Jakobsenduelund0778

The results of the feature importance assessment show that the CNN model makes predictions using clinically meaningful and appropriate features. Finally, we implemented counterfactual explanations for the CNN model. This can help clinicians understand what feature changes for a particular patient would lead to a desirable outcome, i.e. readiness to extubate.Electronic health records (EHR) contain large volumes of unstructured text, requiring the application of information extraction (IE) technologies to enable clinical analysis. We present the open source Medical Concept Annotation Toolkit (MedCAT) that provides (a) a novel self-supervised machine learning algorithm for extracting concepts using any concept vocabulary including UMLS/SNOMED-CT; (b) a feature-rich annotation interface for customizing and training IE models; and (c) integrations to the broader CogStack ecosystem for vendor-agnostic health system deployment. We show improved performance in extracting UMLS concepts from open datasets (F10.448-0.738 vs 0.429-0.650). Further real-world validation demonstrates SNOMED-CT extraction at 3 large London hospitals with self-supervised training over ∼8.8B words from ∼17M clinical records and further fine-tuning with ∼6K clinician annotated examples. We show strong transferability (F1 > 0.94) between hospitals, datasets and concept types indicating cross-domain EHR-agnostic utility for accelerated clinical and research use cases.While Deep Learning (DL) is often considered the state-of-the art for Artificial Intel-ligence-based medical decision support, it remains sparsely implemented in clinical practice and poorly trusted by clinicians due to insufficient interpretability of neural network models. We have approached this issue in the context of online detection of epileptic seizures by developing a DL model from EEG signals, and associating certain properties of the model behavior with the expert medical knowledge. This has conditioned the preparation of the input signals, the network architecture, and the post-processing of the output in line with the domain knowledge. TNO155 Specifically, we focused the discussion on three main aspects (1) how to aggregate the classification results on signal segments provided by the DL model into a larger time scale, at the seizure-level; (2) what are the relevant frequency patterns learned in the first convolutional layer of different models, and their relation with the delta, theta, alpha, beta and gamma frequency bands on which the visual interpretation of EEG is based; and (3) the identification of the signal waveforms with larger contribution towards the ictal class, according to the activation differences highlighted using the DeepLIFT method. Results show that the kernel size in the first layer determines the interpretability of the extracted features and the sensitivity of the trained models, even though the final performance is very similar after post-processing. Also, we found that amplitude is the main feature leading to an ictal prediction, suggesting that a larger patient population would be required to learn more complex frequency patterns. Still, our methodology was successfully able to generalize patient inter-variability for the majority of the studied population with a classification F1-score of 0.873 and detecting 90% of the seizures.Transparent gas barrier materials have extensive applications in packaging, pharmaceutical preservation, and electronics. Herein, we designed transparent films with a symmetric sandwich structure using layer-by-layer assembly of biaxially oriented polypropylene (BOPP) and acrylic resin (AR) followed by a cellulose nanoparticle (CNP) layer. The BOPP as a substrate created a barrier to hinder the transmission of water molecules to the adhesive AR layer and gas barrier functional CNP layer. The aspect ratio of the CNPs was shown to affect the film microstructure, resulting in different values for the oxygen transmission rate (OTR). The well-organized CNP layer exhibited lower OTR when compared with the network layer. The thickness, density, and porosity of the CNP layer exhibited correlations with OTR. The water molecules were able to flow in through an additional pathway, thus increasing the water vapor transmission rate (WVTR). Moreover, these sandwiched cellulose composite films showed excellent light transmittance and tensile strength.Nitrocellulose (NC) membrane can have value-added applications for lateral flow assay (LFA)-based diagnostic tools, which has great potential for the detection of pathogens, such as COVID-19, in different environments. However, poor sensitivity of the NC membrane based LFA limits its further application in many cases. Herein, we developed a facile method for LFA sensitivity enhancement, by incorporating two-sugar barrier into LFAs one between the conjugation pad and the test line, and the other between the test line and the control line. ORF1ab nucleic acid of COVID-19 was used as the model target to demonstrate the concept on the HF120 membrane. Results show that at optimum conditions, the two sugar barrier LFAs have a detection limit of 0.5 nM, which is compared to that of 2.5 nM for the control LFA, achieving a 5-fold sensitivity increase. This low cost, easy-to-fabricate and easy-to-integrate LFA method may have potential applications in other cellulose paper-based platforms.Multifunctional theranostic nanoplatforms integrated of imaging function, multi-modality therapy, stimuli-responsiveness, and targeted delivery are of highly desirable attributes in achieving precise medicine. However, preparation of multifunctional nanoplatforms often involves laborious, multiple steps and inevitably utilizes low-biocompatible or non-functional components. Herein we report a facile, one-step self-assembly strategy to fabricate hyaluronic acid (HA)-based multifunctional tumor theranostic nanoplatform by employing magnetic resonance imaging (MRI) agent Mn2+ as a reversible crosslink agent for histidine-grafted HA, along with simultaneously