Prediction of risk factors for linezolid-induced thrombocytopenia based on neural network model.

Background: Based on real-world medical data, the artificial neural network model was used to predict the risk factors of linezolid-induced thrombocytopenia to provide a reference for better clinical use of this drug and achieve the timely prevention of adverse reactions. Methods: The artificial neural network algorithm was used to construct the prediction model of the risk factors of linezolid-induced thrombocytopenia and further evaluate the effectiveness of the artificial neural network model compared with the traditional Logistic regression model. Results: A total of 1,837 patients receiving linezolid treatment in a hospital in Xi 'an, Shaanxi Province from 1 January 2011 to 1 January 2021 were recruited. According to the exclusion criteria, 1,273 cases that did not meet the requirements of the study were excluded. A total of 564 valid cases were included in the study, with 89 (15.78%) having thrombocytopenia. The prediction accuracy of the artificial neural network model was 96.32%, and the AUROC was 0.944, which was significantly higher than that of the Logistic regression model, which was 86.14%, and the AUROC was 0.796. In the artificial neural network model, urea, platelet baseline value and serum albumin were among the top three important risk factors. Conclusion: The predictive performance of the artificial neural network model is better than that of the traditional Logistic regression model, and it can well predict the risk factors of linezolid-induced thrombocytopenia.

7295 elderly hospitalized patients with catheter-associated urinary tract infection: a case-control study.

BACKGROUND: Catheter-associated urinary tract infection (CAUTI) ranks second among nosocomial infections in elderly patients after lung infections. Improper treatment can lead to death. This study analysed the risk factors, pathogen distribution, clinical characteristics and outcomes of CAUTI in elderly inpatients with a large sample size to provide evidence for clinical prevention and control. METHODS: Based on the HIS and LIS, a case‒control study was conducted on all hospitalized patients with indwelling urinary catheters ≥ 60 years old from January 1, 2019, to December 31, 2022, and the patients were divided into the CAUTI group and the non-CAUTI group. RESULTS: CAUTI occurred in 182 of 7295 patients, and the infection rate was 3.4/per 1000 catheter days. Urine pH ≥ 6.5, moderate dependence or severe dependence in the classification of self-care ability, age ≥ 74 years, male sex, hospitalization ≥ 14 days, indwelling urinary catheter ≥ 10 days, diabetes and malnutrition were independent risk factors for CAUTI (P < 0.05). A total of 276 strains of pathogenic bacteria were detected in urine samples of 182 CAUTI patients at different times during hospitalization. The main pathogens were gram-negative bacteria (n = 132, 47.83%), followed by gram-positive bacteria (n = 91, 32.97%) and fungi (n = 53, 19.20%). Fever, abnormal procalcitonin, positive urinary nitrite and abnormal urination function were the clinical characteristics of elderly CAUTI patients (P < 0.001). Once CAUTI occurred in elderly patients, the hospitalization days were increased by 18 days, the total hospitalization cost increased by ¥18,000, and discharge all-cause mortality increased by 2.314 times (P<0.001). CONCLUSION: The situation of CAUTI in the elderly is not optimistic, it is easy to have a one-person multi-pathogen infection, and the proportion of fungi infection is not low. Urine pH ≥ 6.5, moderate or severe dependence on others and malnutrition were rare risk factors for elderly CAUTI in previous studies. Our study analysed the clinical characteristics of CAUTI in the elderly through a large sample size, which provided a reliable basis for its diagnosis and identified the adverse outcome of CAUTI.

Predicting estimated glomerular filtration rate after partial and radical nephrectomy based on split renal function measured by radionuclide: a large-scale retrospective study.

PURPOSE: The purpose of this study was to develop predictive models for postoperative estimated glomerular filtration rate (eGFR) based on the split glomerular filtration rate measured by radionuclide (rGFR), as choosing radical nephrectomy (RN) or partial nephrectomy (PN) for complex renal masses requires accurate prediction of postoperative eGFR. METHODS: Patients who underwent RN or PN for a single renal mass at Xijing Hospital between 2008 and 2022 were retrospectively included. Preoperative split rGFR was evaluated using technetium-99 m-diethylenetriaminepentaacetic acid (Tc-99 m DTPA) renal dynamic imaging, and the postoperative short-term (< 7 days) and long-term (3 months to 5 years) eGFRs were assessed. Linear mixed-effect models were used to predict eGFRs, with marginal R2 reflecting predictive ability. RESULTS: After excluding patients with missing follow-up eGFRs, the data of 2251 (RN: 1286, PN: 965) and 2447 (RN: 1417, PN: 1030) patients were respectively included in the long-term and short-term models. Two models were established to predict long-term eGFRs after RN (marginal R2 = 0.554) and PN (marginal R2 = 0.630), respectively. Two other models were established to predict short-term eGFRs after RN (marginal R2 = 0.692) and PN (marginal R2 = 0.656), respectively. In terms of long-term eGFRs, laparoscopic and robotic surgery were superior to open surgery in both PN and RN. CONCLUSIONS: We developed novel tools for predicting short-term and long-term eGFRs after RN and PN based on split rGFR that can help in preoperative decision-making.

