A Semi-Supervised Learning Framework for Classifying Colorectal Neoplasia Based on the NICE Classification.

Labelling medical images is an arduous and costly task that necessitates clinical expertise and large numbers of qualified images. Insufficient samples can lead to underfitting during training and poor performance of supervised learning models. In this study, we aim to develop a SimCLR-based semi-supervised learning framework to classify colorectal neoplasia based on the NICE classification. First, the proposed framework was trained under self-supervised learning using a large unlabelled dataset; subsequently, it was fine-tuned on a limited labelled dataset based on the NICE classification. The model was evaluated on an independent dataset and compared with models based on supervised transfer learning and endoscopists using accuracy, Matthew's correlation coefficient (MCC), and Cohen's kappa. Finally, Grad-CAM and t-SNE were applied to visualize the models' interpretations. A ResNet-backboned SimCLR model (accuracy of 0.908, MCC of 0.862, and Cohen's kappa of 0.896) outperformed supervised transfer learning-based models (means: 0.803, 0.698, and 0.742) and junior endoscopists (0.816, 0.724, and 0.863), while performing only slightly worse than senior endoscopists (0.916, 0.875, and 0.944). Moreover, t-SNE showed a better clustering of ternary samples through self-supervised learning in SimCLR than through supervised transfer learning. Compared with traditional supervised learning, semi-supervised learning enables deep learning models to achieve improved performance with limited labelled endoscopic images.

A bioinformatics analysis and an experimental validation of the hypoxia-related prognostic model.

Background: Hypoxia plays an important role in the development of pancreatic cancer (PCA). However, there is few research on the application of hypoxia molecules in predicting the prognosis of PCA. We aimed to establish a prognostic model based on hypoxia-related genes (HRGs) for PCA to discover new biomarkers, and to reveal the potential of this prognostic model for evaluating the tumor microenvironment (TME). Methods: Univariate Cox regression analysis was used to identify HRGs associated with overall survival (OS) of PCA samples. A hypoxia-related prognostic model was established based on least absolute shrinkage and selection operator (LASSO) regression analysis in The Cancer Genome Atlas (TCGA) cohort. The model was validated in the Gene Expression Omnibus (GEO) datasets. The Cell-type Identification by Estimating Relative Subsets of RNA Transcripts (CIBERSORT) algorithm was used to estimate the infiltration of immune cells. A wound healing assay and transwell invasion assay were used to explore the biological functions of target genes in PCA. Results: A total of 18 HRGs were differentially expressed between the tumor and normal pancreatic tissue, 4 (BHLHE40, ENO1, SDC4, and TGM2) of which were selected to construct a prognostic model. According to this model, patients in the high-risk group had a less favorable prognosis. Furthermore, the proportion of M0 macrophages was significantly higher in high-risk tissue-type patients, whereas naïve B cells, plasma cells, CD8+ T cells, and activated CD4+ memory T cells were significantly lower. The expression of BHLHE40 in PCA cells was significantly up-regulated under hypoxic conditions. Moreover, BHLHE40 was shown to regulate the transcription and expression of the downstream target gene TLR3. The wound healing assay and transwell invasion assay indicated that BHLHE40 mediated PCA cell migration and invasion by targeting the downstream gene TLR3. Conclusions: The hypoxia-related prognostic model established by the expression pattern of 4 HRGs can be used to predict the prognosis and assess the TME of PCA patients. Mechanically, activation of the BHLHE40/TLR3 axis is responsible for the promoted invasion and migration of PCA cells in a hypoxic environment.

Application of computed tomography-based radiomics combined with clinical factors in the diagnosis of malignant degree of lung adenocarcinoma.

