Injectable temperature-sensitive hydrogel facilitating endoscopic submucosal dissection.

Purpose: Early gastrointestinal tumors can be removed by endoscopic procedures. Endoscopic mucosal dissection (ESD) requires submucosal fluid injection to provide mucosal elevation and prevent intraoperative perforation. However, the clinically applied normal saline mucosal elevation height is low for a short time, which often requires multiple intraoperative injections that increase the inconvenience and procedure time. In addition, recently researched submucosal injection materials (SIM) suffer from complex preparation, poor economy, and poor biocompatibility. Therefore, there is an urgent need for a new type of SIM that can provide long, safe and effective mucosal elevation in support of the endoscopic procedures. Methods: The FS hydrogel is based on polyethylene-polypropylene glycol (F-127) mixed with sodium alginate (SA). The different physicochemical properties of FS hydrogels were characterized through various experiments. Afterward, various biosafety assessments were carried out. Finally, the performance of FS hydrogels was evaluated by in vitro submucosal injection and in vivo swine ESD. Results: The experimental results show that the FS hydrogel is liquid at room temperature, making it easy to inject, and when injected under the mucosa, it undergoes temperature-induced cross-linking, transforming from a liquid to a solid state to provide long-lasting mucosal augmentation. At the same time, the FS hydrogel exhibits controllable gelation, stability, and biocompatibility. The results of in vitro submucosal injections and in vivo ESD procedures showed that FS achieves high mucosal augmentation and provides good submucosal cushioning in the long term. Conclusion: In summary, the F-127/SA hydrogel is simple to synthesize, cost-effective, safe, easy to store, and able to assist ESD well from the perspective of practical clinical problems, indicating that the FS hydrogel can be an ideal potent submucosal injection substitution.

Reinvestigating the performance of artificial intelligence classification algorithms on COVID-19 X-Ray and CT images.

There are concerns that artificial intelligence (AI) algorithms may create underdiagnosis bias by mislabeling patient individuals with certain attributes (e.g., female and young) as healthy. Addressing this bias is crucial given the urgent need for AI diagnostics facing rapidly spreading infectious diseases like COVID-19. We find the prevalent AI diagnostic models show an underdiagnosis rate among specific patient populations, and the underdiagnosis rate is higher in some intersectional specific patient populations (for example, females aged 20-40 years). Additionally, we find training AI models on heterogeneous datasets (positive and negative samples from different datasets) may lead to poor model generalization. The model's classification performance varies significantly across test sets, with the accuracy of the better performance being over 40% higher than that of the poor performance. In conclusion, we developed an AI bias analysis pipeline to help researchers recognize and address biases that impact medical equality and ethics.

Analysis of the situations and influencing factors of public anxiety in China: based on Baidu index data.

Background: Anxiety disorders have emerged as one of the most prevalent mental health problems and health concerns. However, previous research has paid limited attention to measuring public anxiety from a broader perspective. Furthermore, while we know many factors that influence anxiety disorders, we still have an incomplete understanding of how these factors affect public anxiety. We aimed to quantify public anxiety from the perspective of Internet searches, and to analyze its spatiotemporal changing characteristics and influencing factors. Methods: This study collected Baidu Index from 2014 to 2022 in 31 provinces in mainland China to measure the degree of public anxiety based on the Baidu Index from 2014 to 2022. The spatial autocorrelation analysis method was used to study the changing trends and spatial distribution characteristics of public anxiety. The influencing factors of public anxiety were studied using spatial statistical modeling methods. Results: Empirical analysis shows that the level of public anxiety in my country has continued to rise in recent years, with significant spatial clustering characteristics, especially in the eastern and central-southern regions. In addition, we constructed ordinary least squares (OLS) and geographically weighted regression (GWR) spatial statistical models to examine the relationship between social, economic, and environmental factors and public anxiety levels. We found that the GWR model that considers spatial correlation and dependence is significantly better than the OLS model in terms of fitting accuracy. Factors such as the number of college graduates, Internet traffic, and urbanization rate are significantly positively correlated with the level of public anxiety. Conclusion: Our research results draw attention to public anxiety among policymakers, highlighting the necessity for a more extensive examination of anxiety issues, especially among university graduates, by the public and relevant authorities.

