The nasal microbiota is a potential diagnostic biomarker for sepsis in critical care units.

This study aimed to characterize the composition of intestinal and nasal microbiota in septic patients and identify potential microbial biomarkers for diagnosis. A total of 157 subjects, including 89 with sepsis, were enrolled from the affiliated hospital. Nasal swabs and fecal specimens were collected from septic and non-septic patients in the intensive care unit (ICU) and Department of Respiratory and Critical Care Medicine. DNA was extracted, and the V4 region of the 16S rRNA gene was amplified and sequenced using Illumina technology. Bioinformatics analysis, statistical processing, and machine learning techniques were employed to differentiate between septic and non-septic patients. The nasal microbiota of septic patients exhibited significantly lower community richness (P = 0.002) and distinct compositions (P = 0.001) compared to non-septic patients. Corynebacterium, Staphylococcus, Acinetobacter, and Pseudomonas were identified as enriched genera in the nasal microbiota of septic patients. The constructed machine learning model achieved an area under the curve (AUC) of 89.08, indicating its efficacy in differentiating septic and non-septic patients. Importantly, model validation demonstrated the effectiveness of the nasal microecological diagnosis prediction model with an AUC of 84.79, while the gut microecological diagnosis prediction model had poor predictive performance (AUC = 49.24). The nasal microbiota of ICU patients effectively distinguishes sepsis from non-septic cases and outperforms the gut microbiota. These findings have implications for the development of diagnostic strategies and advancements in critical care medicine.IMPORTANCEThe important clinical significance of this study is that it compared the intestinal and nasal microbiota of sepsis with non-sepsis patients and determined that the nasal microbiota is more effective than the intestinal microbiota in distinguishing patients with sepsis from those without sepsis, based on the difference in the lines of nasal specimens collected.

Phosphoglycerate kinase (PGK) 1 succinylation modulates epileptic seizures and the blood-brain barrier.

Epilepsy is the most common chronic disorder in the nervous system, mainly characterized by recurrent, periodic, unpredictable seizures. Post-translational modifications (PTMs) are important protein functional regulators that regulate various physiological and pathological processes. It is significant for cell activity, stability, protein folding, and localization. Phosphoglycerate kinase (PGK) 1 has traditionally been studied as an important adenosine triphosphate (ATP)-generating enzyme of the glycolytic pathway. PGK1 catalyzes the reversible transfer of a phosphoryl group from 1, 3-bisphosphoglycerate (1, 3-BPG) to ADP, producing 3-phosphoglycerate (3-PG) and ATP. In addition to cell metabolism regulation, PGK1 is involved in multiple biological activities, including angiogenesis, autophagy, and DNA repair. However, the exact role of PGK1 succinylation in epilepsy has not been thoroughly investigated. The expression of PGK1 succinylation was analyzed by Immunoprecipitation. Western blots were used to assess the expression of PGK1, angiostatin, and vascular endothelial growth factor (VEGF) in a rat model of lithium-pilocarpine-induced acute epilepsy. Behavioral experiments were performed in a rat model of lithium-pilocarpine-induced acute epilepsy. ELISA method was used to measure the level of S100ß in serum brain biomarkers' integrity of the blood-brain barrier. The expression of the succinylation of PGK1 was decreased in a rat model of lithium-pilocarpine-induced acute epilepsy compared with the normal rats in the hippocampus. Interestingly, the lysine 15 (K15), and the arginine (R) variants of lentivirus increased the susceptibility in a rat model of lithium-pilocarpine-induced acute epilepsy, and the K15 the glutamate (E) variants, had the opposite effect. In addition, the succinylation of PGK1 at K15 affected the expression of PGK1 succinylation but not the expression of PGK1total protein. Furthermore, the study found that the succinylation of PGK1 at K15 may affect the level of angiostatin and VEGF in the hippocampus, which also affects the level of S100ß in serum. In conclusion, the mutation of the K15 site of PGK1 may alter the expression of the succinylation of PGK1 and then affect the integrity of the blood-brain barrier through the angiostatin / VEGF pathway altering the activity of epilepsy, which may be one of the new mechanisms of treatment strategies.