Background: As an emerging technology, radiomics is being widely used in the diagnosis of early lung cancer due to its excellent diagnostic performance. However, there is a lack of studies that apply radiomics to the diagnosis of malignancy of lung adenocarcinoma. Thus, we used computed tomography (CT)-based radiomics to construct a model for the diagnosis of high-risk lung adenocarcinoma. Methods: Data of 170 patients who underwent surgical treatment at the First Affiliated Hospital of Soochow University and had a maximum nodule diameter ≤2 cm on preoperative CT images between January 2020 and December 2021 were retrospectively analyzed. All enrolled patients were randomly divided into experimental and validation groups according to the ratio of 7:3. The diagnosis of lung adenocarcinoma was based on postoperative pathological results. The region of interest was delineated on preoperative CT images, and the radiomics features were extracted. The least absolute shrinkage and selection operator (LASSO) was used to screen the radiomics features thus obtaining the radiomics score (Radscore), which was the basis of the radiomics model. Based on the multivariate regression analysis, independent predictors were screened from the clinical baseline data and imaging features thus constructing clinical model. Multivariate logistic regression was used to combine independent predictors and the Radscore to form a comprehensive nomogram. The diagnostic performance of constructed models was evaluated based on receiver operating characteristic (ROC) curves and decision curve analysis (DCA). Results: The sensitivity and specificity of the clinical model based on consolidation-to-tumor ratio (CTR), lobulated signs and vascular anomaly signs was 70.0% and 76.7% in the validation group. The radiomics model [area under the curve (AUC) 0.926; 95% confidence interval (CI): 0.857-0.995] and the comprehensive model (AUC 0.922; 95% CI: 0.851-0.992) performed better than clinical model (AUC 0.839; 95% CI: 0.720-0.958) in the validation group. The sensitivity and specificity of the comprehensive model was 85.0% and 80.0% in the validation group. DCA of radiomics model and comprehensive model suggested they have better net survival benefit than clinical model. Conclusions: Compared with clinical model, radiomics model and comprehensive model had better diagnostic performance in distinguishing malignant degree of lung adenocarcinoma.

Discussion of the experience and improvement of an enhanced recovery after surgery procedure for minimally invasive lobectomy: a cohort study.

BACKGROUND: It is of great significance to explore a path for expedited recovery from thoracic surgery for patients undergoing minimally invasive lobectomy to ensure their rapid and smooth recovery and to conserve medical resources. METHODS: We analyzed 629 cases from the Department of Thoracic Surgery, First Affiliated Hospital of Soochow University from January 2018 to January 2020. According to the length of postoperative stay (LOS) and perioperative management, the 629 patients were divided into group A [routine management group (RMG)], group B [rapidly recovery group (RRG), LOS >72 h], and group C (RRG, LOS ≤72 h). The t-test and chi-square test were used to compare the postoperative complications (PC), chest tube indwelling time (CTIT), LOS, postoperative opioid dosages (POD), and total costs (TC) of the 3 groups. RESULTS: Compared with the RMG, the LOS, PC, CTIT, POD, and TC of the RRG were statistically significantly ameliorated (P<0.05). When compared with group A, the PC (18.9% vs. 38.8%), LOS (2.74±0.80 vs. 5.70±1.10 d), CTIT (46.1±18.5 vs. 123.6±34.8 h), and TC (¥51,517±7,217 vs. ¥65,781±8,200) of group C were all decreased. Compared with group A, group B had more preoperative complications, less CTIT, but no significant difference in LOS, PC, and TC during hospital stay. Compared with group B, group C had less preoperative complications, and reduced LOS, CTIT, and TC. The POD in group A was significantly higher than that in groups B and C. CONCLUSIONS: Use of an enhanced recover after surgery (ERAS) protocol can reduce the postoperative complications and shorten the LOS. Some high-risk patients cannot be discharged within 72 h after surgery, but their incidence of grade II postoperative lung complications can be decreased and they can benefit from ERAS. Single intercostal nerve block, COX-2 selective inhibitor, and removal of the chest tube as early as possible can improve the ERAS. KEYWORDS: Enhanced recover after surgery (ERAS); video-assisted thoracic surgery (VATS); minimally invasive lobectomy.

In-situ lignin drives lytic polysaccharide monooxygenases to enhance enzymatic saccharification.