Harnessing artificial intelligence for prostate cancer management.

Prostate cancer (PCa) is a common malignancy in males. The pathology review of PCa is crucial for clinical decision-making, but traditional pathology review is labor intensive and subjective to some extent. Digital pathology and whole-slide imaging enable the application of artificial intelligence (AI) in pathology. This review highlights the success of AI in detecting and grading PCa, predicting patient outcomes, and identifying molecular subtypes. We propose that AI-based methods could collaborate with pathologists to reduce workload and assist clinicians in formulating treatment recommendations. We also introduce the general process and challenges in developing AI pathology models for PCa. Importantly, we summarize publicly available datasets and open-source codes to facilitate the utilization of existing data and the comparison of the performance of different models to improve future studies.

Xiaoqing: A Q&A model for glaucoma based on LLMs.

Glaucoma is one of the leading cause of blindness worldwide. Individuals affected by glaucoma, including patients and their family members, frequently encounter a deficit in dependable support beyond the confines of clinical environments. Seeking advice via the internet can be a difficult task due to the vast amount of disorganized and unstructured material available on these sites, nevertheless. This research explores how Large Language Models (LLMs) can be leveraged to better serve medical research and benefit glaucoma patients. We introduce Xiaoqing, a Natural Language Processing (NLP) model specifically tailored for the glaucoma field, detailing its development and deployment. To evaluate its effectiveness, we conducted two forms of experiments: comparative and experiential. In the comparative analysis, we presented 22 glaucoma-related questions in simplified Chinese to three medical NLP models (Xiaoqing LLMs, HuaTuo, Ivy GPT) and two general models (ChatGPT-3.5 and ChatGPT-4), covering a range of topics from basic glaucoma knowledge to treatment, surgery, research, management standards, and patient lifestyle. Responses were assessed for informativeness and readability. The experiential experiment involved glaucoma patients and non-patients interacting with Xiaoqing, collecting and analyzing their questions and feedback on the same criteria. The findings demonstrated that Xiaoqing notably outperformed the other models in terms of informativeness and readability, suggesting that Xiaoqing is a significant advancement in the management and treatment of glaucoma in China. We also provide a Web-based version of Xiaoqing, allowing readers to directly experience its functionality. The Web-based Xiaoqing is available at https://qa.glaucoma-assistant.com//qa.

Prediction and visualization map for physicochemical indices of kiwifruits by hyperspectral imaging.

Soluble solid content (SSC), firmness, and color (L*, a*, and b*) are important physicochemical indices for assessing the quality and maturity of kiwifruits. Therefore, this research aimed to realize the nondestructive detection and visualization map for the physicochemical indices of kiwifruits at different maturity stages by hyperspectral imaging coupled with the chemometrics. To further improve the detection accuracy and working efficiency of the models, competitive adaptive reweighted sampling (CARS) and successive projection algorithm were employed to choose feature wavelengths for predicting the physicochemical indices of kiwifruits. Multiple linear regression (MLR) was designed to develop simplified detection models based on feature wavelengths for determining the physicochemical indices of kiwifruits. The results showed that 32, 18, 26, 29, and 32 feature wavelengths were extracted from 256 full wavelengths to predict the SSC, firmness, L*, a*, and b*, respectively, with the CARS algorithm. Not only was the working efficiency of the CARS-MLR model improved, but the prediction accuracy of the CARS-MLR model for determining the physicochemical indices was also at its relative best. The residual predictive deviations of the CARS-MLR model for determining the SSC, firmness, L*, a*, and b* were 3.09, 2.90, 2.32, 2.74, and 2.91, respectively, which were all above 2.3. Compared with the model based on the full spectra, the CARS-MLR model could be used to predict the physicochemical indices of kiwifruits. Finally, the visualization map for the physicochemical indices of kiwifruits at different maturity stages was generated by calculating the spectral response of each pixel on the kiwifruit samples with the CARS-MLR model. This made the detection for the physicochemical indices of kiwifruits more intuitive. This study demonstrates that hyperspectral imaging coupled with the chemometrics is promising for the nondestructive detection and visualization map for the physicochemical indices of kiwifruits, and also provides a novel theoretical basis for the nondestructive detection of kiwifruit quality.