Automatic quantification and grading of hip bone marrow oedema in ankylosing spondylitis based on deep learning.

OBJECTIVE: This study has developed a new automatic algorithm for the quantificationy and grading of ankylosing spondylitis (AS)-hip arthritis with magnetic resonance imaging (MRI). METHODS: (1) This study designs a new segmentation network based on deep learning, and a classification network based on deep learning. (2) We train the segmentation model and classification model with the training data and validate the performance of the model. (3) The segmentation results of inflammation in MRI images were obtained and the hip joint was quantified using the segmentation results. RESULTS: A retrospective analysis was performed on 141 cases; 101 patients were included in the derived cohort and 40 in the validation cohort. In the derivation group, median percentage of bone marrow oedema (BME) for each grade was as follows: 36% for grade 1 (<15%), 42% for grade 2 (15-30%),and 22% for grade 3 (≥30%). The accuracy of 44 cases on 835 AS images was 85.7%. Our model made 31 correct decisions out of 40 AS test cases. This study showed that THE accuracy rate 85.7%. CONCLUSIONS: An automatic computer-based analysis of MRI has the potential of being a useful method for the diagnosis and grading of AS hip BME.

Automatic segmentation and grading of ankylosing spondylitis on MR images via lightweight hybrid multi-scale convolutional neural network with reinforcement learning.

Objective.Ankylosing spondylitis (AS) is a disabling systemic disease that seriously threatens the patient's quality of life. Magnetic resonance imaging (MRI) is highly preferred in clinical diagnosis due to its high contrast and tissue resolution. However, since the uncertainty and intensity inhomogeneous of the AS lesions in MRI, it is still challenging and time-consuming for doctors to quantify the lesions to determine the grade of the patient's condition. Thus, an automatic AS grading method is presented in this study, which integrates the lesion segmentation and grading in a pipeline.Approach. To tackle the large variations in lesion shapes, sizes, and intensity distributions, a lightweight hybrid multi-scale convolutional neural network with reinforcement learning (LHR-Net) is proposed for the AS lesion segmentation. Specifically, the proposed LHR-Net is equipped with the newly proposed hybrid multi-scale module, which consists of multiply convolution layers with different kernel sizes and dilation rates for extracting sufficient multi-scale features. Additionally, a reinforcement learning-based data augmentation module is utilized to deal with the subjects with diffuse and fuzzy lesions that are difficult to segment. Furthermore, to resolve the incomplete segmentation results caused by the inhomogeneous intensity distributions of the AS lesions in MR images, a voxel constraint strategy is proposed to weigh the training voxel labels in the lesion regions. With the accurately segmented AS lesions, automatic AS grading is then performed by a ResNet-50-based classification network.Main results. The performance of the proposed LHR-Net was extensively evaluated on a clinically collected AS MRI dataset, which includes 100 subjects. Dice similarity coefficient (DSC), average surface distance, Hausdorff Distance at95thpercentile (HD95), predicted positive volume, and sensitivity were employed to quantitatively evaluate the segmentation results. The average DSC of the proposed LHR-Net on the AS dataset reached 0.71 on the test set, which outperforms the other state-of-the-art segmentation method by 0.04.Significance. With the accurately segmented lesions, 31 subjects in the test set (38 subjects) were correctly graded, which demonstrates that the proposed LHR-Net might provide a potential automatic method for reproducible computer-assisted diagnosis of AS grading.

[Application of Internet of Things Technology in Isolation Ward].

Based on the existing information construction foundation of the isolation ward of the hospital, according to the relevant guidelines issued by the National Health Commission, the management of environmental isolation, disinfection, medical staff management and patient management are discussed, combining the application of Internet of things technology in hospital management, a series of new applications with distinctive features of Internet of Things (IoT) are built, and advanced technology and equipment such as Internet of Things are introduced. Realize the application scenario, implementation method and business mode of intelligent IoT in isolation ward, form an integrated data management center and monitoring system through data intelligent IoT, aggregation and operation, and realize the digital collection, processing, storage, transmission and analysis of medical information, equipment information, personnel information and management information, so as to realize medical closed-loop management, reduce the hidden danger of medical safety in isolated wards and improve the level of medical quality.

Automated Segmentation Method for Low Field 3D Stomach MRI Using Transferred Learning Image Enhancement Network.

