ABSTRACT
Objective:To observe the diagnostic value of six classification intelligent auxiliary diagnosis lightweight model for common fundus diseases based on fundus color photography.Methods:A applied research. A dataset of 2 400 color fundus images from Nanjing Medical University Eye Hospital and Zhejiang Mathematical Medical Society Smart Eye Database was collected, which was desensitized and labeled by a fundus specialist. Of these, 400 each were for diabetic retinopathy, glaucoma, retinal vein occlusion, high myopia, age-related macular degeneration, and normal fundus. The parameters obtained from the classical classification models VGGNet16, ResNet50, DenseNet121 and lightweight classification models MobileNet3, ShuffleNet2, GhostNet trained on the ImageNet dataset were migrated to the six-classified common fundus disease intelligent aid diagnostic model using a migration learning approach during training as initialization parameters for training to obtain the latest model. 1 315 color fundus images of clinical patients were used as the test set. Evaluation metrics included sensitivity, specificity, accuracy, F1-Score and agreement of diagnostic tests (Kappa value); comparison of subject working characteristic curves as well as area under the curve values for different models.Result:Compared with the classical classification model, the storage size and number of parameters of the three lightweight classification models were significantly reduced, with ShuffleNetV2 having an average recognition time per sheet 438.08 ms faster than the classical classification model VGGNet16. All 3 lightweight classification models had Accuracy > 80.0%; Kappa values > 70.0% with significant agreement; sensitivity, specificity, and F1-Score for the diagnosis of normal fundus images were ≥ 98.0%; Macro-F1 was 78.2%, 79.4%, and 81.5%, respectively.Conclusion:The intelligent assisted diagnosis of common fundus diseases based on fundus color photography is a lightweight model with high recognition accuracy and speed; the storage size and number of parameters are significantly reduced compared with the classical classification model.
ABSTRACT
ABSTRACT Huge amounts of fly ash - a substance that does not conform to the ASTM C618 classification due to its chemical properties - have been abandoned in landfills around the world, despite their self-cementing property. It has not been used in concrete making applications due to its large amounts of free lime and sulfate contents. The fly ash in these plants is dumped in landfills, causing serious environmental hazards. Fly ash is disposed to the landfills by belt conveyors after being humidified with water. Therefore, the fly ashes humidified in the landfill areas are hydrated in nature. This hydration is further intensified in landfills by rain and snow. Thus, the free lime content of fly ash decreases due to its long hydration process. In this work, the lightweight masonry blocks were produced by mixing normal and hydrated fly ashes or normal, hydrated fly ash and lime without Portland cement. The compressive strength, water absorption, sorptivity, density, porosity, and thermal conductivity values of the samples produced were determined. The results obtained from these tests showed that lightweight masonry blocks could be produced by using these waste materials in building applications.
RESUMO Enormes quantidades de cinzas volantes - uma substância que não está de acordo com a classificação ASTM C618 devido às suas propriedades químicas - foram abandonadas em aterros sanitários ao redor do mundo, apesar de sua propriedade de autocimentação. Essas substâncias não têm sido usadas em aplicações de fabricação de concreto devido às suas grandes quantidades de cal livre e teores de sulfato. A cinza volante dessas usinas é despejada em aterros sanitários, causando sérios riscos ambientais. Essas cinzas são descartadas em aterros por correias transportadoras após serem umedecidas com água. Portanto, as cinzas volantes umedecidas nas áreas do aterro são hidratadas na natureza. Essa hidratação é ainda mais intensificada em aterros, pela chuva e pela neve. Assim, o teor de cal livre nas cinzas volantes diminui devido ao longo processo de hidratação. Neste trabalho, blocos de alvenaria leves foram produzidos pela mistura de cinzas volantes normais e hidratadas, ou cinza volante normal hidratada e cal sem cimento Portland. Foram determinados os valores de resistência à compressão, absorção de água, sensibilidade, densidade, porosidade e condutividade térmica das amostras produzidas. Os resultados obtidos nesses testes mostraram que blocos de alvenaria leves podem ser produzidos usando esses materiais residuais em aplicações de construção.
