Differential privacy preserved federated learning for prognostic modeling in COVID-19 patients using large multi-institutional chest CT dataset.

BACKGROUND: Notwithstanding the encouraging results of previous studies reporting on the efficiency of deep learning (DL) in COVID-19 prognostication, clinical adoption of the developed methodology still needs to be improved. To overcome this limitation, we set out to predict the prognosis of a large multi-institutional cohort of patients with COVID-19 using a DL-based model. PURPOSE: This study aimed to evaluate the performance of deep privacy-preserving federated learning (DPFL) in predicting COVID-19 outcomes using chest CT images. METHODS: After applying inclusion and exclusion criteria, 3055 patients from 19 centers, including 1599 alive and 1456 deceased, were enrolled in this study. Data from all centers were split (randomly with stratification respective to each center and class) into a training/validation set (70%/10%) and a hold-out test set (20%). For the DL model, feature extraction was performed on 2D slices, and averaging was performed at the final layer to construct a 3D model for each scan. The DensNet model was used for feature extraction. The model was developed using centralized and FL approaches. For FL, we employed DPFL approaches. Membership inference attack was also evaluated in the FL strategy. For model evaluation, different metrics were reported in the hold-out test sets. In addition, models trained in two scenarios, centralized and FL, were compared using the DeLong test for statistical differences. RESULTS: The centralized model achieved an accuracy of 0.76, while the DPFL model had an accuracy of 0.75. Both the centralized and DPFL models achieved a specificity of 0.77. The centralized model achieved a sensitivity of 0.74, while the DPFL model had a sensitivity of 0.73. A mean AUC of 0.82 and 0.81 with 95% confidence intervals of (95% CI: 0.79-0.85) and (95% CI: 0.77-0.84) were achieved by the centralized model and the DPFL model, respectively. The DeLong test did not prove statistically significant differences between the two models (p-value = 0.98). The AUC values for the inference attacks fluctuate between 0.49 and 0.51, with an average of 0.50 ± 0.003 and 95% CI for the mean AUC of 0.500 to 0.501. CONCLUSION: The performance of the proposed model was comparable to centralized models while operating on large and heterogeneous multi-institutional datasets. In addition, the model was resistant to inference attacks, ensuring the privacy of shared data during the training process.

Nurses' Knowledge Regarding Oxygen Therapy; a Cross-Sectional Study.

Introduction: Oxygen therapy, if done correctly, can save patients' life promptly. However, improper use will be just as dangerous. The present study aimed to investigate the level of nurses' knowledge on properly using oxygen. Method: This was a cross-sectional study with a minimum sample size of 72 nurses who were randomly selected from various wards of Masih Daneshvari Hospital, Tehran, Iran. To determine the level of knowledge about oxygen therapy, a questionnaire was used to collect data. This questionnaire consists of seven items, each of which is designed to determine the level of the individual's knowledge about the various details of oxygen therapy. Results: Seventy-eight nurses with the mean age of 35.80±7.42 years participated in the study (87% female). The mean knowledge score of nurses regarding oxygen therapy was 8.89 ± 2.79 out of 16 points. 84.6% of the nurses were able to differentiate various types of oxygen masks. Accordingly, 94.9% of nurses had good knowledge on oxygen humidification. Also, 50% of the nurses had sufficient knowledge about the amount of oxygen flow produced by different masks. 10.3% of the nurses could choose the most appropriate mask for different clinical conditions. 6.4% of the nurses had knowledge of working with flowmeters, and 15.4% of the nurses had sufficient information about the maximum level of oxygen required for the patient. 17.9% of the nurses were familiar with measuring the appropriate amount of oxygen for patients. There was no statistically significant relationship between age (p = 0.57), gender (p = 0.09), employment status (p = 0.38), workplace (p = 0.86), current position (p = 0.11), degree (p = 0.27), and graduation time (p = 0.58) of nurses with good knowledge of using oxygen. However, a statistically significant relationship was reported between nurses' related work experience and their knowledge of the proper use of oxygen (p = 0.03). Conclusion: In general, the nurses' knowledge at Masih Daneshvari Hospital on how to properly use oxygen is at a moderate level. Nurses' knowledge in some areas, such as working with the flowmeter, choosing the suitable mask for specific clinical conditions, and the maximum oxygen required for patients, is meager and requires training intervention.

COVID-19 prognostic modeling using CT radiomic features and machine learning algorithms: Analysis of a multi-institutional dataset of 14,339 patients.

BACKGROUND: We aimed to analyze the prognostic power of CT-based radiomics models using data of 14,339 COVID-19 patients. METHODS: Whole lung segmentations were performed automatically using a deep learning-based model to extract 107 intensity and texture radiomics features. We used four feature selection algorithms and seven classifiers. We evaluated the models using ten different splitting and cross-validation strategies, including non-harmonized and ComBat-harmonized datasets. The sensitivity, specificity, and area under the receiver operating characteristic curve (AUC) were reported. RESULTS: In the test dataset (4,301) consisting of CT and/or RT-PCR positive cases, AUC, sensitivity, and specificity of 0.83 ± 0.01 (CI95%: 0.81-0.85), 0.81, and 0.72, respectively, were obtained by ANOVA feature selector + Random Forest (RF) classifier. Similar results were achieved in RT-PCR-only positive test sets (3,644). In ComBat harmonized dataset, Relief feature selector + RF classifier resulted in the highest performance of AUC, reaching 0.83 ± 0.01 (CI95%: 0.81-0.85), with a sensitivity and specificity of 0.77 and 0.74, respectively. ComBat harmonization did not depict statistically significant improvement compared to a non-harmonized dataset. In leave-one-center-out, the combination of ANOVA feature selector and RF classifier resulted in the highest performance. CONCLUSION: Lung CT radiomics features can be used for robust prognostic modeling of COVID-19. The predictive power of the proposed CT radiomics model is more reliable when using a large multicentric heterogeneous dataset, and may be used prospectively in clinical setting to manage COVID-19 patients.

Effects of External Perturbations on Anticipatory and Compensatory Postural Adjustments in Patients with Multiple Sclerosis and a Fall History.

BACKGROUND: Although previous studies have investigated postural adjustment mechanisms in patients with multiple sclerosis (MS), it seems that no study has yet investigated the relationship between anticipatory and compensatory postural adjustments (APAs and CPAs, respectively) and falls. METHODS: Seventeen MS fallers, 17 MS nonfallers, and 15 controls were exposed to a series of expected and unexpected backward pull perturbations applied at the trunk level. The electrical activity of 12 leg and trunk muscles as well as center of pressure displacement were recorded. RESULTS: The MS fallers had delayed muscle activity onsets compared with MS nonfallers and controls. In addition, a significantly lower level of muscle activity during APAs was detected in MS fallers compared with controls. Moreover, in the unexpected condition of perturbation, significantly smaller CPA was observed in MS fallers compared with controls. Both groups of patients with MS required more time to stabilize their center of pressure after both types of perturbations compared with controls. CONCLUSIONS: The inability to produce efficient APAs and CPAs during perturbations may explain the high rates of postural instability and falls in patients with MS. Findings from this study provide a background for the development of perturbation-based training programs aimed at balance improvement and fall prevention by restoring mechanisms underlying balance impairments.