Mapping anatomical brain networks with graph-theoretic analysis of diffusion tractography has recently gained popularity, because of its presumed value in understanding brain function. However, this approach has seldom been used to compare brain connectomes across species, which may provide insights into brain evolution. Here, we employed a data-driven approach to compare interregional brain connections across three primate species: 1) the intensively studied rhesus macaque, 2) our closest living primate relative, the chimpanzee, and 3) humans. Specifically, we first used random parcellations and surface-based probabilistic diffusion tractography to derive the brain networks of the three species under various network densities and resolutions. We then compared the characteristics of the networks using graph-theoretic measures. In rhesus macaques, our tractography-defined hubs showed reasonable overlap with hubs previously identified using anterograde and retrograde tracer data. Across all three species, hubs were largely symmetric in the two hemispheres and were consistently identified in medial parietal, insular, retrosplenial cingulate and ventrolateral prefrontal cortices, suggesting a conserved structural architecture within these regions. However, species differences were observed in the inferior parietal cortex, polar and medial prefrontal cortices. The potential significance of these interspecies differences is discussed.

Distilling knowledge from convolutional neural networks (CNNs) is a double-edged sword for vision transformers (ViTs). It boosts the performance since the imagefriendly local-inductive bias of CNN helps ViT learn faster and better, but leading to two problems: (1) Network designs of CNN and ViT are completely different, which leads to different semantic levels of intermediate features, making spatial-wise knowledge transfer methods (e.g., feature mimicking) inefficient. (2) Distilling knowledge from CNN limits the network convergence in the later training period since ViT's capability of integrating global information is suppressed by CNN's local-inductive-bias supervision.To this end, we present Cumulative Spatial Knowledge Distillation (CSKD). CSKD distills spatial-wise knowledge to all patch tokens of ViT from the corresponding spatial responses of CNN, without introducing intermediate features. Furthermore, CSKD exploits a Cumulative Knowledge Fusion (CKF) module, which introduces the global response of CNN and increasingly emphasizes its importance during the training. Applying CKF leverages CNN's local inductive bias in the early training period and gives full play to ViT's global capability in the later one. Extensive experiments and analysis on ImageNet-1k and downstream datasets demonstrate the superiority of our CSKD. Code will be publicly available.

BackgroundSkip metastasis is a special type in cervical lymph node metastasis (LNM) of patients diagnosed with papillary thyroid carcinoma (PTC) which induced poor prognosis. There are few studies about skip metastasis and conclusions remained uncertain. Therefore, this study aims to explore the frequency and to investigate risk factors of skip metastasis in PTC.MethodsThrough searching the keyword by PubMed and Embase databases which articles published up to 1st August 2018 about skip metastasis in papillary thyroid carcinoma, we extract data in order to assure whether those materials meet the criteria.ResultsThe prevalence of skip metastasis is 12.02% in light of our meta-analysis of 18 studies with 2165 patients. The upper pole location (RR = 3.35, 95% CI =1.65–6.79, P = 0.0008) and tumors size ≤1 cm (RR = 2.65, 95% CI =1.50–4.70, P = 0.0008) are significantly associated with skip metastasis, whereas lymphovascular invasion (RR = 0.33, 95% CI =0.15–0.75, P = 0.0083) exists lower rate of skip metastasis. Multifocality, gender, age, bilaterality, thyroiditis and Extrathyroidal extension (ETE) are insignificantly associated with skip metastasis. Level II and level III are the most frequently affected areas.ConclusionThe lateral compartment should be carefully examined especially for those PTC patients who present primary tumors in the upper lobe with a primary tumor size ≤10 mm which could be detected with skip metastasis.

Introduction The aim of the study was to systematically review relevant studies to evaluate the diagnostic value of urinary kidney injury molecule 1 (uKIM-1) for acute kidney injury (AKI) in adults. Method We searched PubMed and Embase for literature published up to November 1st, 2019 and used the Quality Assessment Tool for Diagnosis Accuracy Studies (QUADAS-2) to assess the quality. Then, we extracted useful information from each eligible study and pooled sensitivity, specificity, and area under the curve (AUC) values. Results A total of 14 studies with 3300 patients were included. The estimated sensitivity of urinary KIM-1 (uKIM-1) in the diagnosis of AKI was 0.74 (95% CrI 0.62–0.84), and the specificity was 0.84 (95% CrI, 0.76–0.90). The pooled diagnostic odds ratio (DOR) was 15.22 (95% CrI, 6.74–42.20), the RD was 0.55 (95% CrI 0.43–0.70), and the AUC of uKIM-1 in diagnosing AKI was 0.62 (95% CrI 0.41–0.76). The results of the subgroup analysis showed the influence of different factors. Conclusion Urinary KIM-1 is a good predictor for AKI in adult patients with relatively high sensitivity and specificity. However, further research and clinical trials are still needed to confirm whether and how uKIM-1 can be commonly used in clinical diagnosis.