Cross-platform privacy-preserving CT image COVID-19 diagnosis based on source-free domain adaptation.

In the wake of the Coronavirus disease (COVID-19) pandemic, chest computed tomography (CT) has become an invaluable component in the rapid and accurate detection of COVID-19. CT scans traditionally require manual inspections from medical professionals, which is expensive and tedious. With advancements in machine learning, deep neural networks have been applied to classify CT scans for efficient diagnosis. However, three challenges hinder this application of deep learning: (1) Domain shift across CT platforms and human subjects impedes the performance of neural networks in different hospitals. (2) Unsupervised Domain Adaptation (UDA), the traditional method to overcome domain shift, typically requires access to both source and target data. This is not realistic in COVID-19 diagnosis due to the sensitivity of medical data. The privacy of patients must be protected. (3) Data imbalance may exist between easy/hard samples and between data classes which can overwhelm the training of deep networks, causing degenerate models. To overcome these challenges, we propose a Cross-Platform Privacy-Preserving COVID-19 diagnosis network (CP 3 Net) that integrates domain adaptation, self-supervised learning, imbalanced label learning, and rotation classifier training into one synergistic framework. We also create a new CT benchmark by combining real-world datasets from multiple medical platforms to facilitate the cross-domain evaluation of our method. Through extensive experiments, we demonstrate that CP 3 Net outperforms many popular UDA methods and achieves state-of-the-art results in diagnosing COVID-19 using CT scans.

Psychological stress recognition from heart rate variability parameters based on field programmable gate arrays.

Psychological stress is a big threat to people's health. Early detection of psychological stress is important. The design of a stress recognition device based on the ECG (electrocardiograph) signal is presented in this paper. The device features intelligence, precision, portability, fast response, and low power consumption. In the design, the ECG signals are acquired by the AD8232 ECG module and processed by a low power consumption FPGA (Field Programmable Gated Array) development board PYNQ-Z2. Meanwhile, a modified Deep Forest model named Aw-Deep Forest (Adaptive Weight Deep Forest) is proposed. The Aw-Deep Forest has better performance than the Deep Forest model because it improves the fitting quality of the forests. By implementing the Aw-Deep Forest model on the FPGA, the device can assess people's state of psychological stress by analyzing the HRV (heart rate variability) parameters from ECG data. This paper mainly introduces the detailed process of ECG signal collecting, filtering, analog signal to digital signal conversion, HRV parameter analysis, and psychological stress recognition with Aw-Deep Forest. The final accuracy is 81.39%.

Automatic detection of retinopathy with optical coherence tomography images via a semi-supervised deep learning method.

Automatic detection of retinopathy via computer vision techniques is of great importance for clinical applications. However, traditional deep learning based methods in computer vision require a large amount of labeled data, which are expensive and may not be available in clinical applications. To mitigate this issue, in this paper, we propose a semi-supervised deep learning method built upon pre-trained VGG-16 and virtual adversarial training (VAT) for the detection of retinopathy with optical coherence tomography (OCT) images. It only requires very few labeled and a number of unlabeled OCT images for model training. In experiments, we have evaluated the proposed method on two popular datasets. With only 80 labeled OCT images, the proposed method can achieve classification accuracies of 0.942 and 0.936, sensitivities of 0.942 and 0.936, specificities of 0.971 and 0.979, and AUCs (Area under the ROC Curves) of 0.997 and 0.993 on the two datasets, respectively. When comparing with human experts, it achieves expert level with 80 labeled OCT images and outperforms four out of six experts with 200 labeled OCT images. Furthermore, we also adopt the Gradient Class Activation Map (Grad-CAM) method to visualize the key regions that the proposed method focuses on when making predictions. It shows that the proposed method can accurately recognize the key patterns of the input OCT images when predicting retinopathy.

Cross-domain retinopathy classification with optical coherence tomography images via a novel deep domain adaptation method.

