RESUMO
OBJECTIVES: Based on the current clinical routine, we aimed to develop a novel deep learning model to distinguish coronavirus disease 2019 (COVID-19) pneumonia from other types of pneumonia and validate it with a real-world dataset (RWD). METHODS: A total of 563 chest CT scans of 380 patients (227/380 were diagnosed with COVID-19 pneumonia) from 5 hospitals were collected to train our deep learning (DL) model. Lung regions were extracted by U-net, then transformed and fed to pre-trained ResNet-50-based IDANNet (Identification and Analysis of New covid-19 Net) to produce a diagnostic probability. Fivefold cross-validation was employed to validate the application of our model. Another 318 scans of 316 patients (243/316 were diagnosed with COVID-19 pneumonia) from 2 other hospitals were enrolled prospectively as the RWDs to testify our DL model's performance and compared it with that from 3 experienced radiologists. RESULTS: A three-dimensional DL model was successfully established. The diagnostic threshold to differentiate COVID-19 and non-COVID-19 pneumonia was 0.685 with an AUC of 0.906 (95% CI: 0.886-0.913) in the internal validation group. In the RWD cohort, our model achieved an AUC of 0.868 (95% CI: 0.851-0.876) with the sensitivity of 0.811 and the specificity of 0.822, non-inferior to the performance of 3 experienced radiologists, suggesting promising clinical practical usage. CONCLUSIONS: The established DL model was able to achieve accurate identification of COVID-19 pneumonia from other suspected ones in the real-world situation, which could become a reliable tool in clinical routine. KEY POINTS: ⢠In an internal validation set, our DL model achieved the best performance to differentiate COVID-19 from non-COVID-19 pneumonia with a sensitivity of 0.836, a specificity of 0.800, and an AUC of 0.906 (95% CI: 0.886-0.913) when the threshold was set at 0.685. ⢠In the prospective RWD cohort, our DL diagnostic model achieved a sensitivity of 0.811, a specificity of 0.822, and AUC of 0.868 (95% CI: 0.851-0.876), non-inferior to the performance of 3 experienced radiologists. ⢠The attention heatmaps were fully generated by the model without additional manual annotation and the attention regions were highly aligned with the ROIs acquired by human radiologists for diagnosis.
Assuntos
COVID-19 , Aprendizado Profundo , Pneumonia Viral , Humanos , Redes Neurais de Computação , Pneumonia Viral/diagnóstico por imagem , Estudos Prospectivos , SARS-CoV-2 , Tomografia Computadorizada por Raios XRESUMO
The construction of a smart city puts forward new requirements for lighting systems, such as variable color temperature adapting to environment and low energy consumption. We introduce a variable color temperature laser lighting system that produces uniform light with minimum energy. The color temperature is controlled by tri-color RGB diode lasers, and uniform lighting is achieved by microlens arrays. Tri-color diode lasers with wavelengths of 650, 556, and 450 nm are used as the lighting sources, and the white light laser output is achieved by combining the three beams. The color temperature is controlled by changing the power ratio of each lighting source. Finally, the homogenization of laser energy is regulated by the microlens arrays, and the energy uniformity reaches 91.1%. Moreover, we do an experiment to compare LED street lighting and laser street lighting, finding that the street lighting system with this design can increase the energy utilization rate by 113.33%, and the color temperature of the car headlamps with this design can be changed according to the environment. Therefore, this laser lighting system is an effective solution for modern smart lighting systems and energy saving, which have vast application.
RESUMO
Double-sided microlens arrays (DSMLAs) include combinations of two single-sided MLAs to overcome positioning errors and greatly improve light transmissivity compared to other types of lenses. Precision glass molding (PGM) is used to fabricate DSMLAs, but controlling alignment errors during this process is challenging. In this paper, a mold assembly was manufactured with a novel combination of materials to improve the alignment accuracy of mold cores during PGM by using the nonlinear thermal expansion characteristics of the various materials to improve the DSMLA alignment accuracy. By establishing a mathematical model of the DSMLA alignment error and a thermal expansion model of the mold-sleeve pair, the relationship between the maximum alignment error of the DSMLA and the mold-sleeve gap was determined. This research provides a method to optimize the mold-sleeve gap and minimize the alignment error of the DSMLA. The measured DSMLA alignment error was 10.56 µm, which is similar to the predicted maximum alignment error. Optical measurements showed that the uniformity of the homogenized beam spot was 97.81%, and the effective homogeneous area accounted for 91.66% of the total area. This proposed method provides a novel strategy to improve the performance of DSMLAs.
