Review on Microreactors for Photo-Electrocatalysis Artificial Photosynthesis Regeneration of Coenzymes.

In recent years, with the outbreak of the global energy crisis, renewable solar energy has become a focal point of research. However, the utilization efficiency of natural photosynthesis (NPS) is only about 1%. Inspired by NPS, artificial photosynthesis (APS) was developed and utilized in applications such as the regeneration of coenzymes. APS for coenzyme regeneration can overcome the problem of high energy consumption in comparison to electrocatalytic methods. Microreactors represent a promising technology. Compared with the conventional system, it has the advantages of a large specific surface area, the fast diffusion of small molecules, and high efficiency. Introducing microreactors can lead to more efficient, economical, and environmentally friendly coenzyme regeneration in artificial photosynthesis. This review begins with a brief introduction of APS and microreactors, and then summarizes research on traditional electrocatalytic coenzyme regeneration, as well as photocatalytic and photo-electrocatalysis coenzyme regeneration by APS, all based on microreactors, and compares them with the corresponding conventional system. Finally, it looks forward to the promising prospects of this technology.

Epilepsy alters brain networks in patients with insular glioma.

AIMS: We intend to elucidate the alterations of cerebral networks in patients with insular glioma-related epilepsy (GRE) based on resting-state functional magnetic resonance images. METHODS: We collected 62 insular glioma patients, who were subsequently categorized into glioma-related epilepsy (GRE) and glioma with no epilepsy (GnE) groups, and recruited 16 healthy individuals matched to the patient's age and gender to form the healthy control (HC) group. Graph theoretical analysis was applied to reveal differences in sensorimotor, default mode, visual, and executive networks among different subgroups. RESULTS: No significant alterations in functional connectivity were found in either hemisphere insular glioma. Using graph theoretical analysis, differences were found in visual, sensorimotor, and default mode networks (p < 0.05). When the glioma located in the left hemisphere, the degree centrality was reduced in the GE group compared to the GnE group. When the glioma located in the right insula, the degree centrality, nodal efficiency, nodal local efficiency, and nodal clustering coefficient of the GE group were lower than those of the GnE group. CONCLUSION: The impact of insular glioma itself and GRE on the brain network is widespread. The networks altered by insular GRE differ depending on the hemisphere location. GRE reduces the nodal properties of brain networks than that in insular glioma.

Survival outcome and predictors of WHO grade 2 and 3 insular gliomas: A classification based on the tumor spread.

OBJECTIVE: The study aimed to identify if clinical features and survival outcomes of insular glioma patients are associated with our classification based on the tumor spread. METHODS: Our study included 283 consecutive patients diagnosed with histological grade 2 and 3 insular gliomas. A new classification was proposed, and tumors restricted to the paralimbic system were defined as type 1. When tumors invaded the limbic system (referred to as the hippocampus and its surrounding structures in this study) simultaneously, they were defined as type 2. Tumors with additional internal capsule involvement were defined as type 3. RESULTS: Tumors defined as type 3 had a higher age at diagnosis (p = 0.002) and a higher preoperative volume (p < 0.001). Furthermore, type 3 was more likely to be diagnosed as IDH wild type (p < 0.001), with a higher rate of Ki-67 index (p = 0.015) and a lower rate of gross total resection (p < 0.001). Type 1 had a slower tumor growth rate than type 2 (mean 3.3%/month vs. 19.8%/month; p < 0.001). Multivariate Cox regression analysis revealed the extent of resection (HR 0.259, p = 0.004), IDH status (HR 3.694, p = 0.012), and tumor spread type (HR = 1.874, p = 0.012) as independent predictors of overall survival (OS). Tumor grade (HR 2.609, p = 0.008), the extent of resection (HR 0.488, p = 0.038), IDH status (HR 2.225, p = 0.025), and tumor spread type (HR 1.531, p = 0.038) were significant in predicting progression-free survival (PFS). CONCLUSION: The current study proposes a classification of the insular glioma according to the tumor spread. It indicates that the tumors defined as type 1 have a relatively better nature and biological characteristics, and those defined as type 3 can be more aggressive and refractory. Besides its predictive value for prognosis, the classification has potential value in formulating surgical strategies for patients with insular gliomas.