Deep learning based retinopathy classification with optical coherence tomography (OCT) images has recently attracted great attention. However, existing deep learning methods fail to work well when training and testing datasets are different due to the general issue of domain shift between datasets caused by different collection devices, subjects, imaging parameters, etc. To address this practical and challenging issue, we propose a novel deep domain adaptation (DDA) method to train a model on a labeled dataset and adapt it to an unlabelled dataset (collected under different conditions). It consists of two modules for domain alignment, that is, adversarial learning and entropy minimization. We conduct extensive experiments on three public datasets to evaluate the performance of the proposed method. The results indicate that there are large domain shifts between datasets, resulting a poor performance for conventional deep learning methods. The proposed DDA method can significantly outperform existing methods for retinopathy classification with OCT images. It achieves retinopathy classification accuracies of 0.915, 0.959 and 0.990 under three cross-domain (cross-dataset) scenarios. Moreover, it obtains a comparable performance with human experts on a dataset where no labeled data in this dataset have been used to train the proposed DDA method. We have also visualized the learnt features by using the t-distributed stochastic neighbor embedding (t-SNE) technique. The results demonstrate that the proposed method can learn discriminative features for retinopathy classification.

Imaging sebaceous gland using optical coherence tomography with deep learning assisted automatic identification.

Imaging sebaceous glands and evaluating morphometric parameters are important for diagnosis and treatment of serum problems. In this article, we investigate the feasibility of high-resolution optical coherence tomography (OCT) in combination with deep learning assisted automatic identification for these purposes. Specifically, with a spatial resolution of 2.3 µm × 6.2 µm (axial × lateral, in air), OCT is capable of clearly differentiating sebaceous gland from other skin structures and resolving the sebocyte layer. In order to achieve efficient and timely imaging analysis, a deep learning approach built upon ResNet18 is developed to automatically classify OCT images (with/without sebaceous gland), with a classification accuracy of 97.9%. Based on the result of automatic identification, we further demonstrate the possibility to measure gland size, sebocyte layer thickness and gland density.

Diagnostic Value of Structural and Functional Neuroimaging in Autoimmune Epilepsy.

Epilepsy is a common nervous system disease, which affects about 70 million people all over the world. In 2017, the International League Against Epilepsy (ILAE) considered immune factors as its independent cause, and the concept of autoimmune epilepsy (AE) was widely accepted. Early diagnosis and timely treatment can effectively improve the prognosis of the disease. However, due to the diversity of clinical manifestations, the expensive cost of autoantibody detection, and the increased prevalence in Western China, the difficulty for clinicians in early diagnosis and treatment has increased. Fortunately, convenient and fast imaging examinations are expected to help even more. The imaging manifestations of AE patients were characteristic, especially the combined application of structural and functional neuroimaging, which improved the diagnostic value of imaging. In this paper, several common autoantibodies associated with AE and their structure and function changes in neuroimaging were reviewed to provide help for neurologists to achieve the goal of precision medicine.

Olfactomedin-3 Enhances Seizure Activity by Interacting With AMPA Receptors in Epilepsy Models.

Background: OLFM3 (olfactomedin-3) is a member of the olfactomedin domain family, which has been found to stimulate the formation and adhesion of tight cell connections and to regulate cytoskeleton formation and cell migration. Differences in the gene coding for OLFM3 have been found between patients with epilepsy and controls. However, the exact role of OLFM3 in epilepsy has not been thoroughly investigated. Methods: Biochemical methods were used to assess OLFM3 expression and localization in the cortex of patients with temporal lobe epilepsy and in the hippocampus and cortex of epileptic mice. Electrophysiological recordings were used to measure the role of OLFM3 in regulating hippocampal excitability in a model of magnesium-free-induced seizure in vitro. Behavioral experiments were performed in a pentylenetetrazol (PTZ)-induced seizure model, and electroencephalograms (EEGs) were recorded in the chronic phase of the kainic acid (KA)-induced epilepsy model in vivo. OLFM3 and its interaction with AMPAR (α-amino-3-hydroxy-5-methyl-4-isoxazole-propionic acid receptor) subunits were analyzed by co-immunoprecipitation. Results: The expression of OLFM3 was increased in the cortex of patients with temporal lobe epilepsy and in the hippocampus and cortex of epileptic mice compared with controls. Interestingly, lentivirus-mediated overexpression of OLFM3 in the hippocampus increased the susceptibility of mice to PTZ-induced seizures, and OLFM3 knockdown had the opposite effect. OLFM3 affected AMPAR currents in a brain-slice model of epileptiform activity induced by Mg2+-free medium. We found that OLFM3 co-immunoprecipitation with GluA1 and GluA2. Furthermore, downregulation or overexpression of OLFM3 in the hippocampus affected the membrane expression of GluA1 and GluA2 in epileptic mice. Conclusion: These findings reveal that OLFM3 may enhance seizure activity by interacting with GluA1 and GluA2, potentially indicating a molecular mechanism for new therapeutic strategies.