loading chemotherapeutic agent doxorubicin hydrochloride (DOX) and photodynamic therapy agent chlorin e6, to realize MRI-guided targeted chemo-photodynamic cancer therapy. The targeted delivery and stimuli-responsive payload release were demonstrated in vitro and in vivo. Furthermore, the combined chemo-photodynamic therapy of the nanoassembly dramatically improved the cancer therapeutic outcome, in comparison with that of free DOX and nanoplatform solely loaded DOX in a melanoma bearing mice. Our one step assemble strategy is of great potential in clinic transformation.Inspired by the natural electrostatic interaction of cationic growth factors with anionic sulfated glycosaminoglycans in the extracellular matrix, we developed electrospun poly(hydroxybutyrate)/gelatin (PG) fibers conjugated with anionic sulfated carboxymethylcellulose (sCMC) to enable growth factor immobilization via electrostatic interaction for tissue engineering. The fibrous scaffold bound cationic molecules, was cytocompatible and exhibited a remarkable morphological and functional stability. Transforming growth factor-β1 immobilized on the sCMC conjugated fibers was retained for at least 4 weeks with negligible release (3%). Immobilized fibroblast growth factor-2 and connective tissue growth factor were bioactive and induced proliferation and fibrogenic differentiation of infrapatellar fat pad derived mesenchymal stem cells respectively with efficiency similar to or better than free growth factors. Taken together, our studies demonstrate that sCMC conjugated PG fibers can immobilize and retain function of cationic growth factors and hence show potential for use in various tissue engineering applications.Atherosclerotic cardiovascular disease became one of the major causes of morbidity and mortality worldwide. As a sulfated polysaccharide with anti-inflammatory and hypolipidemic activities, fucoidan can induce autophagy. We show here that fucoidan reduces lipid accumulation in foam cells, which is one of the causes of atherosclerosis. Further studies show that fucoidan promotes autophagy showed by the expression of p62/SQSTM1 and microtubule-associated protein light chain 3 (LC3) II, which can be blocked by autophagy inhibitors 3-MA and bafilomycin A1. In addition, the expression of transcription factor EB (TFEB), master regulator of autophagy and lysosome function, is upregulated after the treatment with fucoidan. Moreover, the knockout of TFEB with small interfering RNA suppressed the effect of fucoidan. Together, fucoidan reduces lipid accumulation in foam cells by enhancing autophagy through the upregulation of TFEB. In view of the role of foam cells in atherosclerosis, fucoidan can be valuable for the treatment of atherosclerosis.The inferior tendon healing after surgery is inextricably linked to the surgical suture. Poor load transfer along the suture often results in a high tendon re-tear rate. Besides, the severe inflammation and infection induced by sutures even cause a second surgery. Herein, to alleviate the above-mentioned issues, a multifunctional suture was fabricated by decorating chitosan/gelatin-tannic acid (CS/GE-TA) on the porous tape suture. The porous tape suture ensured the required mechanical properties and sufficient space for tissue integration. Compared to the pristine suture, the CS/GE-TA decorated suture (TA100) presented a 332% increase in pull-out force from the tendon, indicating potentially decreased re-tear rates. Meanwhile, TA100 showed superior anti-inflammatory and antibacterial performances. In vivo experiments further proved that TA100 could not only reduce inflammatory action but also facilitate collagen deposition and blood vessel formation. These results indicate that the multifunctional sutures are promising candidates for accelerating tendon healing.Self-healing hydrogels with pH-responsiveness could protect loaded drugs from being destroyed till it arrives to the target. The pectin-based hydrogel is a candidate due to the health benefit, anti-inflammation, antineoplastic activity, nontoxicity, and biospecific degradation, et al. However, the abundant existence of water-soluble branched heteropolysaccharide chains influenced its performance resulting in limitation of the potential. In the present study, we prepared a series of self-healing pectin/chitosan hydrogels via the Diels-Alder reaction. Moreover, pectin/chitosan composite hydrogel was prepared as a contrast. By comparison, it can be seen that the Diels-Alder reaction greatly improved the cross-linking density of hydrogels. The self-healing experiments showed excellent self-healing performance. In different swelling mediums, significant transformation in the swelling ratio was shown, indicating well-swelling property, pH- and thermo-responsiveness. The drug loading and release studies presented high loading efficiency and sustained release performance. The cytotoxicity assay that showed a high cell proliferation ratio manifested great cytocompatibility.Polylactide (PLA) nanocomposites with spray-and freeze-dried cellulose nanocrystals (i.e., SCNC and FCNC) were prepared through solution casting using four different solvents tetrahydrofuran (THF), chloroform (CHL), dimethylformamide (DMF), and dimethyl sulfoxide (DMSO). Small amplitude oscillatory shear rheological analysis was extensively employed to explore the CNC dispersion quality in PLA. Overall, the rheological properties differences of PLA/SCNC and PLA/FCNC nanocomposites were not very significant. Moreover, the use of THF and CHL did not lead to a proper dispersion of CNCs in PLA due to their low dielectric constants. On the other hand, while the use of DMF was effective on the enhancement of CNC dispersion, DMSO could more dramatically lead to such enhancement due to its higher dielectric constant. The percolation threshold in PLA/SCNC nanocomposites prepared with DMF and DMSO was predicted around 1.52 and 0.12 wt% CNC, respectively. The crystallization behavior of PLA/nanocomposites prepared with DMF and DMSO were also explored.