ABSTRACT
Objective:To evaluate the application of a multi-task learning-based light-weight convolution neural network (MTLW-CNN) for the automatic segmentation of organs at risk (OARs) in thorax.Methods:MTLW-CNN consisted of several layers for sharing features and 3 branches for segmenting 3 OARs. 497 cases with thoracic tumors were collected. Among them, the computed tomography (CT) images encompassing lung, heart and spinal cord were included in this study. The corresponding contours delineated by experienced radiation oncologists were ground truth. All cases were randomly categorized into the training and validation set ( n=300) and test set ( n=197). By applying MTLW-CNN on the test set, the Dice similarity coefficients (DSCs) of 3 OARs, training and testing time and space complexity (S) were calculated and compared with those of Unet and DeepLabv3+ . To evaluate the effect of multi-task learning on the generalization performance of the model, 3 single-task light-weight CNNs (STLW-CNNs) were built. Their structures were totally the same as the corresponding branches in MTLW-CNN. After using the same data and algorithm to train STLW-CNN, the DSCs were statistically compared with MTLW-CNN on the testing set. Results:For MTLW-CNN, the averages (μ) of lung, heart and spinal cord DSCs were 0.954, 0.921 and 0.904, respectively. The differences of μ between MTLW-CNN and other two models (Unet and DeepLabv3+ ) were less than 0.020. The training and testing time of MTLW-CNN were 1/3 to 1/30 of that of Unet and DeepLabv3+ . S of MTLW-CNN was 1/42 of that of Unet and 1/1 220 of that of DeepLabv3+ . The differences of μ and standard deviation (σ) of lung and heart between MTLW-CNN and STLW-CNN were approximately 0.005 and 0.002. The difference of μ of spinal cord was 0.001, but σof STLW-CNN was 0.014 higher than that of MTLW-CNN.Conclusions:MTLW-CNN spends less time and space on high-precision automatic segmentation of thoracic OARs. It can improve the application efficiency and generalization performance of the models.
ABSTRACT
Objective:To design a multimodality wheelchair in order to meet many use requirements of the lower limbs dyskinesia patients. Methods:For some shortcomings of the existing multimodality wheelchair such as heavy structure and drive redundancy, the design of multimodality wheelchair based on double parallelogram mechanism is studied, and a simple and reliable multimodality wheelchair mechanism is presented. Rotary actuation drives the double parallelogram mechanism and makes parts of the wheelchair move, and three posture transformations are realized. The coupling movement of the back and leg rest realizes the transformation of sitting and supine posture, and the coupling movement of seat and back realizes the change of sitting and standing posture. The mechanical analysis of wheelchairs based on uniform load was carried out, and the rotation moment is calculated. Results:When the driving torque changed from sitting position to supine position, it increased with the increase of rotation angle and reached a peak value. When the driving torque changed from sitting position to standing position, it gradually decreased with the increase of rotation angle. Then the dynamics simulation is carried out in Adams software, the calculated results were basically consistent with the simulation results. Conclusion:The proposed mechanism is helpful to realize the lightweight design of multimodality wheelchair, and meets the design needs of intelligent multimodality wheelchair.
ABSTRACT
Objective To study the application value of lightweight mesh in the ventral hernia tension-free hernia repair in middle-age and elderly patients. Methods The retrospective cohort study method was used to analyze the clinical data and followed information about 284 cases of tension-free hernia repair in our hospital. The patients were randomly divid-ed into experimental group and control group. The lightweight mesh and heavyweight mesh were used in experimental group and control group respectively. The information was collected including age, sex, inguinal hernia type, combined chronic dis-ease, operative time, blood loss, the duration of hospitalization, surgical expenses and other indicators. Data were followed up in patients including postoperative chronic pain, limited mobility, surgical numbness, the recurrent, foreign body sensation. Results The average cost of surgery was higher in experimental group[ (9 404.54±314.87) yuan]than that in control group [(7 613.94±329.68) yuan]. There were no significant differences in the operative time, blood loss and length of hospital stay between two groups. Results of follow-up data showed that the early postoperative chronic pain, early and long-term activity limitation, long-term surgery area numbness and foreign body sensation were lower in experimental group than those in con-trol group. Conclusion The cost of lightweight mesh was higher than that of heavyweight mesh in the ventral tension-free hernia repair therapy. It is worth to apply the lightweight mesh, which can reduce the incidence of postoperative pain, numb-ness and other adverse reactions.
ABSTRACT
Objective:To establish hospital infection data sharing and service platform for collecting and issuing the hospital infection data. Methods:The relational database was used to realize data storage and J2EE light structure was employed to enhance the development and the operating efficiency and flexibility of the system according to the principles of Web information system.Results and Conclusion:An open and sharing hospital infection data platform for data processing, retrieving and analyzing was established. The established hospital infection data set and surgery infection data set contained more than 20 000 records. The system provided an effective software platform support for the utilization of hospital infection data and the prevention and control of hospital infection. This article introduces key technologies of the design and realization of the system.
ABSTRACT
This paper describes an inexpensive and secure cage system for housing venomous snakes. The cages are easily constructed from commercially available plastic containers and are lightweight and can be stacked, minimizing the area needed to house numerous animals. They allow easy access to the animal and can be adequately disinfected. These cages can be individually locked and also allow for full viewing of the animal.