Recently low-molecular lignin was reported to activate lytic polysaccharide monooxygenases (LPMOs) to oxidize cellulose. However, whether lignin formed in cell wall can play the role as electron donor for LPMOs is still largely unknown due to the complex ultrastructure of lignocellulosic biomass. In this study, we presented a new strategy to elucidate in-situ lignin function in LPMOs reaction. A lignocellulosic mimicking model was used as substrate, which was equipped with a polysaccharide template of self-assembled bacterial cellulose film and synthesized lignin. Remarkably, it has been demonstrated that lignin polymer deposited on cellulose can reduce LPMOs in-situ for cellulose oxidation and then boost cellulose hydrolysis, and the cellulose conversion ratio of the mimicked lignocellulosic film was increased by 26.0%. More importantly, lignin in-situ might exceed the well-known reductant of ascorbic acid to drive LPMOs for cellulase enzymatic hydrolysis with equivalent cellulose oxidation efficiency and extremely lower H2O2 generation, avoiding the inactivation of enzymes. The maximum H2O2 yield from lignin-driven LPMO reaction was 75.8% lower than that from ascorbic acid-driven reaction. Therefore, by using the lignocellulosic mimicking model, we have elucidated the function of in-situ lignin in boosting enzymatic hydrolysis. Such understanding could significantly promote current utilization of LPMOs in lignocellulosic biorefinery.

Enzymatic treatment improves fast pyrolysis product selectivity of softwood and hardwood lignin.

Fast pyrolysis of lignin is still struggling in efficiency and scalable utilization. The low product selectivity thereby represents one of the most challenging issues. White-rot fungi have been widely used in bio-pretreatment of lignocellulosic biomass, where ligninolytic enzymes have been evidenced to modify lignin structures and enhance bio-refining efficiency. We thus treated lignin from both softwood (ginkgo) and hardwood (poplar) with enzymatic cocktail from white-rot fungus for fast pyrolysis. Both ginkgo and poplar lignin had much improved product selectivity at lower temperature after enzymatic modification, in particular, the 2-methoxy-phenol production from ginkgo lignin. Besides the improved product selectivity, the residue bio-char from pyrolysis had much improved surface area with more porous structures. Mechanistic study showed that the improvement of lignin pyrolysis products might attribute to demethoxylation and interunit linkage cleavage of lignin during enzymatic treatment. All these results highlighted that the product selectivity and bio-char performances have been synergistically improved by enzymatic treatment, which could thus pave a new way for enhancing fast pyrolysis efficiency. Overall, using softwood and hardwood lignin, this research has presented a new strategy using ligninolytic enzyme to modify lignin for synergistically improving product selectivity and bio-char performances, which opened up a new avenue for lignin valorization.

Heterologous expression and characterization of a xylanase and xylosidase from white rot fungi and their application in synergistic hydrolysis of lignocellulose.

Endo-xylanase and ß-xylosidase are the major enzymes for hemicellulose hydrolysis, which play a significant role in biomass conversion. In our previous work, the white-rot fungi Pleurotus ostreatus HAUCC 162 and Irpex lacteus CD2 were demonstrated to have strong ability in lignocellulose degradation, and the related lignin degradation enzymes were characterized. However, little was known about their hemicellulases. In this work, a novel endo-1, 4-xylanase and a ß-xylosidase from Pleurotus ostreatus HAUCC 162 and Irpex lacteus CD2 were heterologously expressed and characterized. The optima of pH and temperature were 5.0 and 55 °C for rXyn162, and 6.5 and 30 °C for rXylCD2. rXyn162 showed high tolerance to metal ions such as Ca2+, Cr3+, Zn2+, Na+, and Al3+. The recombinant rXyn162 and rXylCD2 exhibited synergistic hydrolysis of oat spelts xylan and sodium hydroxide pretreated cornstalk (SHPC), where the degree of synergy (DS) was 2.26 for SHPC hydrolysis. MALDI-TOF-MS and HPLC analysis showed that xylooligosaccharides (XOS) with small degrees of polymerization (DP2-DP4) were the major XOS hydrolyzate during SHPC degradation by rXyn162 and rXylCD2. In addition, rXyn162 and rXylCD2 could efficiently improve the hydrolysis of SHPC by commercial cellulase. The present study suggested the potential application of rXyn162 and rXylCD2 in the field of biomass pretreatment and biofuel production.

[Difficulties in registration for export of traditional Chinese medicines to EU under directive 2004/24/EC and countermeasures].

During the seven-year transitional period of European Union Directive 2004/24/EC, only a few of traditional Chinese medicines had been approved for registration. In other words, the EU directive has become an unavoidable registration barrier to hinder Chinese enterprises from entering EU market. By analyzing difficulties of enterprises in registration in EU and studying the only successful case in China--Di Ao Group, this article proposes countermeasures in the hope of providing effective reference for Chinese enterprises in expanding EU market, and promoting the internationalization progress of traditional Chinese medicine.