Exploring the impact of hydrogen sulfide on hematologic malignancies: A review.

Hydrogen sulfide (H2S) is one of the three most crucial gaseous messengers in the body. The discovery of H2S donors, coupled with its endogenous synthesis capability, has sparked hope for the treatment of hematologic malignancies. In the last decade, the investigation into the impact of H2S has expanded, particularly within the fields of cardiovascular function, inflammation, infection, and neuromodulation. Hematologic malignancies refer to a diverse group of cancers originating from abnormal proliferation and differentiation of blood-forming cells, including leukemia, lymphoma, and myeloma. In this review, we delve deeply into the complex interrelation between H2S and hematologic malignancies. In addition, we comprehensively elucidate the intricate molecular mechanisms by which both H2S and its donors intricately modulate the progression of tumor growth. Furthermore, we systematically examine their impact on pivotal aspects, encompassing the proliferation, invasion, and migration capacities of hematologic malignancies. Therefore, this review may contribute novel insights to our understanding of the prospective therapeutic significance of H2S and its donors within the realm of hematologic malignancies.

Cu(I)-Catalyzed Three-Component Annulation for the Synthesis of 3-Acyl Imidazo[1, 5-a]Pyridines from 2-Pyridinyl-Substituted p-Quinone Methides, Terminal Alkynes, and TsN3 Using O2 as the Oxygen Source.

Currently, most conventional methods to achieve imidazo[1,5-a]pyridines have limitations for the synthesis of 3-acyl imidazo[1,5-a]pyridines. Herein, a novel and efficient Cu(I)-catalyzed three-component annulation method for the synthesis of valuable 3-acyl imidazo[1,5-a]pyridines by the reaction of 2-pyridinyl-substituted p-QMs, terminal alkynes, and TsN3 in the presence of O2 under mild conditions have successfully been developed. The investigation indicated that molecular oxygen (O2) and TsN3, respectively, serving as oxygen and nitrogen sources, were essential for the successful completion of the reaction system.

Research on UAV Flight Parameter Identification Method Based on Launch Force and Airspeed.

Flight parameters are crucial criteria for UAV control, playing a significant role in ensuring the safe and efficient completion of missions. Launch force and airspeed information are key parameters in the early and middle stages of flight, serving as important data for monitoring the UAV's flight status. In response to challenges such as weak launch force, low identification rates, small airspeed, and low recognition accuracy in UAVs, a method for identifying UAV flight parameters based on launch force and airspeed is proposed. From the aspect of launch force identification, a recognition method based on a low-g value accelerometer information source is proposed, utilizing a 'multi-level time window + threshold' approach. For airspeed identification, an optimization method for airspeed measurement under the Kalman filter architecture is introduced. A device for airspeed measurement based on pressure sensors is designed, and the recommended installation position is determined through simulation. Furthermore, the feasibility and robustness of the proposed launch force identification and airspeed measurement optimization methods are validated through simulation. Finally, the effectiveness of the design is verified through centrifuge and wind tunnel experiments. This research provides technical support for the identification of the launch force and airspeed measurement in UAVs.

Application of postoperative adjuvant radiotherapy in limited-stage small cell lung cancer: A systematic review and meta-analysis.

