Notes

Perforator-Plus Flap throughout Lower Extremity Renovation.
There are two challenges associated with the interpretability of deep learning models in medical image analysis applications that need to be addressed confidence calibration and classification uncertainty. Confidence calibration associates the classification probability with the likelihood that it is actually correct - hence, a sample that is classified with confidence X% has a chance of X% of being correctly classified. Classification uncertainty estimates the noise present in the classification process, where such noise estimate can be used to assess the reliability of a particular classification result. Both confidence calibration and classification uncertainty are considered to be helpful in the interpretation of a classification result produced by a deep learning model, but it is unclear how much they affect classification accuracy and calibration, and how they interact. In this paper, we study the roles of confidence calibration (via post-process temperature scaling) and classification uncertainty (computed either from classification entropy or the predicted variance produced by Bayesian methods) in deep learning models. Results suggest that calibration and uncertainty improve classification interpretation and accuracy. This motivates us to propose a new Bayesian deep learning method that relies both on calibration and uncertainty to improve classification accuracy and model interpretability. Experiments are conducted on a recently proposed five-class polyp classification problem, using a data set containing 940 high-quality images of colorectal polyps, and results indicate that our proposed method holds the state-of-the-art results in terms of confidence calibration and classification accuracy. For simultaneous positron-emission-tomography and magnetic-resonance-imaging (PET-MRI) systems, while early methods relied on independently reconstructing PET and MRI images, recent works have demonstrated improvement in image reconstructions of both PET and MRI using joint reconstruction methods. The current state-of-the-art joint reconstruction priors rely on fine-scale PET-MRI dependencies through the image gradients at corresponding spatial locations in the PET and MRI images. In the general context of image restoration, compared to gradient-based models, patch-based models (e.g., sparse dictionaries) have demonstrated better performance by modeling image texture better. Thus, we propose a novel joint PET-MRI patch-based dictionary prior that learns inter-modality higher-order dependencies together with intra-modality textural patterns in the images. We model the joint-dictionary prior as a Markov random field and propose a novel Bayesian framework for joint reconstruction of PET and accelerated-MRI images, using expectation maximization for inference. We evaluate all methods on simulated brain datasets as well as on in vivo datasets. MS4078 mw We compare our joint dictionary prior with the recently proposed joint priors based on image gradients, as well as independently applied patch-based priors. Our method demonstrates qualitative and quantitative improvement over the state of the art in both PET and MRI reconstructions. A series of liposome ligands (Bio-Chol, Bio-Bio-Chol, tri-Bio-Chol and tetra-Bio-Chol) modified by different branched biotins that can recognize the SMVT receptors over-expressed in breast cancer cells were synthesized. And four liposomes (Bio-Lip, Bio-Bio-Lip, tri-Bio-Lip and tetra-Bio-Lip) modified by above mentioned ligands as well as the unmodified liposome (Lip) were prepared to study the targeting ability for breast cancer. The cytotoxicity study and apoptosis assay of paclitaxel-loaded liposomes showed that tri-Bio-Lip had the strongest anti-proliferative effect on breast cancer cells. The cellular uptake studies on mice breast cancer cells (4T1) and human breast cancer cells (MCF-7) indicated tri-Bio-Lip possessed the strongest internalization ability, which was 5.21 times of Lip, 2.60 times of Bio-Lip, 1.67 times of Bio-Bio-Lip and 1.17 times of tetra-Bio-Lip, respectively. Moreover, the 4T1 tumor-bearing BALB/c mice were used to evaluate the in vivo targeting ability. The data showed the enrichment of liposomes at tumor sites were tri-Bio-Lip > tetra-Bio-Lip > Bio-Bio-Lip > Bio-Lip > Lip, which were consistent with the results of in vitro targeting studies. In conclusion, increasing the density of targeting molecules on the surface of liposomes can effectively enhance the breast cancer targeting ability, and the branching structure and spatial distance of biotin residues may also have an important influence on the affinity to SMVT receptors. Therefore, tri-Bio-Lip could be a promising drug delivery system for targeting breast cancer. After spinal cord injury (SCI), endogenous neural/progenitor stem cells (NSPCs) were activated in neural tissue adjacent to the injured segment, but few cells migrated to the injury epicenter and differentiated into neurons. N-cadherin regulates mechanical adhesion between NSPCs, and also drives NSPCs migration and promotes NSPCs differentiation. In this study, linearly ordered collagen scaffold (LOCS) was modified with N-cadherin through a two-step cross-linking between thiol and amino group. The results indicated that N-cadherin modification improved the adhesion of NSPCs on collagen scaffold and increased the differentiation into neurons. When LOCS-Ncad was transplanted into complete transected rat spinal cords, more NSPCs migrated to the lesion center and more newborn neurons appeared within the injury site. Furthermore, rats transplanted with LOCS-Ncad showed significantly improved locomotor recovery compared with the rats without implants. Collectively, our results suggest that LOCS-Ncad may be a promising treatment option to facilitate SCI repair by recruiting endogenous NSPCs to the lesion center and promoting neuronal differentiation. Stochastic optical reconstruction microscopy (STORM) is a promising method for the visualization of ultra-fine mitochondrial structures. However, this approach is limited to monitoring dynamic intracellular events owing to its low temporal resolution. We developed a new strategy to capture mitochondrial dynamics using a compressed sensing STORM algorithm following raw data pre-treatments by a noise-corrected principal component analysis and K-factor image factorization. Using STORM microscopy with a vicinal-dithiol-proteins targeting probe, visualizing mitochondrial dynamics was attainable with spatial and temporal resolutions of 45 nm and 0.8 s, notably, dynamic mitochondrial tubulation retraction of ~746 nm in 1.2 s was monitored. The labeled conjugate was observed as clusters (radii, ~90 nm) distributed on the outer mitochondrial membranes, not yet reported as far as we know. This strategy is promising for the quantitative analysis of intracellular behaviors below the optical diffraction limit.
Read More: https://www.selleckchem.com/products/ms4078.html