BCL-XL inhibitors enhance the apoptotic efficacy of BRAF inhibitors in BRAFV600E colorectal cancer.

Metastatic BRAFV600E colorectal cancer (CRC) carries an extremely poor prognosis and is in urgent need of effective new treatments. While the BRAFV600E inhibitor encorafenib in combination with the EGFR inhibitor cetuximab (Enc+Cet) was recently approved for this indication, overall survival is only increased by 3.6 months and objective responses are observed in only 20% of patients. We have found that a limitation of Enc+Cet treatment is the failure to efficiently induce apoptosis in BRAFV600E CRCs, despite inducing expression of the pro-apoptotic protein BIM and repressing expression of the pro-survival protein MCL-1. Here, we show that BRAFV600E CRCs express high basal levels of the pro-survival proteins MCL-1 and BCL-XL, and that combining encorafenib with a BCL-XL inhibitor significantly enhances apoptosis in BRAFV600E CRC cell lines. This effect was partially dependent on the induction of BIM, as BIM deletion markedly attenuated BRAF plus BCL-XL inhibitor-induced apoptosis. As thrombocytopenia is an established on-target toxicity of BCL-XL inhibition, we also examined the effect of combining encorafenib with the BCL-XL -targeting PROTAC DT2216, and the novel BCL-2/BCL-XL inhibitor dendrimer conjugate AZD0466. Combining encorafenib with DT2216 significantly increased apoptosis induction in vitro, while combining encorafenib with AZD0466 was well tolerated in mice and further reduced growth of BRAFV600E CRC xenografts compared to either agent alone. Collectively, these findings demonstrate that combined BRAF and BCL-XL inhibition significantly enhances apoptosis in pre-clinical models of BRAFV600E CRC and is a combination regimen worthy of clinical investigation to improve outcomes for these patients.

Significance of N6-methyladenosine RNA methylation regulators in diagnosis and subtype classification of primary Sjögren's syndrome.

The importance of N6-methyladenine (m6A) in mRNA metabolism, physiology, pathology and other life processes is well recognized. However, the exact role of m6A regulators in primary Sjögren's syndrome (PSS) remains unclear. In this study, we used bioinformatics and machine learning random forest approach to screen eight key m6A regulators from the Gene Expression Omnibus GSE7451, GSE40611 and GSE84844 datasets. An accurate nomogram model for predicting PSS risk was established based on these regulators. And using consensus clustering, patients diagnosed with PSS were classified into two different m6A patterns. We found that patients in group B had higher m6A scores compared to those in group A: furthermore, both groups were closely related to immunity and possibly to other diseases. These results emphasise the important role of m6A regulators in the pathogenesis of PSS. Our study of m6A patterns may inform future immunotherapy strategies for PSS.

Advanced gastrointestinal stromal tumor: reliable classification of imatinib plasma trough concentration via machine learning.

AIM: Patients with advanced gastrointestinal stromal tumors (GISTs) exhibiting an imatinib plasma trough concentration (IM Cmin) under 1100 ng/ml may show a reduced drug response rate, leading to the suggestion of monitoring for IM Cmin. Consequently, the objective of this research was to create a customized IM Cmin classification model for patients with advanced GISTs from China. METHODS: Initial data and laboratory indicators from patients with advanced GISTs were gathered, and the above information was segmented into a training set, validation set, and testing set in a 6:2:2 ratio. Key variables associated with IM Cmin were identified to construct the classification model using the least absolute shrinkage and selection operator (LASSO) regression and forward stepwise binary logistic regression. Within the training and validation sets, nine ML classification models were constructed via the resampling method and underwent comparison through the Brier scores, the areas under the receiver-operating characteristic curve (AUROC), the decision curve, and the precision-recall (AUPR) curve to determine the most suitable model for this dataset. Two methods of internal validation were used to assess the most suitable model's classification performance: tenfold cross-validation and random split-sample validation (test set), and the value of the test set AUROC was used to evaluate the model's classification performance. RESULTS: Six key variables (gender, daily IM dose, metastatic site, red blood cell count, platelet count, and percentage of neutrophils) were ultimately selected to construct the classification model. In the validation set, it is found by comparison that the Extreme Gradient Boosting (XGBoost) model has the largest AUROC, the lowest Brier score, the largest area under the decision curve, and the largest AUPR value. Furthermore, as evaluated via internal verification, it also performed well in the test set (AUROC = 0.725). CONCLUSION: For patients with advanced GISTs who receive IM, initial data and laboratory indicators could be used to accurately estimate whether the IM Cmin is below 1100 ng/ml. The XGBoost model may stand a chance to assist clinicians in directing the administration of IM.