BACKGROUND AND PURPOSE: One of the most important treatments for small cell lung cancer (SCLC) is radiation therapy. Currently, the criteria for administering postoperative adjuvant radiotherapy (PORT) in SCLC remain uncertain. Therefore, we conducted a meta-analysis to investigate the influence of PORT on the prognosis of limited-stage SCLC (LS-SCLC). METHODS: We conducted a comprehensive search across three databases, PubMed, Embase, and the Cochrane Library. Data analysis involved utilizing both random-effects and fixed-effects models for pooling the results. A comparative analysis was performed to assess the prognostic outcomes of patients with LS-SCLC who did and did not undergo PORT. The primary outcome assessed was overall survival (OS), while the secondary outcome was disease-free survival (DFS). RESULTS: This analysis included 11 retrospective studies comprising 7694 eligible participants. Among the entire population of LS-SCLC patients, the OS was superior in those receiving PORT than in those not receiving it (hazard ratio [HR]: 0.79, 95 % confidence interval [CI]: 0.71-0.87; P < 0.0001). In pN0 stage LS-SCLC patients, PORT was associated with a detrimental effect on OS (HR: 1.22, 95 % CI: 1.04-1.43; P = 0.01). In pN1 stage LS-SCLC patients, additionally administering PORT did not provide a significant OS advantage as compared to not administering it (HR: 0.82, 95 % CI: 0.60-1.12; P = 0.21). In pN2 stage LS-SCLC patients, those receiving PORT demonstrated a significant improvement in OS (HR: 0.59; 95 % CI: 0.50-0.70; P < 0.0001) as compared to those not receiving it. Regarding DFS in LS-SCLC patients, the difference in the protective effect with and without the administration of PORT was less pronounced (HR: 0.76, 95 % CI: 0.58-1.00; P = 0.053). CONCLUSIONS: With respect to OS, PORT is not advisable in patients with pN0 or pN1 stage LS-SCLC but is highly recommended in pN2 stage LS-SCLC. Further research is warranted to confirm these findings.

Portable rotary PCR system for real-time detection of Pseudomonas aeruginosa in milk.

The rapid quantitative detection of Pseudomonas aeruginosa in milk is of great significance to food safety. Quantitative real-time polymerase chain reaction (qPCR) technology is a good choice to meet this requirement. A good qPCR system should show the advantages of being low cost, having low-power consumption, having potential for miniaturization and be portable. However, most of the time-domain-based qPCR systems reported to date do not meet these requirements. In this study, we propose a novel real-time rotary PCR reaction system (RRP) that meets all the abovementioned specifications, and contains four modules: a heating control module, a disposable PCR capillary tube, a mechanical control module, and a photoelectric detection module. The volume of our homemade-PCR capillary tube is only 3 µL. The total manufacturing cost is cheaper than $200, and the capillary tube is about 1.4 cents. The size parameter of the RRP is less than 300 mm × 150 mm × 150 mm, using low mobile power sources to operate. All the features mean that the RRP meets the advantages of low sample volumes, enhanced thermal conductivity and being portable. Through conducting the experimental quantitative detection of Pseudomonas aeruginosa in milk and theoretical simulations by COMSOL, we prove the feasibility of this rotary PCR real-time detection system, which has broad application prospects in the rapid detection of bacteria and food safety.

Spatial distribution of supratentorial diffuse gliomas: A retrospective study of 990 cases.

Background: Gliomas distribute unevenly in the supratentorial brain space. Many factors were linked to tumor locations. This study aims to describe a more detailed distributing pattern of these tumors with age and pathological factors concerned. Methods: A consecutive series of 990 adult patients with newly-diagnosed supratentorial diffuse gliomas who underwent resection in Beijing Tiantan Hospital between January 2013 and January 2017 were retrospectively reviewed. For each patient, the anatomic locations were identified by the preoperative MRI, and the pathological subtypes were reviewed for histological grade and molecular status (if any) from his medical record. The MNI template was manually segmented to measure each anatomic location's volume, and its invaded ratio was then adjusted by the volume to calculate the frequency density. Factors of age and pathological subtypes were also compared among locations. Results: The insulae, hippocampi, and corpus callosum were locations of the densest frequencies. The frequency density decreased from the anterior to posterior (frontal - motor region - sensory region - parietal - occipital), while the grade (p < 0.0001) and the proportion of IDH-wt (p < 0.0001) increased. More tumors invading the right basal ganglion were MGMT-mt (p = 0.0007), and more of those invading the left frontal were TERT-wt (p = 0.0256). Age varied among locations and pathological subtypes. Conclusions: This study demonstrated more detailed spatial disproportions of supratentorial gliomas. There are potential interactions among age, pathological subtypes, and tumor locations.

