OBJECTIVES: To design a deep learning model that balances model complexity and performance to enable its integration into wearable ECG monitoring devices for automated diagnosis of atrial fibrillation. METHODS: This study was performed based on data from 84 patients with atrial fibrillation, 25 patients with atrial fibrillation, and 18 subjects without obvious arrhythmia collected from the publicly available datasets LTAFDB, AFDB, and NSRDB, respectively. A lightweight attention network based on depthwise separable convolution and fusion of channel-spatial information, namely DSC-AttNet, was proposed. Depthwise separable convolution was introduced to replace standard convolution and reduce model parameters and computational complexity to realize high efficiency and light weight of the model. The multilayer hybrid attention mechanism was embedded to compute the attentional weights of the channels and spatial information at different scales to improve the feature expression ability of the model. Ten-fold cross-validation was performed on LTAFDB, and external independent testing was conducted on AFDB and NSRDB datasets. RESULTS: DSC-AttNet achieved a ten-fold average accuracy of 97.33% and a precision of 97.30% on the test set, both of which outperformed the other 4 comparison models as well as the 3 classical models. The accuracy of the model on the external test set reached 92.78%, better than those of the 3 classical models. The number of parameters of DSC-AttNet was 1.01M, and the computational volume was 27.19G, both smaller than the 3 classical models. CONCLUSIONS This proposed method has a smaller complexity, achieves better classification performance, and has a better generalization ability for atrial fibrillation classification.
Atrial Fibrillation/diagnosis* ; Humans ; Electrocardiography ; Deep Learning ; Wearable Electronic Devices ; Neural Networks, Computer

OBJECTIVES: We propose a myocardial infarction (MI) detection and localization model for improving the diagnostic accuracy for MI to provide assistance to clinical decision-making. METHODS: The proposed model was constructed based on multi-scale field residual blocks fusion modified channel attention (MSF-RB-MCA). The model utilizes lead II electrocardiogram (ECG) signals to detect and localize MI, and extracts different levels of feature information through the multi-scale field residual block. A modified channel attention for automatic adjustment of the feature weights was introduced to enhance the model's ability to focus on the MI region, thereby improving the accuracy of MI detection and localization. RESULTS: A 5-fold cross-validation test of the model was performed using the publicly available Physikalisch-Technische Bundesanstalt (PTB) dataset. For MI detection, the model achieved an accuracy of 99.96% on the test set with a specificity of 99.84% and a sensitivity of 99.99%. For MI localization, the accuracy, specificity and sensitivity were 99.81%, 99.98% and 99.65%, respectively. The performances of the model for MI detection and localization were superior to those of other comparison models. CONCLUSIONS The proposed MSF-RB-MCA model shows excellent performance in AI detection and localization based on lead II ECG signals, demonstrating its great potential for application in wearable devices.
Myocardial Infarction/diagnosis* ; Humans ; Electrocardiography/methods* ; Signal Processing, Computer-Assisted ; Algorithms ; Sensitivity and Specificity

Objective:To observe the effect of a new type of endotracheal tube occluder,which can be finely tuned,combined with rehabilitation training on respiratory and speech function of patients with severe tracheotomy. Method:A total of 30 patients with severe tracheotomy were randomly divided into test group and control group,15 cases in each group of patients received severe clinical and rehabilitation treatment,including lung rehabilitation,swallowing and phonation training and exercise therapy.Patients in the test group wore adjust-able tracheal tube occluder during rehabilitation period.The rehabilitation treatment time of the two groups was twice a day for two consecutive weeks.After 2 weeks of treatment,clinical pulmonary infection score(CPIS),semi quantitative cough strength score(SCSS),maximum phonation time(MPT),volume and tracheal tube extubation rate were compared between the two groups. Result:After treatment,the CPIs score,SCSS score,MPT and maximum volume value of the two groups were significantly improved in both groups(P＜0.001).The SCSS score of the test group improved more than that of the control group(P＜0.05).The MPT and maximum volume of the test group were higher than those of the control group(P＜0.05).The extubation rate of the test group was higher than that of the control group. Conclusion:The adjustable tracheal tube occluder is safe and effective,and combining with rehabilitation therapy,has positive significance for improving the respiratory and speech function of patients with severe tra-cheotomy.

