Lung cancer is the most threatening tumor disease to human health. Early detection is crucial to improve the survival rate and recovery rate of lung cancer patients. Existing methods use the two-dimensional multi-view framework to learn lung nodules features and simply integrate multi-view features to achieve the classification of benign and malignant lung nodules. However, these methods suffer from the problems of not capturing the spatial features effectively and ignoring the variability of multi-views. Therefore, this paper proposes a three-dimensional (3D) multi-view convolutional neural network (MVCNN) framework. To further solve the problem of different views in the multi-view model, a 3D multi-view squeeze-and-excitation convolution neural network (MVSECNN) model is constructed by introducing the squeeze-and-excitation (SE) module in the feature fusion stage. Finally, statistical methods are used to analyze model predictions and doctor annotations. In the independent test set, the classification accuracy and sensitivity of the model were 96.04% and 98.59% respectively, which were higher than other state-of-the-art methods. The consistency score between the predictions of the model and the pathological diagnosis results was 0.948, which is significantly higher than that between the doctor annotations and the pathological diagnosis results. The methods presented in this paper can effectively learn the spatial heterogeneity of lung nodules and solve the problem of multi-view differences. At the same time, the classification of benign and malignant lung nodules can be achieved, which is of great significance for assisting doctors in clinical diagnosis.
Objective To compare the inter-observer agreement, consistency with the gold standard, and accuracy of the 2007 and 2018 versions of the AO/OTA classification in femoral intertrochanteric fractures, and to identify easily confused fracture types. Methods X-ray images of patients with femoral intertrochanteric fractures at Daping Hospital, Army Medical University between 2017 and 2021 were retrospectively collected. Three senior orthopedic trauma surgeons independently classified the fractures using both the 2007 and 2018 AO/OTA versions. A committee of five experts established the gold standard. Kappa coefficients were used to evaluate inter-observer agreement and consistency with the gold standard, while a confusion matrix was used to analyze accuracy and confusion points. Results A total of 236 patients were included. Regarding inter-observer agreement, the 2007 version was superior to the 2018 version at the subtype level [Kappa value: (0.473-0.739) vs. (0.322-0.658)], with no significant difference at the subgroup level [Kappa value: (0.234-0.453) vs. (0.204-0.442)]. Regarding consistency with the gold standard, the 2018 version was slightly better than the 2007 version [Kappa value: (0.332-0.629) vs. (0.269-0.581)] at the subgroup level. In terms of accuracy, the 2007 version showed higher accuracy at the subtype level (72.50% vs. 70.11%), whereas the 2018 version demonstrated better accuracy at the subgroup level (59.04% vs. 51.99%). The most easily confused subtypes in both versions were A1 and A2. At the subgroup level, A2.2 was the most easily confused type in both versions. Conclusions There is inconsistency in the application of both classification versions by surgeons. The 2007 version demonstrates slightly better inter-observer agreement at the subtype level, while the 2018 version shows better accuracy at the subgroup level. The A2.2 subgroup is a major point of confusion, suggesting that clinical attention should be focused on this type or that auxiliary tools may be needed to improve accuracy.
The gait acquisition system can be used for gait analysis. The traditional wearable gait acquisition system will lead to large errors in gait parameters due to different wearing positions of sensors. The gait acquisition system based on marker method is expensive and needs to be used by combining with the force measurement system under the guidance of rehabilitation doctors. Due to the complex operation, it is inconvenient for clinical application. In this paper, a gait signal acquisition system that combines foot pressure detection and Azure Kinect system is designed. Fifteen subjects are organized to participate in gait test, and relevant data are collected. The calculation method of gait spatiotemporal parameters and joint angle parameters is proposed, and the consistency analysis and error analysis of the gait parameters of proposed system and camera marking method are carried out. The results show that the parameters obtained by the two systems have good consistency (Pearson correlation coefficient r ≥ 0.9, P < 0.05) and have small error (root mean square error of gait parameters is less than 0.1, root mean square error of joint angle parameters is less than 6). In conclusion, the gait acquisition system and its parameter extraction method proposed in this paper can provide reliable data acquisition results as a theoretical basis for gait feature analysis in clinical medicine.