Hydrogen sulfide and its donors: Novel antitumor and antimetastatic agents for liver cancer.

Hepatocellular carcinoma (HCC) is the sixth most frequent human cancer and the world's third most significant cause of cancer mortality. HCC treatment has recently improved, but its mortality continues to increase worldwide due to its extremely complicated and heterogeneous genetic abnormalities. After nitric oxide (NO) and carbon monoxide (CO), the third gas signaling molecule discovered is hydrogen sulfide (H2S), which has long been thought to be a toxic gas. However, numerous studies have proven that H2S plays many pathophysiological roles in mammals. Endogenous or exogenous H2S can decrease cell proliferation, promote apoptosis, block cell cycle, invasion and migration through various cellular signaling pathways. This review analyzes and discusses the recent literature on the function and molecular mechanism of H2S and H2S donors in HCC, so as to provide convenience for the scientific research and clinical application of H2S in the treatment of liver cancer.

A nomogram model to predict the acute venous thromboembolism risk after surgery in patients with glioma.

INTRODUCTION: Postoperative venous thromboembolism (VTE) is a common complication for glioma patients, with an incidence rate of about 20 %. The purpose of this study was to explore the risk factors of acute VTE after glioma surgery, which may provide an essential reference for clinical guidance on the prevention of acute VTE. MATERIALS AND METHODS: A total of 435 patients who underwent glioma surgery from 2012 to 2021 were included in this study. Duplex ultrasonography was performed routinely 3-5 days after the surgery to define VTE. Univariate and multivariate logistic regression analyses were performed to explore the independent predictor of acute VTE after glioma surgery and use these selected risk factors to construct and validate a nomogram. RESULTS: Several risk factors for predicting acute VTE after glioma surgery were identified and used to build the nomogram: age, operation time, systemic immune-inflammation index (SII), hypertension, and diabetes mellitus. The area under the curve of the nomogram was 0.834, indicating good discrimination. Hosmer-Lemeshow of the calibration curve was 3.05 (P = 0.98), showing a high degree of agreement between the prediction and actual outcome. Decision curve analysis indicated that the nomogram model was helpful when the incidence of VTE was 5-80 %. CONCLUSIONS: A nomogram to predict acute VTE after glioma surgery was constructed and validated. Clinicians can use this predictive model to achieve risk assessment and take different treatment measures to prevent acute postoperative VTE and improve patients' quality of life effectively.

Video Based Cocktail Causal Container for Blood Pressure Classification and Blood Glucose Prediction.

With the development of modern cameras, more physiological signals can be obtained from portable devices like smartphone. Some hemodynamically based non-invasive video processing applications have been applied for blood pressure classification and blood glucose prediction objectives for unobtrusive physiological monitoring at home. However, this approach is still under development with very few publications. In this paper, we propose an end-to-end framework, entitled cocktail causal container, to fuse multiple physiological representations and to reconstruct the correlation between frequency and temporal information during multi-task learning. Cocktail causal container processes hematologic reflex information to classify blood pressure and blood glucose. Since the learning of discriminative features from video physiological representations is quite challenging, we propose a token feature fusion block to fuse the multi-view fine-grained representations to a union discrete frequency space. A causal net is used to analyze the fused higher-order information, so that the framework can be enforced to disentangle the latent factors into the related endogenous association that corresponds to down-stream fusion information to improve the semantic interpretation. Moreover, a pair-wise temporal frequency map is developed to provide valuable insights into extraction of salient photoplethysmograph (PPG) information from fingertip videos obtained by a standard smartphone camera. Extensive comparisons have been implemented for the validation of cocktail causal container using a Clinical dataset and PPG-BP benchmark. The root mean square error of 1.329±0.167 for blood glucose prediction and precision of 0.89±0.03 for blood pressure classification are achieved in Clinical dataset.

Analyzing and mapping the research status, hotspots, and frontiers of biological wound dressings: An in-depth data-driven assessment.