OBJECTIVE To evaluate the effectiveness， safety， economy， innovation， suitability and accessibility of recombinant Mycobacterium tuberculosis fusion protein （EC）， and to provide evidence for selecting skin detection methods for tuberculosis infection diagnosis and auxiliary diagnosis of tuberculosis. METHODS The effectiveness and safety of EC compared with purified protein derivative of tuberculin （TB-PPD） were analyzed by the method of systematic review. Cost minimization analysis， cost-effectiveness analysis and cost-utility analysis were used to evaluate the short-term economy of EC compared with TB-PPD， and cost-utility analysis was used to evaluate the long-term economy. The evaluation dimensions of innovation， suitability and accessibility were determined by systematic review and improved Delphi expert consultation， and the comprehensive score of EC and TB-PPD in each dimension were calculated by the weight of each indicator. RESULTS The scores of effectiveness， safety， economy， innovation and suitability of EC were all higher than those of TB-PPD. The affordability scores of the two drugs were consistent， while the availability score of EC was lower than those of TB-PPD. After considering dimensions and index weight， the scores of effectiveness， safety， economy， innovation， suitability， accessibility and the comprehensive score of EC were all higher than those of TB-PPD. CONCLUSIONS Compared with TB-PPD， EC performs better in all dimensions of effectiveness， safety， economy， innovation， suitability and accessibility. However， it is worth noting that EC should further improve its availability in the dimension of accessibility.

ObjectiveTo observe the early and mid-term clinical effect of occupational therapy on patients after total knee arthroplasty (TKA) during perioperative period. MethodsFrom July to August, 2018, 100 patients who underwent the first TKA in the First Affiliated Hospital of USTC were divided into control group (n = 50) and observation group (n = 50) according to the treatment plan. Both groups received routine preoperative rehabilitation education, surgical treatment and postoperative physical therapy, in addition, the observation group received perioperative occupational therapy. They were assessed with modified Barthel Index (MBI) and Instrumental Activities of Daily Living (IADL) before operation; one, two and three days after operation; on the day of discharge; one month, three months and six months after operation. ResultsOne day, two days, one month, three months and six months after operation, the scores of MBI were higher in the observation group than in the control group (t > 2.113, P < 0.05). One month, three months after operation, the scores of IADL were significantly higher in the observation group than in the control group (t > 5.125, P < 0.001). The postoperative hospitalization time was significantly shorter (t = -5.356, P < 0.001), and the total treatment cost was lower (t = -2.455, P < 0.05) in the observation group than in the control group. ConclusionOccupational therapy could effectively improve the early and mid-term activities of daily living of TKA patients and help them return to normal life earlier.

Objective:To document the blood indexes of middle-aged and elderly intracerebral hemorrhage (ICH) patients complicated with deep vein thrombosis (DVT).Methods:A retrospective analysis was conducted of 77 hospitalized ICH patients using venous color Doppler ultrasonography within 3 days of admission. According to the results, they were divided into a DVT group (18 cases) and a non-DVT group (59 cases). The blood routine, biochemistry, coagulation, and D-dimer examinations were conducted on the 2nd day after admission. T-tests and rank sum tests tested the significance of any differences between the groups in average white blood cell counts, neutrophil percentages, platelets, albumin, globulin, fasting blood glucose, urea nitrogen, creatinine, uric acid, electrolytes, fibrinogen or D-dimer.Results:The average levels of albumin, uric acid and calcium in the DVT group were significantly lower than in the non-DVT group. The average levels of fasting blood glucose and D-dimer were significantly higher.Conclusions:Decreased serum uric acid, calcium and albumin levels, together with increased fasting blood glucose and D-dimer are related to the occurrence of DVT in ICH patients. To reduce the risk of DVT it is important to maintain normal levels of serum uric acid, calcium and albumin and to limit D-dimer and fasting blood glucose.

Objective:To observe factors influencing the risk of urinary tract infection (UTI) after cerebral hemorrhage (ICH).Methods:A total of 77 ICH patients undergoing rehabilitation between August 2015 and August 2017 were studied. Among them, 24 were diagnosed with a UTI according to clinical manifestations and urine routine tests and urine culture. They formed the UTI group. The other 53 were the non-UTI group. Complete clinical evaluations were available for all 77, and those data were used to identify risk factors for UTI using univariate and multivariate logistic regression analysis.Results:The univariate analysis showed that age, side of paralysis, consciousness dysfunction, Brunnstrom staging of the lower extremity, tracheotomy, indwelling catheter, lung infection, white blood cell count, neutrophil percentage, serum natrium, uric acid, D-dimer level and fibrinogen were all significant predictors of UTI. The multivariate analysis identified age, right side hemiplegia and D-dimer level as useful predictors.Conclusions:Advanced age and a high D-dimer level are independent risk factors for UTI after ICH, while hemiplegia on the right side is a protective factor.

