors, present some advantages over warfarin, such as weak interaction with drugs, no need for routine monitoring of coagulation function, and no frequent adjustment of medication dose. DOACs have been approved in the antithrombotic treatment of venous thromboembolism and nonvalvular atrial fibrillation. However, the application of thrombin inhibitors in patients with MHV is contraindicated because of the RE-ALIGN trial results, and FXa inhibitors are not recommended. Considering the mechanism underlying coagulation, it may be more efficient to inhibit the upstream coagulation FXa than thrombin. Recently, several preclinical and clinical investigations have looked into the role of FXa inhibitors in patients with MHVs, and have provided some encouraging evidence. Herein, we review the completed and ongoing clinical trials, and express our opinion that the following patients with a low risk of thromboembolic events should be considered for treatment with FXa inhibitors: patients with MHVs replacement 3 months postoperatively, patients with isolated aortic valve replacement, patients with a new, low thrombogenicity mechanical valve, and Asian patients. We hope this summary will provide meaningful insights for further clinical investigations.

Detection of gadolinium deposition in cortical bone with ultrashort echo time T1 mapping: an ex vivo study in a rabbit model.

OBJECTIVES: To investigate the capacity of ultrashort echo time (UTE) T1 mapping to non-invasively assess gadolinium deposition in cortical bone after gadolinium-based contrast agent (GBCA) administration. METHODS: Twenty-eight New Zealand rabbits (male, 3.0-3.5 kg) were randomly allocated into control, macrocyclic, high-dose macrocyclic, and linear GBCA groups (n = 7 for each group), and respectively given daily doses of 0.9 ml/kg bodyweight saline, 0.3 mmol/kg bodyweight gadobutrol, 0.9 mmol/kg bodyweight gadobutrol, and 0.3 mmol/kg bodyweight gadopentetate dimeglumine for five consecutive days per week over a period of 4 weeks. After a subsequent 4 weeks of recovery, the rabbits were sacrificed and their tibiae harvested. T1 value of cortical bone was measured using a combination of UTE actual flip angle imaging and variable repetition time on a 7T animal scanner. Gadolinium concentration in cortical bone was measured using inductively coupled plasma mass spectrometry (ICP-MS). Pearson's correlation between R1 value (R1 = 1/T1) and gadolinium concentration in cortical bone was assessed. RESULTS: Bone T1 values were significantly lower in the lower-dose macrocyclic (329.2 ± 21.0 ms, p < 0.05), higher-dose macrocyclic (316.8 ± 21.7 ms, p < 0.01), and linear (296.8 ± 24.1 ms, p < 0.001) GBCA groups compared with the control group (356.3 ± 19.4 ms). Gadolinium concentrations measured by ICP-MS in the control, lower-dose macrocyclic, higher-dose macrocyclic, and linear GBCA groups were 0.04 ± 0.02 µg/g, 2.60 ± 0.48 µg/g, 4.95 ± 1.17 µg/g, and 13.62 ± 1.55 µg/g, respectively. There was a strong positive correlation between R1 values and gadolinium concentrations in cortical bone (r = 0.73, p < 0.001). CONCLUSIONS: These results suggest that UTE T1 mapping has the potential to provide a non-invasive assessment of gadolinium deposition in cortical bone following GBCA administration. KEY POINTS: • Changes in T1 value related to gadolinium deposition were found in bone after both linear and macrocyclic GBCA administrations. • R1 relaxometry correlates strongly with gadolinium concentration in cortical bone. • UTE T1 mapping provides a potential tool for non-invasively monitoring gadolinium deposition in cortical bone.