Association of LDL-C/HDL-C ratio with coronary heart disease: A meta-analysis.

BACKGROUND: Coronary heart disease (CHD) is a common heart disease and a leading cause of death in developed countries and some developing countries such as China. It is recognized as a multifactorial disease, with dyslipidemia being closely associated with the progression of coronary atherosclerosis. Numerous studies have confirmed the relationship between a single indicator of low-density lipoprotein cholesterol (LDL-C) or high-density lipoprotein cholesterol (HDL-C) and CHD. However, the association between LDL-C to HDL-C ratio (LHR) and CHD remains unclear. This study aimed to comprehensively explore the association between LHR and CHD. METHODS: This meta-analysis was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses. PubMed, Embase, Web of Science, and China National Knowledge Infrastructure databases were comprehensively searched up to June 15, 2023, to find the studies that indicated the connection between LHR and CHD. A total of 12 published studies were selected. The random-effects model was used to pool the data and mean difference (MD), and the 95% confidence intervals (CI) were taken as the overall outcome. No language restrictions existed in the study selection. The Review Manager 5.4 and Stata 12 were used to analyze the data. RESULTS: Twelve high-quality clinical studies involving 5544 participants, including 3009 patients with CHD, were enrolled in the meta-analysis. The findings revealed that the LHR was higher by 0.65 in patients with CHD than in those without CHD (MD, 0.65; 95% CI, 0.50-0.80). CONCLUSION: The LHR was found to be positively correlated with CHD, suggesting that it may serve as a potential indicator of CHD.

Establishing a 3D culture system for early organogenesis of monkey embryos ex vivo and single-cell transcriptome analysis of cultured embryos.

Creating in vitro culture platforms for monkey embryos is crucial for understanding the initial 4 weeks of early primate embryogenesis. Here, we present a protocol to culture cynomolgus monkey embryos in vitro for 25 days post-fertilization and to delineate the key developmental events of gastrulation and early organogenesis. We describe steps for culturing with a 3D system, immunofluorescence analysis, single-cell RNA sequencing, and bioinformatic analysis. For complete details on the use and execution of this protocol, please refer to Gong et al. (2023).1.

Quantification of Hypsarrhythmia in Infantile Spasmatic EEG: A Large Cohort Study.

Infantile spasms (IS) is a neurological disorder causing mental and/or developmental retardation in many infants. Hypsarrhythmia is a typical symptom in the electroencephalography (EEG) signals with IS. Long-term EEG/video monitoring is most frequently employed in clinical practice for IS diagnosis, from which manual screening of hypsarrhythmia is time consuming and lack of sufficient reliability. This study aims to identify potential biomarkers for automatic IS diagnosis by quantitative analysis of the EEG signals. A large cohort of 101 IS patients and 155 healthy controls (HC) were involved. Typical hypsarrhythmia and non-hypsarrhythmia EEG signals were annotated, and normal EEG were randomly picked from the HC. Root mean square (RMS), teager energy (TE), mean frequency, sample entropy (SamEn), multi-channel SamEn, multi-scale SamEn, and nonlinear correlation coefficient were computed in each sub-band of the three EEG signals, and then compared using either a one-way ANOVA or a Kruskal-Wallis test (based on their distribution) and the receiver operating characteristic (ROC) curves. The effects of infant age on these features were also investigated. For most of the employed features, significant ( ) differences were observed between hypsarrhythmia EEG and non-hypsarrhythmia EEG or HC, which seem to increase with increased infant age. RMS and TE produce the best classification in the delta and theta bands, while entropy features yields the best performance in the gamma band. Our study suggests RMS and TE (delta and theta bands) and entropy features (gamma band) to be promising biomarkers for automatic detection of hypsarrhythmia in long-term EEG monitoring. The findings of our study indicate the feasibility of automated IS diagnosis using artificial intelligence.