Objective:To compare the accuracy and generalized robustness of three predictive models of knowledge-based treatment strategies for radiotherapy for optimized model selection.Methods:The clinical radiotherapy plans of 45 prostate cancer (PC) cases and 25 nasopharyngeal cancer (NPC) cases were collected, and analyzed using three models (Z, L and S model), proposed by Zhu et al, Appenzoller et al and Shiraishi et al, respectively, to predict the dose-volume histogram (DVH) of bladder and rectum on PC cases and that of left and right parotid on NPC cases. The prediction error was measured by the difference of area under the predicted DVH and the clinical DVH curves (|V (pre_DVH)-V (clin_DVH)|), where a smaller prediction error implies a greater prediction accuracy. The accuracies of these three models were compared on the single organ at risk (OAR), and the generalized robustness of models was evaluated and compared by calculating the standard deviation of the prediction accuracy on different OAR. Results:For bladder and rectum, the prediction error of L model (0.114 and 0.163, respectively) was significantly higher than those values of Z and S models (≤0.071, P<0.05); for left parotid gland, the predicted error of S model (0.033) did not present significant difference from those values of Z and L models (≤0.025, P>0.05); for right parotid gland, S model (0.033) demonstrated significantly higher prediction error than those of Z and L models (≤0.028, P<0.05). Regarding different OAR, S model showed a lower standard deviation of prediction accuracy when comparing to Z and L models (0.016, 0.018 and 0.060, respectively). Conclusions:In the prediction of DVH in bladder and rectum of PC, Z and S models were more accurate than L model. In contrast, Z and L models demonstrated higher accuracy than S model in the prediction of left and right parotid glands of NPC. In respect to different OAR, the generalized robustness of S model was superior than the other two models.

Objective:To observe risk factors for lung infection (LI) after cerebral hemorrhage (ICH).Methods:A total of 118 patients with ICH were retrospectively analyzed. Among them, 63 were diagnosed with an LI according to clinical manifestations and X-ray or CT on admission. They were selected into an LI group, while the rest formed the non-LI group. Gender, age, duration of disease, side of paralysis, position and type of ICH, complications, cognitive disorders, consciousness, dysphagia, motor function, and serum levels of albumin, sodium, potassium, calcium and magnesium were recorded on the second day after admission. The risk factors for LI were analyzed using univariate and multivariate logistic regression analysis.Results:Age, duration of disease, side of paralysis, position and type of ICH, tracheotomy, urinary tract infection, cognitive impairment, dysphagia, motor function, and serum levels of albumin, sodium, calcium, and magnesium were all significant predictors of LI after an ICH. Multivariate logistic regression analysis showed that right side hemiplegia, bilateral paralysis, subarachnoid hemorrhage, urinary tract infection, and serum magnesium are all significant predictors.Conclusions:Right or bilateral paralysis, subarachnoid hemorrhage and urinary tract infection are independent risk factors for LI after an ICH. High serum magnesium within the normal range can reduce the incidence of LI.

Objective:To explore the factors related to cognitive dysfunction after traumatic brain injury.Methods:A total of 55 patients with traumatic brain injury were analyzed retrospectively. The clinical information about the patients, including age, gender, paralyzed side, course of the disease, type of injury, degree of injury, treatment, speech function (grades of Boston aphasia severity), swallowing function (7-level evaluation method), limb motor function (Brunnstrom stages of the upper limb, hand and lower limb), complications (tracheotomy, lung infection, urinary tract infection, pressure sores, electrolyte imbalance and hypoproteinemia), whether the patient received rehabilitation therapy, and D-dimer and coagulation examination results were recorded. Univariate analysis and ordered multi-class logistic regression analysis were conducted to screen the related factors using Rancho Los Amigos cognitive function grades as the dependent variable.Results:The results of the univariate analysis showed that consciousness, aphasia severity, swallowing ability, Brunnstrom stages, having received a tracheotomy, urinary tract infection, hypoproteinemia, fibrinogen and D-dimer all significantly predicted Rancho Los Amigos cognitive function grades. The logistic regression analysis showed that disordered consciousness, aphasia severity, Brunnstrom stage of the upper limbs, and urinary tract infection were significant predictors.Conclusions:Disordered consciousness, aphasia severity, Brunnstrom stage of the upper limbs, and urinary tract infection are factors related to cognitive dysfunction after traumatic brain injury.