Septic patients in the intensive care unit present different nasal microbiotas.

AIM: The primary objective of this study was to evaluate correlations among mortality, intensive care unit (ICU) length of stay and airway microbiotas in septic patients. MATERIALS & METHODS: A deep-sequencing analysis of the 16S rRNA gene V4 region was performed. RESULTS: The nasal microbiota in septic patients was dominated by three nasal bacterial types (Corynebacterium, Staphylococcus and Acinetobacter). The Acinetobacter type was associated with the lowest diversity and longest length of stay (median: 9 days), and the Corynebacterium type was associated with the shortest length of stay. We found that the Acinetobacter type in the >9-day group was associated with the highest mortality (33%). CONCLUSION: Septic patients have three nasal microbiota types, and the nasal microbiota is related to the length of stay and mortality.

Estimation of Refractive Index for Biological Tissue Using Micro-Optical Coherence Tomography.

The refractive index of a biological tissue is required for investigating the tissue's optical properties. Efforts have been made to characterize the refractive indices of biological tissues at a single wavelength, but it is more convenient to know the Cauchy's coefficients, which provide refractive index over a wide range of wavelengths. We demonstrate a method to noninvasively provide the Cauchy's dispersion coefficients of biological tissues using a micro-optical coherence tomography. Using the short-frequency Fourier transforms, the relative optical thickness of the sample in the wavelength range of the broadband source was obtained from interferograms. The coefficients of the Cauchy's equation were estimated based on the wavelength-dependent sample thickness. We validated the proposed method using distilled water and fresh rat cornea ex vivo, and the experimental results were consistent with the reference data.

Extending the depth of focus of fiber-optic optical coherence tomography using a chromatic dual-focus design.

We report a dual-focus fiber-optic probe designed to extend depth of focus (DOF) in high-resolution endoscopic optical coherence tomography. We exploited the broad spectral bandwidth of a supercontinuum source and, in the fiber probe, the foci of the 750-1000 nm and 1100-1450 nm inputs were axially chromatically shifted. The interference signals from the two spectral bands were measured with a Si camera-based spectrometer and an InGaAs camera-based spectrometer, respectively. We verified the feasibility of the design using a phantom composed of microparticles and swine small intestine tissue ex vivo. The results showed that a transverse resolution below 5 µm over 300 µm could be maintained, and that the extended DOF was 2 times larger than that of the single focus probe via the use of dual spectral band inputs and a chromatic focal shift.

[A machine learning model based on initial gut microbiome data for predicting changes of Bifidobacterium after prebiotics consumption].

OBJECTIVE: To investigate the effects of prebiotics supplementation for 9 days on gut microbiota structure and function and establish a machine learning model based on the initial gut microbiota data for predicting the variation of Bifidobacterium after prebiotic intake. METHODS: With a randomized double-blind self-controlled design, 35 healthy volunteers were asked to consume fructo-oligosaccharides (FOS) or galacto-oligosaccharides (GOS) for 9 days (16 g per day). 16S rRNA gene high-throughput sequencing was performed to investigate the changes of gut microbiota after prebiotics intake. PICRUSt was used to infer the differences between the functional modules of the bacterial communities. Random forest model based on the initial gut microbiota data was used to identify the changes in Bifidobacterium after 5 days of prebiotic intake and then to build a continuous index to predict the changes of Bifidobacterium. The data of fecal samples collected after 9 days of GOS intervention were used to validate the model. RESULTS: Fecal samples analysis with QIIME revealed that FOS intervention for 5 days reduced the intestinal flora alpha diversity, which rebounded on day 9; in GOS group, gut microbiota alpha diversity decreased progressively during the intervention. Neither FOS nor GOS supplement caused significant changes in ß diversity of gut microbiota. The area under the curve (AUC) of the prediction model was 89.6%. The continuous index could successfully predict the changes in Bifidobacterium (R=0.45, P=0.01), and the prediction accuracy was verified by the validation model (R=0.62, P=0.01). CONCLUSION: Short-term prebiotics intervention can significantly decrease α-diversity of the intestinal flora. The machine learning model based on initial gut microbiota data can accurately predict the changes in Bifidobacterium, which sheds light on personalized nutrition intervention and precise modulation of the intestinal flora.

