RESUMEN
Glomerulus morphology on renal pathology images provides valuable diagnosis and outcome prediction information. To provide better care, an efficient, standardized, and scalable method is urgently needed to optimize the time-consuming and labor-intensive interpretation process by renal pathologists. This paper proposes a deep convolutional neural network (CNN)-based approach to automatically detect and classify glomeruli with different stains in renal pathology images. In the glomerulus detection stage, this paper proposes a flattened Xception with a feature pyramid network (FX-FPN). The FX-FPN is employed as a backbone in the framework of faster region-based CNN to improve glomerulus detection performance. In the classification stage, this paper considers classifications of five glomerulus morphologies using a flattened Xception classifier. To endow the classifier with higher discriminability, this paper proposes a generative data augmentation approach for patch-based glomerulus morphology augmentation. New glomerulus patches of different morphologies are generated for data augmentation through the cycle-consistent generative adversarial network (CycleGAN). The single detection model shows the F1 score up to 0.9524 in H&E and PAS stains. The classification result shows that the average sensitivity and specificity are 0.7077 and 0.9316, respectively, by using the flattened Xception with the original training data. The sensitivity and specificity increase to 0.7623 and 0.9443, respectively, by using the generative data augmentation. Comparisons with different deep CNN models show the effectiveness and superiority of the proposed approach.
Asunto(s)
Aprendizaje Profundo , Glomérulos Renales , Humanos , Glomérulos Renales/diagnóstico por imagen , Glomérulos Renales/patología , Redes Neurales de la Computación , Interpretación de Imagen Asistida por Computador/métodos , Enfermedades Renales/diagnóstico por imagen , Riñón/diagnóstico por imagen , Riñón/patología , Procesamiento de Imagen Asistido por Computador/métodosRESUMEN
BACKGROUND: Resting-state functional magnetic resonance imaging (rs-fMRI) is a promising method for the study of brain function. Typically, rs-fMRI is performed on anesthetized animals. Although different functional connectivity (FC) in various anesthetics on whole brain have been studied, few studies have focused on different FC in the aged brain. Here, we measured FC under three commonly used anesthesia methods and analyzed data to determine if the FC in whole brain analysis were similar among groups. METHODS: Twenty-four male aged Wistar rats were randomly divided into three groups (nâ=â8 in each group). Anesthesia was performed under either isoflurane (ISO), combined ISO + dexmedetomidine (DEX) or α-chloralose (AC) according to the groups. Data of rs-fMRI was analyzed by FC in a voxel-wise way. Differences in the FC maps between the groups were analyzed by one-way analysis of variance and post hoc two-sample t tests. RESULTS: Compared with ISO + DEX anesthesia, ISO anesthesia caused increased FC in posterior brain and decreased FC in the middle brain of the aged rat. AC anesthesia caused global suppression as no increase in FC was observed. CONCLUSION: ISO could be used as a substitute for ISO + DEX in rat default mode network studies if the left temporal association cortex is not considered important.
Asunto(s)
Anestesia , Isoflurano , Animales , Encéfalo/diagnóstico por imagen , Mapeo Encefálico , Imagen por Resonancia Magnética , Masculino , Ratas , Ratas WistarRESUMEN
In this paper, we integrate a cellular automaton model with game theory to simulate crowd evacuation from a room with consideration of obstacle removal. The room has one or more exits, one of which is blocked by obstacles. The obstacles at the exit can be removed by volunteers. We investigate the cooperative and defective behaviors of pedestrians during evacuation. The yielder game and volunteer's dilemma game are employed to resolve interpedestrian conflict. An anticipation floor field is proposed to guide the pedestrians to avoid obstacles that are being removed. We conducted experiments to determine how a variety of conditions affect overall crowd evacuation and volunteer evacuation times. The conditions were the start time of obstacle removal, number of obstacles, placement of obstacles, time spent in obstacle removal, strength of the anticipation floor field, and obstacle visibility distance. We demonstrate how reciprocity can be achieved among pedestrians and increases the efficiency of the entire evacuation process.
RESUMEN
OBJECTIVE: To investigate the apoptosis-inducing effect of honokiol on human non-Hodgkin lymphoma Raji cells and the possible mechanism. METHODS: Raji cells were treated with different concentrations of honokiol, and the proliferation of the cells was detected using MTT assay. Flow cytometry was employed to analyze the cell cycle changes and apoptosis of honokiol-treated cells. Caspase 8 activity in the cells was measured by caspase 8 kit, and RT-PCR was used to detect the expression of apoptosis-related genes Bcl-2, Bad, and Bax. RESULTS: Honokiol significantly inhibited the growth of Raji cells in a time- and dose-dependent manner, with IC(50) concentration of 17.53, 12.61, and 7.4 µg/ml at 12, 24, 48 h, respectively. Flow cytometry revealed cell cycle arrest at G0/G1 phase following honokiol treatment. The apoptosis rates of Raji cells treated with 7.5 and 15 µg/ml honokiol were significantly higher than that of the control cells [(18.24∓2.53)%, (28.44∓2.48)% vs (4.84∓1.15)%, P<0.01]. Caspase 8 activity in Raji cells was significantly enhanced by honokiol (P<0.05). The mRNA expression of the apoptosis-promoting gene Bad was significantly increased following honokiol treatment (P<0.01), while the expressions of Bcl-2 and Bax remained unchanged. CONCLUSION: Honokiol can induce apoptosis in Raji cells possibly in relation to enhancement of caspase 8 activity and Bad gene expression.