RESUMO
BACKGROUND AND PURPOSE: As the only ionotropic receptor in the 5-HT receptor family, the 5-HT3 receptor (5-HT3 R) is involved in psychiatric disorders and its modulators have potential therapeutic effects for cognitive impairment in these disorders. However, it remains unclear how 5-HT3 Rs shape synaptic plasticity for memory function. EXPERIMENTAL APPROACH: Extracellular as well as whole-cell electrophysiological recordings were used to monitor hippocampal LTP and synaptic transmission in hippocampal slices in 5-HT3 AR knockout or 5-HT3 AR-GFP mice. Immunocytochemistry, qRT-PCR and western blotting were used to measure receptor expression. We also assessed hippocampal dependent cognition and memory, using the Morris water maze (MWM) and novel object recognition. KEY RESULTS: We found that 5-HT3 R dysfunction impaired hippocampal LTP in Schaffer collateral (SC)-CA1 pathway in hippocampal slices, by facilitating GABAergic inputs in pyramidal cells. This effect was dependent on 5-HT3 Rs on axon terminals. It resulted from reduced expression and function of the cannabinoid receptor 1 (CB1 R) co-localized with 5-HT3 Rs on axon terminals, and then led to diminishment of tonic inhibition of GABA release by CB1 Rs. Inhibition of CB1 Rs mimicked the facilitation of GABAergic transmission by 5-HT3 R disruption. Consequently, mice with hippocampal 5-HT3 R disruption exhibited impaired spatial memory in MWM tasks. CONCLUSION AND IMPLICATIONS: These results suggest that 5-HT3 Rs are crucial in enabling hippocampal synaptic plasticity via a novel CB1 R-GABAA -dependent pathway to regulate spatial memory.
Assuntos
Potenciação de Longa Duração , Memória Espacial , Animais , Região CA1 Hipocampal/metabolismo , Hipocampo/metabolismo , Humanos , Potenciação de Longa Duração/fisiologia , Transtornos da Memória/metabolismo , Camundongos , Receptor CB1 de Canabinoide/genética , Receptor CB1 de Canabinoide/metabolismo , Receptores de GABA-A/metabolismo , Serotonina/metabolismo , Ácido gama-Aminobutírico/metabolismoRESUMO
OBJECTIVES: Sinus of Valsalva aneurysms (SVAs) are rare and are often complicated by aortic insufficiency (AI). Treating AI is important for achieving good long-term results in patients with SVA. Here, we have summarized our experience with the surgical management of patients with SVA with concomitant AI. METHODS: Patients who were diagnosed with SVA and underwent surgical treatment between January 2008 and May 2016 were included. Clinical characteristics, including age, gender, SVA anatomy and concurrent anomalies, were analysed. The surgical strategies, intraoperative results and early and late outcomes were evaluated. RESULTS: A total of 178 patients (age 37.4 ± 13.1 years, 114 men) were identified. Eighty-seven (48.9%) patients had at least 2+ AI preoperatively. Patients with AI had a significant higher incidence of right coronary SVA with concomitant ventricular septal defects (80.5% vs 54.9%, P < 0.001). Concurrent aortic valve surgery was performed in 70 patients with 63 valve replacements and 7 valve repairs. The other 17 patients underwent SVA repair only. The mean follow-up time was 44.4 ± 33.8 months. During the follow-up, no patients who underwent valve replacement experienced perivalvular leakage, whereas 7 patients who underwent aortic valve repair had trivial-to-mild AI and 3 of the 17 patients who underwent SVA repair only presented with moderate AI. CONCLUSIONS: Various surgical techniques can be utilized to manage concomitant AI without compromising short-term outcomes. Valve replacement yields good long-term results. According to our experience, valve repair, especially valve-sparing procedures performed by experienced surgeons, could be an acceptable option when managing concomitant AI in young patients.
Assuntos
Aneurisma Aórtico/cirurgia , Insuficiência da Valva Aórtica/cirurgia , Valva Aórtica/cirurgia , Procedimentos Cirúrgicos Cardíacos/métodos , Seio Aórtico/cirurgia , Adolescente , Adulto , Idoso , Aneurisma Aórtico/complicações , Aneurisma Aórtico/diagnóstico , Insuficiência da Valva Aórtica/complicações , Insuficiência da Valva Aórtica/diagnóstico , Ecocardiografia , Feminino , Humanos , Masculino , Pessoa de Meia-Idade , Seio Aórtico/diagnóstico por imagem , Resultado do Tratamento , Adulto JovemRESUMO
Objectives: Coarctation of the aorta (CoA) in adolescents and adults is often complicated by other cardiac or aortic problems and may carry additional surgical difficulties. Limited studies have reported the surgical outcomes of CoA repair in this particular patient population. We reviewed our contemporary experience of open surgical management of CoA in adolescents and adults. Methods: From 2008 to 2016, a total of 60 adolescents and adults (mean age, 32 years) underwent CoA repair at our centre. Of this cohort, 43 patients only underwent CoA repair (isolated group), while the remaining 17 underwent other concomitant cardiac procedures (concomitant group). Ascending-to-descending aortic bypass was the preferred repair technique in the concomitant group. Aortic valve replacement and a Bentall/David procedure were the two most frequently combined procedures. Results: There were no in-hospital deaths in the isolated group compared with three in the concomitant group (0% vs 17.6%, P = 0.029). The median intensive care unit stay (1 vs 2 days, P <0.01) and postoperative hospital stay (6 vs 9 days, P <0.01) were significantly longer in the concomitant group. Postoperative respiratory failure occurred more frequently in the concomitant group (0% vs 17.6%, P = 0.029). The isolated group had better overall survival during follow-up ( P = 0.037). The 5-year overall survival rate was 98% in the isolated group and 82% in the concomitant group. Conclusions: Surgical management of coarctation in adolescents and adults can achieve similar satisfactory outcomes as their neonatal counterparts, but the complexity of the concomitant procedures is a risk factor for this particular patient population.