Exploring the causal association between rheumatoid arthritis and the risk of cervical cancer: a two-sample Mendelian randomization study.

OBJECTIVE: Whether rheumatoid arthritis patients have an increased risk of cervical cancer remains controversial, and further research is needed on this clinical question. This study aims to investigate the association between rheumatoid arthritis and the susceptibility to cervical cancer by employing Mendelian randomization methodology, utilizing the extensive dataset from human genome-wide association data analysis. METHODS: The publicly accessible MR base database was utilized to obtain the complete genome, relevant research findings, and summarized data pertaining to rheumatoid arthritis and cervical cancer. Genetic tool variables, specifically single-nucleotide polymorphisms closely linked to rheumatoid arthritis, were chosen for analysis. Four methods, namely inverse variance weighted analysis, weighted median analysis, weighted mode, and MR-Egger regression, were employed. Statistical analysis was conducted to explore the potential association between rheumatoid arthritis and susceptibility to cervical cancer. RESULTS: The results of the inverse variance weighted analysis (OR = 1.096, 95% CI: 1.018-1.180, P = 0.015) indicate a significant causal relationship between rheumatoid arthritis and an increased risk of cervical cancer. Furthermore, the absence of horizontal pleiotropic effects (MR-Egger intercept = 0.00025, P = 0.574) and heterogeneity (QEgger = 2.239, I2Egger = 0.225, PEgger = 0.268, QIVW = 2.734, I2IVW = 0.220, PIVW = 0.999) suggests that the observed association is not influenced by confounding factors. Sensitivity analysis and other statistical methods also support the conclusion that genetic pleiotropy does not introduce bias to the findings. CONCLUSION: There is a causal relationship between rheumatoid arthritis and the occurrence of cervical cancer. People with rheumatoid arthritis is one of the high-risk groups for early screening of cervical cancer. The IL-18 may play a significant role in elevating the risk of cervical cancer among rheumatoid arthritis patients.

Integrated machine learning methods with oversampling technique for regional suitability prediction of waste-to-energy incineration projects.

China's tiered strategy to enhance county-level waste incineration for energy aligns with the sustainable development goals (SDGs), emphasizing the need for comprehensive assessments of waste-to-energy (WtE) plant suitability. Traditional assessment methodologies face challenges, particularly in suggesting innovative site alternatives, adapting to new data sets, and their dependence on strict assumptions. This study introduced enhancements in three pivotal dimensions. Methodologically, it leverages data-driven machine learning (ML) approaches to capture the complex relationships essential for site selection, reducing dependency on strict assumptions. In terms of predictive performance, the integration of oversampling with stacked ensemble models enhances the diversity and generalizability of ML models. The area under curve (AUC) scores from four ML models, enhanced by the oversampled dataset, demonstrated significant improvements compared to the original dataset. The stacking model excelled, achieving a score of 92%. It also led in overall Precision and Recall, reaching 85.2% and 85.08% respectively. Nevertheless, a noticeable discrepancy existed in Precision and Recall for positive classes. The stacking model topped Precision scores at 83.1%, followed by eXtreme Gradient Boosting (XGBoost) (82.61%). In terms of Recall, XGBoost recorded the lowest at 85.07%, while the other three classifiers all marked 88.06%. From an industry applicability standpoint, the stacking model provides innovative location alternatives and demonstrates adaptability in Hunan province, offering a reusable tool for WtE location. In conclusion, this study not only enhances the methodological aspects of WtE site selection but also provides practical and adaptable solutions, contributing positively to sustainable waste management practices.

ERNet: Edge Regularization Network for Cerebral Vessel Segmentation in Digital Subtraction Angiography Images.

Stroke is a leading cause of disability and fatality in the world, with ischemic stroke being the most common type. Digital Subtraction Angiography images, the gold standard in the operation process, can accurately show the contours and blood flow of cerebral vessels. The segmentation of cerebral vessels in DSA images can effectively help physicians assess the lesions. However, due to the disturbances in imaging parameters and changes in imaging scale, accurate cerebral vessel segmentation in DSA images is still a challenging task. In this paper, we propose a novel Edge Regularization Network (ERNet) to segment cerebral vessels in DSA images. Specifically, ERNet employs the erosion and dilation processes on the original binary vessel annotation to generate pseudo-ground truths of False Negative and False Positive, which serve as constraints to refine the coarse predictions based on their mapping relationship with the original vessels. In addition, we exploit a Hybrid Fusion Module based on convolution and transformers to extract local features and build long-range dependencies. Moreover, to support and advance the open research in the field of ischemic stroke, we introduce FPDSA, the first pixel-level semantic segmentation dataset for cerebral vessels. Extensive experiments on FPDSA illustrate the leading performance of our ERNet.