Accurate segmentation of abdominal organs has always been a difficult problem, especially for organs with cavities. And MRI-guided radiotherapy is particularly attractive for abdominal targets compared with low CT contrast. But in the limit of radiotherapy environment, only low field MRI segmentation can be used for stomach location, tracking, and treatment planning. In clinical applications, the existing 3D segmentation network model is trained by the low field MRI, and the segmentation result cannot be used in radiotherapy plan since the bad segmentation performance. Another way is that historical high field intensity MR images are directly used for data expansion to network learning; there will be a domain shift problem. How to use different domain images to improve the segmentation accuracy of deep neural network? A 3D low field MRI stomach segmentation method based on transfer learning image enhancement is proposed in this paper. In this method, Cycle Generative Adversarial Network (CycleGAN) is used to construct and learn the mapping relationship between high and low field intensity MRI and to overcome domain shift. Then, the image generated by the high field intensity MRI through the CycleGAN network is with transferred information as the extended data. The low field MRI combines these extended datasets to form the training data for training the 3D Res-Unet segmentation network. Furthermore, the convolution layer, batch normalization layer, and Relu layer together were replaced with a residual module to relieve the gradient disappearance of the neural network. The experimental results show that the Dice coefficient is 2.5 percent better than the baseline method. The over segmentation and under segmentation are reduced by 0.7 and 5.5 percent, respectively. And the sensitivity is improved by 6.4 percent.

Establishment and internal validation of preoperative nomograms for predicting the possibility of testicular salvage in patients with testicular torsion.

This study aimed to establish nomograms to preoperatively predict the possibility of testicular salvage (TS) in patients with testicular torsion. The clinical data of 204 patients with testicular torsion diagnosed at Xijing Hospital and Tangdu Hospital (Xi'an, China) between August 2008 and November 2019 were retrospectively analyzed. Univariate and multivariate logistic regression analyses were used to determine the independent predictors of TS. Based on multivariate regression coefficients, nomograms to predict possibility of TS were established. The predictive ability of the nomograms was internally validated by receiver operating characteristic (ROC) curves and calibration plots. The duration of symptoms ranged from 2 h to 1 month, with a median of 3.5 days. Thirty (14.7%) patients underwent surgical reduction and contralateral orchiopexy, while the remaining 174 (85.3%) underwent orchiectomy and contralateral orchiopexy. Finally, long symptom duration was an independent risk predictor for TS, while visible intratesticular blood flow and homogeneous testicular echotexture under color Doppler ultrasound were independent protective predictors. Internal validation showed that the nomograms, which were established by integrating these three predictive factors, had good discrimination ability in predicting the possibility of TS (areas under the ROC curves were 0.851 and 0.828, respectively). The calibration plots showed good agreement between the nomogram-predicted possibility of TS and the actual situation. In conclusion, this brief preoperative prediction tool will help clinicians to quickly determine the urgency of surgical exploration.

Risk factors for tissue expander infection in scar reconstruction: a retrospective cohort study of 2374 consecutive cases.

BACKGROUND: Tissue expansion is used for scar reconstruction owing to its excellent clinical outcomes; however, the complications that emerge from tissue expansion hinder repair. Infection is considered a major complication of tissue expansion. This study aimed to analyze the perioperative risk factors for expander infection. METHODS: A large, retrospective, single-institution observational study was carried out over a 10-year period. The study enrolled consecutive patients who had undergone tissue expansion for scar reconstruction. Demographics, etiological data, expander-related characteristics and postoperative infection were assessed. Univariate and multivariate logistic regression analysis were performed to identify risk factors for expander infection. In addition, we conducted a sensitivity analysis for treatment failure caused by infection as an outcome. RESULTS: A total of 2374 expanders and 148 cases of expander infection were assessed. Treatment failure caused by infection occurred in 14 expanders. Multivariate logistic regression analysis identified that disease duration of ≤1 year (odds ratio (OR), 2.07; p < 0.001), larger volume of expander (200-400 ml vs <200 ml; OR, 1.74; p = 0.032; >400 ml vs <200 ml; OR, 1.76; p = 0.049), limb location (OR, 2.22; p = 0.023) and hematoma evacuation (OR, 2.17; p = 0.049) were associated with a high likelihood of expander infection. Disease duration of ≤1 year (OR, 3.88; p = 0.015) and hematoma evacuation (OR, 10.35; p = 0.001) were so related to high risk of treatment failure. CONCLUSIONS: The rate of expander infection in patients undergoing scar reconstruction was 6.2%. Disease duration of <1 year, expander volume of >200 ml, limb location and postoperative hematoma evacuation were independent risk factors for expander infection.