直肠癌淋巴结转移的深度神经网络评判研究进展

正目前,临床上对于直肠癌常用的影像评估方法有MRI、螺旋CT、PET-CT、直肠腔内超声(ERUS)等。而MRI作为首选检查方式,对肿瘤位置、浸润深度、淋巴结转移、血管侵犯、环周切缘及周围器官侵犯等方面的评估均具有明显优势~([1-2])。通过MRI诊断淋巴结的方法通常是影像科医师逐层浏览每一幅图像,从中识别淋巴结的形状、界限及密度来判断,这种传统方式耗时较长且存在主观偏倚,导致     

Rapid Increase in Neural Conduction Time in the Adult Human Auditory Brainstem Following Sudden Unilateral Deafness

Individuals with sudden unilateral deafness offer a unique opportunity to study plasticity of the binaural auditory system in adult humans. Stimulation of the intact ear results in increased activity in the auditory cortex. However, there are no reports of changes at sub-cortical levels in humans. Therefore, the aim of the present study was to investigate changes in sub-cortical activity immediately before and after the onset of surgically induced unilateral deafness in adult humans. Click-evoked auditory brainstem responses (ABRs) to stimulation of the healthy ear were recorded from ten adults during the course of translabyrinthine surgery for the removal of a unilateral acoustic neuroma. This surgical technique always results in abrupt deafferentation of the affected ear. The results revealed a rapid (within minutes) reduction in latency of wave V (mean pre = 6.55 ms; mean post = 6.15 ms; p < 0.001). A latency reduction was also observed for wave III (mean pre = 4.40 ms; mean post = 4.13 ms; p < 0.001). These reductions in response latency are consistent with functional changes including disinhibition or/and more rapid intra-cellular signalling affecting binaurally sensitive neurons in the central auditory system. The results are highly relevant for improved understanding of putative physiological mechanisms underlying perceptual disorders such as tinnitus and hyperacusis.     

Engraftable neural crest stem cells derived from cynomolgus monkey embryonic stem cells

Neural crest stem cells (NCSCs), a population of multipotent cells that migrate extensively and give rise to diverse derivatives, including peripheral and enteric neurons and glia, craniofacial cartilage and bone, melanocytes and smooth muscle, have great potential for regenerative medicine. Non-human primates provide optimal models for the development of stem cell therapies. Here, we describe the first derivation of NCSCs from cynomolgus monkey embryonic stem cells (CmESCs) at the neural rosette stage. CmESC-derived neurospheres replated on polyornithine/laminin-coated dishes migrated onto the substrate and showed characteristic expression of NCSC markers, including Sox10, AP2α, Slug, Nestin, p75, and HNK1. CmNCSCs were capable of propagating in an undifferentiated state in vitro as adherent or suspension cultures, and could be subsequently induced to differentiate towards peripheral nervous system lineages (peripheral sympathetic neurons, sensory neurons, and Schwann cells) and mesenchymal lineages (osteoblasts, adipocytes, chondrocytes, and smooth muscle cells). CmNCSCs transplanted into developing chick embryos or fetal brains of cynomolgus macaques survived, migrated, and differentiated into progeny consistent with a neural crest identity. Our studies demonstrate that CmNCSCs offer a new tool for investigating neural crest development and neural crest-associated human disease and suggest that this non-human primate model may facilitate tissue engineering and regenerative medicine efforts.     

A Multi-Scale DNN Algorithm for Nonlinear Elliptic Equations with Multiple Scales

Algorithms based on deep neural networks (DNNs) have attracted increasing attention from the scientific computing community. DNN based algorithms are easy to implement, natural for nonlinear problems, and have shown great potential to overcome the curse of dimensionality. In this work, we utilize the multi-scale DNN-based algorithm (MscaleDNN) proposed by Liu, Cai and Xu (2020) to solve multi-scale elliptic problems with possible nonlinearity, for example, the p-Laplacian problem. We improve the MscaleDNN algorithm by a smooth and localized activation function. Several numerical examples of multi-scale elliptic problems with separable or non-separable scales in low-dimensional and high-dimensional Euclidean spaces are used to demonstrate the effectiveness and accuracy of the MscaleDNN numerical scheme.     

Convergence and attractivity of memristor-based cellular neural networks with time delays

This paper presents theoretical results on the convergence and attractivity of memristor-based cellular neural networks (MCNNs) with time delays. Based on a realistic memristor model, an MCNN is modeled using a differential inclusion. The essential boundedness of its global solutions is proven. The state of MCNNs is further proven to be convergent to a critical-point set located in saturated region of the activation function, when the initial state locates in a saturated region. It is shown that the state convergence time period is finite and can be quantitatively estimated using given parameters. Furthermore, the positive invariance and attractivity of state in non-saturated regions are also proven. The simulation results of several numerical examples are provided to substantiate the results.     

Comparison of diagnostic usefulness of predictive models in preliminary differentiation of adnexal masses

The purpose of this study was to compare prognostic models evaluating the probability of an ovarian cancer occurrence based on a number of clinical and ultrasonographic data in women with adnexal masses. A total of 686 women with adnexal masses underwent the examinations between 1994 and 2002. The recorded parameters included: age, menopausal status, body mass index, the grayscale and Doppler ultrasonographic examination, and selected markers concentration levels. In order to find the best combination of features, which significantly influences the probability of malignancy, stepwise logistic regression analysis, as well as artificial neural network, was used. The diagnostic efficiency of received models was estimated and compared using receiver-operating characteristics (ROC) curve. The results indicate that 431 and 255 patients had a benign and malignant ovarian tumor, respectively. Application of stepwise logistic regression analysis revealed statistically significant importance of eight features. The sensitivity and specificity for the received model were 65.71% and 77.59%, respectively. Three-layer perceptron network shows 13 features as significant predictors of malignancy. The network gave a sensitivity of 85.7% and specificity of 93.1%. Comparison of area under ROC curve for received models was 0.9679 vs 0.9716. Prognostic values of the analyzed neural model are not optimal but seem to surpass logistic regression model in terms of the predictive possibilities.