Neuroprotective Effects of CXCR2 Antagonist SB332235 on Traumatic Brain Injury Through Suppressing NLRP3 Inflammasome.

The inflammatory process mediated by nucleotide-binding oligomerization domain (NOD)-like receptor family pyrin domain comprising 3 (NLRP3) inflammasome plays a predominant role in the neurological dysfunction following traumatic brain injury (TBI). SB332235, a highly selective antagonist of chemokine receptor 2 (CXCR2), has been demonstrated to exhibit anti-inflammatory properties and improve neurological outcomes in the central nervous system. We aimed to determine the neuroprotective effects of SB332235 in the acute phase after TBI in mice and to elucidate its underlying mechanisms. Male C57BL/6J animals were exposed to a controlled cortical impact, then received 4 doses of SB332235, with the first dose administered at 30 min after TBI, followed by additional doses at 6, 24, and 30 h. Neurological defects were assessed by the modified neurological severity score, while the motor function was evaluated using the beam balance and open field tests. Cognitive performance was evaluated using the novel object recognition test. Brain tissues were collected for pathological, Western blot, and immunohistochemical analyses. The results showed that SB332235 significantly ameliorated TBI-induced deficits, including motor and cognitive impairments. SB332235 administration suppressed expression of both CXCL1 and CXCR2 in TBI. Moreover, SB332235 substantially mitigated the augmented expression levels and activation of the NLRP3 inflammasome within the peri-contusional cortex induced by TBI. This was accompanied by the blocking of subsequent production of pro-inflammatory cytokines. Additionally, SB332235 hindered microglial activity induced by TBI. These findings confirmed the neuroprotective effects of SB332235 against TBI, and the involved mechanisms were in part due to the suppression of NLRP3 inflammasome activity. This study suggests that SB332235 may act as an anti-inflammatory agent to improve functional outcomes in brain injury when applied clinically.

LYSTO: The Lymphocyte Assessment Hackathon and Benchmark Dataset.

We introduce LYSTO, the Lymphocyte Assessment Hackathon, which was held in conjunction with the MICCAI 2019 Conference in Shenzhen (China). The competition required participants to automatically assess the number of lymphocytes, in particular T-cells, in images of colon, breast, and prostate cancer stained with CD3 and CD8 immunohistochemistry. Differently from other challenges setup in medical image analysis, LYSTO participants were solely given a few hours to address this problem. In this paper, we describe the goal and the multi-phase organization of the hackathon; we describe the proposed methods and the on-site results. Additionally, we present post-competition results where we show how the presented methods perform on an independent set of lung cancer slides, which was not part of the initial competition, as well as a comparison on lymphocyte assessment between presented methods and a panel of pathologists. We show that some of the participants were capable to achieve pathologist-level performance at lymphocyte assessment. After the hackathon, LYSTO was left as a lightweight plug-and-play benchmark dataset on grand-challenge website, together with an automatic evaluation platform.

Endoscopic resection of giant esophageal subepithelial lesions: experience from a large tertiary center.