Wound dressings can be applied over the wound sites to provide long-lasting wound management and improve wound healing. Biological wound dressings are superior to synthetic materials due to biodegradability and biocompatibility. These biomaterials have demonstrated huge potential in the field of wound dressings. Applying bibliometric analysis combined with results-based descriptions to characterize the research status, hotspots, and cutting-edge topics, this study is the first in-depth qualitative, quantitative, data-driven overview of biological wound dressings research in recent decades. Filtered data were used to construct co-citation, heatmaps, bi-clustering, strategy maps, and other analyses and visualization. The results show that research on biological wound dressings has progressed considerably in the last 5 years with extensive global collaboration. A clear knowledge base has been developed. Chitosan hydrogels, bacterial cellulose, active agents (silver nanoparticles, growth factors, curcumin, etc.), and electrospinning fibers stand out as research hotspots. The research frontiers include novel starting materials, precise and controlled release systems, and clinical and regenerative medicine applications. We interpreted an overview of the excavated topics and expected the findings here to provide a guide and inspire innovations for developing the next generation wound dressings.

APOBEC3B and CD274 as Combined Biomarkers for Predicting Response to Immunotherapy in Urothelial Carcinoma of the Bladder.

Immunotherapy has become a promising form of treatment for cancers. There is a need to predict response to immunotherapy accurately. In the UCSC Xena, pan-cancer analysis revealed a positive relationship between APOBEC3B (A3B) and tumor mutational burden (R = 0.28, P < 0.001) and microsatellite instability (R = 0.12, P < 0.05). Naturally, the A3B high expression group had higher tumor mutational burden and microsatellite instability than the low expression group. The bladder cancer (BLCA) cohort in The Cancer Genome Atlas (TCGA) revealed tumor mutational signatures of A3B high and low expression groups. Compared to the low expression group, the high expression group had a higher number of SNPs and mutations. Subsequently, A3B was profiled for immune cell infiltration and immune checkpoints in bladder cancer. The results showed that A3B was positively correlated with most immune cells. Compared with the A3B low expression group, the A3B high expression group had higher expression of immune checkpoints. A3B was positively correlated with CD274 (R = 0.12, P = 0.016). This indicated that the high expression of A3B may have a better response to immunotherapy. Furthermore, data from the IMvigor210 immunotherapy clinical trial was used to confirm the findings of this study. The combined survival analysis of A3B and CD274 showed that the group of patients with high expression of CD274 and A3B was found to have a significantly higher survival rate than the rest of the patient group (P < 0.047). The results demonstrated that A3B has a significant role in immunotherapy. Moreover, the combined biomarkers of A3B and CD274 were more effective in predicting response to immunotherapy in bladder urothelial carcinoma. The findings of this study provide valuable insights for precision medicine.

EIF2S2 is a novel independent prognostic biomarker and correlated with immune infiltrates in hepatocellular carcinoma.

Background: Hepatocellular carcinoma (HCC) is a highly malignant disease with poor prognosis. It is urgent to find effective biomarkers. Eukaryotic Translation Initiation Factor 2 Subunit Beta (EIF2S2) is a subunit of heterotrimeric G protein EIF2, and its function is still unclear. We studied the role of EIF2S2 in the malignant progression of liver cancer and its relationship with immune infiltration. Methods: Download the RNA expression and clinical information of EIF2S2 from the Cancer Genome Atlas (TCGA) database, analyze the relationship between the expression of EIF2S2 and the prognosis and clinicopathological characteristics of HCC, analyze the differential genes by Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) and tumor related immune infiltrating cells. The Protein expression level of EIF2S2 was obtained from Human Protein Atlas (HPA) databases. The relationship between EIF2S2 expression and immune infiltrates in HCC was analyzed on TIMER 2.0. The data processing analysis based on R language. Drug Sensitivity data from Genomics of Drug Sensitivity in Cancer (GDSC). Results: EIF2S2 is highly expressed in HCC patients and is associated with poor prognosis. The expression of EIF2S2 was also correlated with age, clinical stage and pathological grade. Univariate and multivariate COX regression analysis showed that EIF2S2 was an independent risk factor for survival. The receiver operating characteristic (ROC) curve of EIF2S2 also confirmed the diagnostic value of EIF2S2 in HCC patients. Through GO and KEGG enrichment analysis, EIF2S2 expression was found to be closely related to some immune pathways. The expression of EIF2S2 was correlated with memory B cell, plasma B cell, CD8+ T cell, CD4+ resting memory T cell and the expression of some immune checkpoints, such as PDCD1, TIGIT and CTLA-4. It is also more sensitive to paclitaxel, sunitinib and other drugs. Conclusion: This study shows that EIF2S2 can be used as a prognostic factor for HCC, which is closely related to immune infiltration and immune checkpoints, and may play a potential regulatory role in predicting drug sensitivity.