Multiple aperture synthetic optical coherence tomography for biological tissue imaging.

An inherent compromise must be made between transverse resolution and depth of focus (DOF) in spectral domain optical coherence tomography (SD-OCT). Thus far, OCT has not been capable of providing a sufficient DOF to stably acquire cellular-resolution images. We previously reported a novel technique named multiple aperture synthesis (MAS) to extend the DOF in high-resolution OCT [Optica4, 701 (2017)]. In this technique, the illumination beam is scanned across the objective lens pupil plane by being steered at the pinhole using a custom-made microcylindrical lens. Images captured via multiple distinctive apertures were digitally refocused, which is similar to synthetic aperture radar. In this study, we applied this technique for the first time to image both a homemade microparticle sample and biological tissue. The results demonstrated the feasibility and efficacy of high-resolution biological tissue imaging with a dramatic DOF extension.

Endomicroscopic optical coherence tomography for cellular resolution imaging of gastrointestinal tracts.

Our ability to detect neoplastic changes in gastrointestinal (GI) tracts is limited by the lack of an endomicroscopic imaging tool that provides cellular-level structural details of GI mucosa over a large tissue area. In this article, we report a fiber-optic-based micro-optical coherence tomography (µOCT) system and demonstrate its capability to acquire cellular-level details of GI tissue through circumferential scanning. The system achieves an axial resolution of 2.48 µm in air and a transverse resolution of 4.8 µm with a depth-of-focus (DOF) of ~150 µm. To mitigate the issue of limited DOF, we used a rigid sheath to maintain a circular lumen and center the distal-end optics. The sensitivity is tested to be 98.8 dB with an illumination power of 15.6 mW on the sample. With fresh swine colon tissues imaged ex vivo, detailed structures such as crypt lumens and goblet cells can be clearly resolved, demonstrating that this fiber-optic µOCT system is capable of visualizing cellular-level morphological features. We also demonstrate that time-lapsed frame averaging and imaging speckle reduction are essential for clearly visualizing cellular-level details. Further development of a clinically viable µOCT endomicroscope is likely to improve the diagnostic outcome of GI cancers.

Fructooligosaccharide (FOS) and Galactooligosaccharide (GOS) Increase Bifidobacterium but Reduce Butyrate Producing Bacteria with Adverse Glycemic Metabolism in healthy young population.

The gut microbiota has been implicated in glucose intolerance and its progression towards type-2 diabetes mellitus (T2DM). Relevant randomized clinical trial with prebiotic intervention was inadequate. We sought to evaluate the impact of fructooligosaccharides (FOS) and galactooligosaccharides (GOS) on glycemia during oral glucose tolerance test (OGTT) and intestinal microbiota. A randomized double-blind cross-over study was performed with 35 adults treated with FOS and GOS for 14 days (16 g/day). Faeces sampling, OGTT and anthropometric parameters were performed. Short-term intake of high-dose prebiotics had adverse effect on glucose metabolism, as in FOS intervention demonstrated by OGTT (P < 0.001), and in GOS intervention demonstrated by fasting glucose (P < 0.05). A significant increase in the relative abundance of Bifidobacterium was observed both in FOS and GOS group, while the butyrate-producing bacteria like Phascolarctobacterium in FOS group and Ruminococcus in GOS group were decreased. A random forest model using the initial microbiota was developed to predict OGTT levels after prebiotic intervention with relative success (R = 0.726). Our study alerted even though FOS and GOS increased Bifidobacterium, they might have adverse effect on glucose metabolism by reducing butyrate-producing microbes. Individualized prebiotics intervention based on gut microbiome needs to be evaluated in future.