BACKGROUND AND AIMS: Increased reports on endoscopic resection (ER) of esophageal giant subepithelial lesions (g-SELs) have emerged in recent years. The aim of this study was to evaluate the efficacy, technical difficulty, and safety through our single-center experience. METHODS: Seventy-five patients with g-SELs undergoing endoscopic resection were included in the training set. Clinicopathologic features, procedure-related characteristics, postprocedural outcomes, and follow-up data were analyzed. A predictive nomogram model for procedural difficulty was proposed based on the multivariable logistic regression analysis. Internal and external validations were conducted to verify the model performance. RESULTS: The overall en bloc resection rate was 93.3%. Intraoperative and postoperative adverse events occurred in 7 (9.3%) and 13 (17.3%) patients, respectively. No recurrence or metastasis was observed. Thirty-two (42.7%) patients underwent a difficult procedure. Age (adjusted odds ratio [aOR], .915; P = .004), maximal tumor diameter ≥8 cm (aOR, 9.896; P = .009), irregular shape (aOR, 4.081; P = .053), extraluminal growth pattern (aOR, 5.419; P = .011), and submucosal tunneling endoscopic resection (aOR, .109; P = .042) were found to be statistically or clinically significant factors for predicting endoscopic resection difficulty, based on which a nomogram model was developed. Internal and external validations of the nomogram via receiver-operating characteristic curves and calibration curves achieved favorable results. CONCLUSIONS: Endoscopic resection serves as a promising therapeutic option for esophageal g-SELs. A younger patient age, large tumor size, irregular shape, and extraluminal growth may indicate increased endoscopic resection difficulty, whereas a submucosal tunneling endoscopic resection procedure tends to be of lower difficulty. Our nomogram model performs well for predicting endoscopic resection difficulty for esophageal g-SELs.

Light mixed-supervised segmentation for 3D medical image data.

BACKGROUND: Accurate 3D semantic segmentation models are essential for many clinical applications. To train a model for 3D segmentation, voxel-level annotation is necessary, which is expensive to obtain due to laborious work and privacy protection. To accurately annotate 3D medical data, such as MRI, a common practice is to annotate the volumetric data in a slice-by-slice contouring way along principal axes. PURPOSE: In order to reduce the annotation effort in slices, weakly supervised learning with a bounding box (Bbox) was proposed to leverage the discriminating information via a tightness prior assumption. Nevertheless, this method requests accurate and tight Bboxes, which will significantly drop the performance when tightness is not held, that is when a relaxed Bbox is applied. Therefore, there is a need to train a stable model based on relaxed Bbox annotation. METHODS: This paper presents a mixed-supervised training strategy to reduce the annotation effort for 3D segmentation tasks. In the proposed approach, a fully annotated contour is only required for a single slice of the volume. In contrast, the rest of the slices with targets are annotated with relaxed Bboxes. This mixed-supervised method adopts fully supervised learning, relaxed Bbox prior, and contrastive learning during the training, which ensures the network exploits the discriminative information of the training volumes properly. The proposed method was evaluated on two public 3D medical imaging datasets (MRI prostate dataset and Vestibular Schwannoma [VS] dataset). RESULTS: The proposed method obtained a high segmentation Dice score of 85.3% on an MRI prostate dataset and 83.3% on a VS dataset with relaxed Bbox annotation, which are close to a fully supervised model. Moreover, with the same relaxed Bbox annotations, the proposed method outperforms the state-of-the-art methods. More importantly, the model performance is stable when the accuracy of Bbox annotation varies. CONCLUSIONS: The presented study proposes a method based on a mixed-supervised learning method in 3D medical imaging. The benefit will be stable segmentation of the target in 3D images with low accurate annotation requirement, which leads to easier model training on large-scale datasets.

Synthesis-based imaging-differentiation representation learning for multi-sequence 3D/4D MRI.

Multi-sequence MRIs can be necessary for reliable diagnosis in clinical practice due to the complimentary information within sequences. However, redundant information exists across sequences, which interferes with mining efficient representations by learning-based models. To handle various clinical scenarios, we propose a sequence-to-sequence generation framework (Seq2Seq) for imaging-differentiation representation learning. In this study, not only do we propose arbitrary 3D/4D sequence generation within one model to generate any specified target sequence, but also we are able to rank the importance of each sequence based on a new metric estimating the difficulty of a sequence being generated. Furthermore, we also exploit the generation inability of the model to extract regions that contain unique information for each sequence. We conduct extensive experiments using three datasets including a toy dataset of 20,000 simulated subjects, a brain MRI dataset of 1251 subjects, and a breast MRI dataset of 2101 subjects, to demonstrate that (1) top-ranking sequences can be used to replace complete sequences with non-inferior performance; (2) combining MRI with our imaging-differentiation map leads to better performance in clinical tasks such as glioblastoma MGMT promoter methylation status prediction and breast cancer pathological complete response status prediction. Our code is available at https://github.com/fiy2W/mri_seq2seq.