Research Progress and Future Trends of Microfluidic Paper-Based Analytical Devices in In-Vitro Diagnosis.

In vitro diagnosis (IVD) has become a hot topic in laboratory research and achievement transformation. However, due to the high cost, and time-consuming and complex operation of traditional technologies, some new technologies are being introduced into IVD, to solve the existing problems. As a result, IVD has begun to develop toward point-of-care testing (POCT), a subdivision field of IVD. The pandemic has made governments and health institutions realize the urgency of accelerating the development of POCT. Microfluidic paper-based analytical devices (µPADs), a low-cost, high-efficiency, and easy-to-operate detection platform, have played a significant role in advancing the development of IVD. µPADs are composed of paper as the core material, certain unique substances as reagents for processing the paper, and sensing devices, as auxiliary equipment. The published reviews on the same topic lack a comprehensive and systematic introduction to µPAD classification and research progress in IVD segmentation. In this paper, we first briefly introduce the origin of µPADs and their role in promoting IVD, in the introduction section. Then, processing and detection methods for µPADs are summarized, and the innovative achievements of µPADs in IVD are reviewed. Finally, we discuss and prospect the upgrade and improvement directions of µPADs, in terms of portability, sensitivity, and automation, to help researchers clarify the progress and overcome the difficulties in subsequent µPAD research.

Risk Model and Immune Signature of m7G-Related lncRNA Based on Lung Adenocarcinoma.

Lung cancer is a major cause of cancer-related deaths globally, with a dismal prognosis. N7-methylguanosine (m7G) is essential for the transcriptional phenotypic modification of messenger RNA (mRNA) and long noncoding RNA (lncRNA). However, research on m7G-related lncRNAs involved in lung adenocarcinoma (LUAD) regulation is still limited. Herein, we aim to establish a prognostic model of m7G-related lncRNAs and investigate their immune properties. Eight prognostic m7G-related lncRNAs were identified using univariate Cox analysis. Six m7G-related lncRNAs were identified using LASSO-Cox regression analysis to construct risk models, and all LUAD patients in The Cancer Genome Atlas (TCGA) cohort was divided into low-risk and high-risk subgroups. The accuracy of the model was verified by Kaplan-Meier analysis, time-dependent receiver operating characteristic, principal component analysis, independent prognostic analysis, nomogram, and calibration curve. Further studies were conducted on the gene set enrichment and disease ontology enrichment analyses. The gene set enrichment analysis (GSEA) revealed that the high-risk group enriched for cancer proliferation pathways, and the enrichment analysis of disease ontology (DO) revealed that lung disease was enriched, rationally explaining the superiority of the risk model. Finally, we found that the low-risk group had higher immune infiltration and checkpoint expression. It can be speculated that the low-risk group has a better effect on immunotherapy. Susceptibility to antitumor drugs in different risk subgroups was assessed, and it found that the high-risk group showed high sensitivity to first-line treatment drugs for non-small cell lung cancer. In conclusion, a risk model based on 6 m7G-related lncRNAs can not only predict the overall survival (OS) rate of LUAD patients but also guide individualized treatment for these patients.

CMS2-Net: Semi-Supervised Sleep Staging for Diverse Obstructive Sleep Apnea Severity.