Single-camera full-range high-resolution spectral domain optical coherence tomography.

We developed spectral domain optical coherence tomography using a dual-channel spectrometer for complex conjugate artifacts (CCA) suppression. We used a three-line charge coupled device to simultaneously detect two interferometric spectra with 2π/3 phase difference. The complex interferometric signal was reconstructed by trigonometric manipulation of two real interferometric spectra, and then full-range images were obtained by use of inverse Fourier transform. Artifacts at direct current (DC) and the ghost remnant of the CCA are common issues with the previously reported two-spectrometer method because the slight mismatching between two spectral detection channels had strong negative effects on CCA suppression and appeared to be the limiting factor on system performance. This novel dual-channel spectrometer uses the same spectrometer optics for the two spectral detection channels and, therefore, achieves better matching between two spectral detection channels and consequently better performance in CCA suppression as compared with the dual spectrometer solution. Full-range imaging with CCA suppression up to ∼25 dB was demonstrated when imaging an attenuated reflector. The efficacy of both CCA and DC suppressions also was validated by imaging the anterior segment of a rat eye ex vivo. The quality of CCA-suppressed images was significantly improved with regard to those obtained with the dual-spectrometer design.

Multifiber angular compounding optical coherence tomography for speckle reduction.

We report on an integrated fiber optic design to implement multifiber angular compounding optical coherence tomography, which enables angular compounding for speckle reduction. A multi-facet fiber array delivers three light beams to the sample with different incident angles. Back-reflective/back-scattered signals from these channels were simultaneously detected by a three-channel spectrometer. The axial and lateral resolution was measured to be ∼3 and ∼3.5 µm, respectively, in air with ∼100 dB sensitivity. We conducted ex vivo experiments on a rat esophagus to demonstrate a contrast to noise improvement of 1.58.

Modeling of Mechanical Stress Exerted by Cholesterol Crystallization on Atherosclerotic Plaques.

Plaque rupture is the critical cause of cardiovascular thrombosis, but the detailed mechanisms are not fully understood. Recent studies have found abundant cholesterol crystals in ruptured plaques, and it has been proposed that the rapid expansion of cholesterol crystals in a limited space during crystallization may contribute to plaque rupture. To evaluate the effect of cholesterol crystal growth on atherosclerotic plaques, we modeled the expansion of cholesterol crystals during the crystallization process in the necrotic core and estimated the stress on the thin cap with different arrangements of cholesterol crystals. We developed a two-dimensional finite element method model of atherosclerotic plaques containing expanding cholesterol crystals and investigated the effect of the magnitude and distribution of crystallization on the peak circumferential stress born by the cap. Using micro-optical coherence tomography (µOCT), we extracted the cross-sectional geometric information of cholesterol crystals in human atherosclerotic aorta tissue ex vivo and applied the information to the model. The results demonstrate that (1) the peak circumference stress is proportionally dependent on the cholesterol crystal growth; (2) cholesterol crystals at the cap shoulder impose the highest peak circumference stress; and (3) spatial distributions of cholesterol crystals have a significant impact on the peak circumference stress: evenly distributed cholesterol crystals exert less peak circumferential stress on the cap than concentrated crystals.

Enhancement and bias removal of optical coherence tomography images: An iterative approach with adaptive bilateral filtering.

Optical coherence tomography (OCT) has continually evolved and expanded as one of the most valuable routine tests in ophthalmology. However, noise (speckle) in the acquired images causes quality degradation of OCT images and makes it difficult to analyze the acquired images. In this paper, an iterative approach based on bilateral filtering is proposed for speckle reduction in multiframe OCT data. Gamma noise model is assumed for the observed OCT image. First, the adaptive version of the conventional bilateral filter is applied to enhance the multiframe OCT data and then the bias due to noise is reduced from each of the filtered frames. These unbiased filtered frames are then refined using an iterative approach. Finally, these refined frames are averaged to produce the denoised OCT image. Experimental results on phantom images and real OCT retinal images demonstrate the effectiveness of the proposed filter.