Objectives
To evaluate the reporting quality of Bland-Altman method consistency evaluation in China from 2014 to 2016.
Methods
WanFang Data, VIP and CNKI databases were electronically searched to collect literature about the application of Bland-Altman method from 2014 to 2016 in China. Two reviewers screened literature, extracted data, and the data were then statistically analyzed by SPSS 22.0 software.
Results
A total of 376 articles were included. The published articles on Bland-Altman method had major flaws (not conforming to reporting standards) in the application conditions, evaluation indexes, graphic depiction and so on. Merely 11.4% of the literature set the clinically acceptable consensus values in the pre-period studies. Merely one literature (0.3%) correctly compared the 95%CI of 95%LoA with the clinically acceptable threshold which had been set previously. The offer rates of the differences between the two measurements and the 95%CI, 95%LoA and 95%CI of 95%LoA in the figure were 95.9%, 9.5%, 94.6% and 4.4%, respectively.
Conclusions
The reporting quality of Bland-Altman method consistency evaluation in China is of low quality, specifically not conforming to reporting standards. We should strengthen the introduction of Bland-Altman methodology to improve the reporting quality.
The aim of this paper is to reveal the change of the brain function for nicotine addicts after smoking cessation, and explore the basis of neural physiology for the nicotine addicts in the process of smoking cessation. Fourteen subjects, who have a strong dependence on nicotine, have agreed to give up smoking and insist on completing the test, and 11 volunteers were recruited as the controls. The resting state functional magnetic resonance imaging and the regional homogeneity (ReHo) algorithm have been used to study the neural activity before and after smoking cessation. A two factors mixed design was used to investigate within-group effects and between-group effects. After 2 weeks’ smoking cessation, the increased ReHo value were exhibited in the brain area of supplementary motor area, paracentral lobule, calcarine, cuneus and lingual gyrus. It suggested that the synchronization of neural activity was enhanced in these brain areas. And between-group interaction effects were appeared in supplementary motor area, paracentral lobule, precentral gyrus, postcentral gyrus, and superior frontal gyrus. The results indicate that the brain function in supplementary motor area of smoking addicts would be enhanced significantly after 2 weeks’ smoking cessation.
Compared with traditional head to head meta-analysis, network meta-analysis has more confounding factors and difficulties to handle. Due to the mutual transitivity of evidence in network meta-analysis, heterogeneity may be brought into indirect meta-analysis. Hence, effective differentiation and correct handling of heterogeneity are being current focus. In order to ensure the reliability of the results of network meta-analysis, the concept of homogeneity is proposed and a series of methods are developed for differentiation and handling of homogeneity. Based on the extension of Bucher methods, current methods for differentiation and handling of homogeneity has extended to ten quantitative measures (eg., node analysis method, hypothesis tests, and two-step method). However, because of the differences and the focus of fundamental methodological theories as well as the limitation of statistics power, no highly-effective method has been worked out. Therefore, the exploration of highly-effective, simple and high-resolved methods are still needed.
Objective To investigate an evaluation method of medical literature applicability to clinical work， and provide a convenient way for physicians to search for the best evidence. Methods Delphi method was used to choose appropriate evaluating indexes， analytic hierarchy process was performed to determine the weighing of each index， and the formula to calculate medical literature applicability was formed. The practicability of this formula was evaluated by consistency checking between the formula’s results and experts’ opinions on literature applicability. Results Five evaluating indexes were determined， including literature’s publishing year （X1）， whether the target questions were covered （X2）, sample size （X3）， trial category （X4）， and journal level （X5）. The formula to calculate medical literature applicability was Y=3.93 X1＋11.78 X2＋14.83 X3＋44.53 X4＋24.93 X5. The result of consistency checking showed that the formula’s results were highly consistent with experts’ opinions （Kappa=0.75， P＜0.001）. Conclusion The applicability formula is a valuable tool to evaluate medical literature applicability.