Although the development of computer-aided algorithms for sleep staging is integrated into automatic detection of sleep disorders, most supervised deep learning-based models might suffer from insufficient labeled data. While the adoption of semi-supervised learning (SSL) can mitigate the issue, the SSL models are still limited to the lack of discriminative feature extraction for diverse obstructive sleep apnea (OSA) severity. This model deterioration might be exacerbated during the domain adaptation. Such exploration on the alleviation of domain-shift of SSL model between different OSA conditions has attracted more and more attentions from the clinic. In this work, a co-attention meta sleep staging network (CMS2-net) is proposed to simultaneously deal with two issues: the inter-class disparity problem and the intra-class selection problem. Within CMS2-net, a co-attention module and a triple-classifier are designed to explicitly refine the coarse feature representations by identifying the class boundary inconsistency. Moreover, the mutual information with meta contrastive variance is introduced to supervise the gradient stream from a multi-scale view. The performance of the proposed framework is demonstrated on both public and local datasets. Furthermore, our approach achieves the state-of-the-art SSL results on both datasets.

Hybrid manifold-deep convolutional neural network for sleep staging.

Analysis of electroencephalogram (EEG) is a crucial diagnostic criterion for many sleep disorders, of which sleep staging is an important component. Manual stage classification is a labor-intensive process and usually suffered from many subjective factors. Recently, more and more computer-aided techniques have been applied to this task, among which deep convolutional neural network has been performing well as an effective automatic classification model. Despite some comprehensive models have been developed to improve classification results, the accuracy for clinical applications has not been reached due to the lack of sufficient labeled data and the limitation of extracting latent discriminative EEG features. Therefore, we propose a novel hybrid manifold-deep convolutional neural network with hyperbolic attention. To overcome the shortage of labeled data, we update the semi-supervised training scheme as an optimal solution. In order to extract the latent feature representation, we introduce the manifold learning module and the hyperbolic module to extract more discriminative information. Eight subjects from the public dataset are utilized to evaluate our pipeline, and the model achieved 89% accuracy, 70% precision, 80% sensitivity, 72% f1-score and kappa coefficient of 78%, respectively. The proposed model demonstrates powerful ability in extracting feature representation and achieves promising results by using semi-supervised training scheme. Therefore, our approach shows strong potential for future clinical development.

Hybrid adversarial-discriminative network for leukocyte classification in leukemia.

PURPOSE: Leukemia is a lethal disease that is harmful to bone marrow and overall blood health. The classification of white blood cell images is crucial for leukemia diagnosis. The purpose of this study is to classify white blood cells by extracting discriminative information from cell segmentation and combining it with the fine-grained features. We propose a hybrid adversarial residual network with support vector machine (SVM), which utilizes the extracted features to improve the classification accuracy for human peripheral white cells. METHODS: Firstly, we segment the cell and nucleus by utilizing an adversarial residual network, which contains a segmentation network and a discriminator network. To extract features that can handle the inter-class consistency problem effectively, we introduce the adversarial residual network. Then, we utilize convolutional neural network (CNN) features and histogram of oriented gradient (HOG) features, which can extract discriminative features from images of segmented cell nuclei. To utilize the representative features fully, a discriminative network is introduced to deal with neighboring information at different scales. Finally, we combine the vectors of HOG features with those of CNN features and feed them into a linear SVM to classify white blood cells into six types. RESULTS: We used three methods to evaluate the effect of leukocyte classification based on 5000 leukocyte images acquired from a local hospital. The first approach is to use the CNN features as the input of SVM to classify leukocytes, which achieved 94.23% specificity, 95.10% sensitivity, and 94.41% accuracy. The use of the HOG features for SVM achieved 83.50% specificity, 87.50% sensitivity, and 85.00% accuracy. The use of combined CNN and HOG features achieved 94.57% specificity, 96.11% sensitivity, and 95.93% accuracy. CONCLUSIONS: We propose a novel hybrid adversarial-discriminative network for the classification of microscopic leukocyte images. It improves the accuracy of cell classification, reduces the difficulty and time pressure of doctors' work, and economizes the valuable time of doctors in daily clinical diagnosis.