RESUMO
In the last few years, many types of research have been conducted on the most harmful pandemic, COVID-19. Machine learning approaches have been applied to investigate chest X-rays of COVID-19 patients in many respects. This study focuses on the deep learning algorithm from the standpoint of feature space and similarity analysis. Firstly, we utilized Local Interpretable Model-agnostic Explanations (LIME) to justify the necessity of the region of interest (ROI) process and further prepared ROI via U-Net segmentation that masked out non-lung areas of images to prevent the classifier from being distracted by irrelevant features. The experimental results were promising, with detection performance reaching an overall accuracy of 95.5%, a sensitivity of 98.4%, a precision of 94.7%, and an F1 score of 96.5% on the COVID-19 category. Secondly, we applied similarity analysis to identify outliers and further provided an objective confidence reference specific to the similarity distance to centers or boundaries of clusters while inferring. Finally, the experimental results suggested putting more effort into enhancing the low-accuracy subspace locally, which is identified by the similarity distance to the centers. The experimental results were promising, and based on those perspectives, our approach could be more flexible to deploy dedicated classifiers specific to different subspaces instead of one rigid end-to-end black box model for all feature space.
Assuntos
COVID-19 , Conjuntos de Dados como Assunto , Aprendizado Profundo , Raios X , Humanos , Algoritmos , Radiografia Pulmonar de MassaRESUMO
BACKGROUND: The purpose of this study is to investigate whether the risks of rehospitalization caused by acute coronary syndrome (ACS) or stroke would have significant differences between diabetic and non-diabetic patients from ACS. METHODS: This was a retrospective study of 364 inpatients with ACS from 2017 to 2019. Logistic regression models included gender, age group, and the principal diagnosis of hospitalization as controlling variables which were used to analyze the dataset. RESULTS: About 10% of patients are hospitalized after recovery. Moreover, regardless of suffering from diabetes, the risk of rehospitalization does not appear to show a significant difference. In comparison with non-diabetic patients, the odds ratio of rehospitalization of diabetic patients was 0.94 (95% CI: 0.46-1.93, p-value = 0.8639) after controlling for the effects of gender, age group, and the principal diagnosis of hospitalization. CONCLUSIONS: Diabetic patients seem to perform well in controlling LDL-C (low-density lipoprotein cholesterol) after ACS recoveries.
RESUMO
Hot compress modalities are used to ameliorate pain despite prevalent confusion about which modality should be used and when. Most recommendations for hot compresses are based on empirical experience, with limited evidence to support its efficacy. To obtain insight into the nerve transmission mechanism of hot compresses and to identify the nerve injury marker proteins specifically associated with sciatic nerve pain, we established a rat model of chronic constriction injury (CCI) and performed mechanical allodynia, electrophysiology, and histopathological analysis. All CCI rats exhibited geometric representation of the affected hind paw, which indicated a hyper-impact on both mechanical gait and asymmetry of gait on day 28. The CCI model after 28 days of surgery significantly reduced compound muscle action potential (CMAP) amplitude, but also significantly reduced latency. Administration of hot compress for 3 weeks (heated at 40-42°C, cycle of 40 min, and rest for 20 min, three cycles each time, three times per week) significantly increased the paw withdrawal thresholds in response to stimulation by Von Frey fibers and reversed the CCI-induced reduced sciatic functional index (SFI) scores. Hot compress treatment in the CCI model improved CMAP amplitude and latency. The S100 protein expression level in the CCI+Hot compression group was 1.5-fold higher than in the CCI group; it dramatically reduced inflammation, such as tumor necrosis factor alpha and CD68 expression in nerve injury sites. Synaptophysin (Syn) expression in the CCI+Hot compression group was less than threefold in the CCI group at both nerve injury sites and brain (somatosensory cortex and hippocampus). This finding indicates that local nerve damage and inflammation cause significant alterations in the sensorimotor strip, and hot compress treatment could significantly ameliorate sciatic nerve pain by attenuating Syn and inflammatory factors from local pathological nerves to the brain. This study determines the potential efficacy and safety of hot compress, and may have important implications for its widespread use in sciatic nerve pain treatment.
RESUMO
BACKGROUND: Intramedullary (IM) or extramedullary (EM) mechanical guides are used as alignment tools during total knee arthroplasty (TKA) surgery. The EM guide is less invasive; however, the IM mechanical guide is the preferred option since it has shown superior outcomes in several studies. Picture archive and communication system (PACS) images, if available, are extensively used for preoperative planning and intraoperative guidance. This retrospective study compared TKA outcomes using the conventional IM guide and a new EM technique which uses PACS image for preoperative and intraoperative assessment bone resection. To the best of our knowledge, this is the first study with the new EM technique. MATERIALS AND METHODS: The study was performed on 205 knees (190 patients) for TKA from 2011 to 2013. The perioperative blood loss and the postoperative alignment angles were assessed for both mechanical guides. The angles were measured on the radiographs of the patient. The blood loss was assessed by the blood accumulated in the hemovac drain during the surgery and until 3 days after the surgery. RESULTS: The new EM guide provided similar postoperative alignment as that obtained with the IM guide. CONCLUSION: The EM-guided method for femoral bone cutting using PACS films in TKA is as good as the IM method. The additional advantages of less injury to the bone and less fat emboli load to the cardiopulmonary system with the EM method makes it an attractive choice for routine, especially in the elderly and/or simultaneous bilateral, TKA in hospitals without modern computer-assisted navigation systems.
RESUMO
Electrical instrumentation is now a real-time phenomenon and an active part of smart phone trends. Modern electronic devices and controllers can be seen everywhere. Personal information detectors provide timely and effective medical care. In previous times, personnel electronic devices, such as the smart phone, i-watch, and e-clothes were diffi cult to link. Nowadays, it's a good time to prevent a health tragedy in certain people. This study tries to focus on discussing the remote control and smart phone display topic, through ECG instrumentation from e-clothes to prevent the human body from taking excessive exercise. Besides, the research also built a database in a remote server, where it stored historical ECG records. Through the model, a real sport (running) case study and system validation was proposed. Humans have a possibility to optimize their sporting style. The e-clothes can send GPS signals to a remote server. The system would then track the routes taken and monitor human body loading. Future research suggests that mass human trials or more mathematical hybrid model designs should take place.