Search

Abstract Background Data accuracy is essential for scientific research and policy development. The National Violent Death Reporting System (NVDRS) data is widely used for discovering the patterns and causing factors of death. Recent studies suggested the annotation inconsistencies within the NVDRS and the potential impact on erroneous suicide-circumstance attributions. Methods We present an empirical Natural Language Processing (NLP) approach to detect annotation inconsistencies and adopt a cross-validation-like paradigm to identify possible label errors. We analyzed 267,804 suicide death incidents between 2003 and 2020 from the NVDRS. We measured annotation inconsistency by the degree of changes in the F-1 score. Results Our results show that incorporating the target state’s data into training the suicide-circumstance classifier brings an increase of 5.4% to the F-1 score on the target state’s test set and a decrease of 1.1% on other states’ test set. Conclusions To conclude, we present an NLP framework to detect the annotation inconsistencies, show the effectiveness of identifying and rectifying possible label errors, and eventually propose an improvement solution to improve the coding consistency of human annotators.

Abstract Major symbiotic organisms have evolved to establish beneficial relationships with hosts. However, understanding the interactions between symbionts and insect hosts, particularly for their roles in defense against pathogens, is still limited. In a previous study, we proposed that the fungus Metarhizium anisopliae can infect the brown planthopper Nilaparvata lugens, a harmful pest for rice crops. To expand on this, we investigated changes in N. lugens’ intestinal commensal community after M. anisopliae infection and identified key gut microbiotas involved. Our results showed significant alterations in gut microbiota abundance and composition at different time points following infection with M. anisopliae. Notably, certain symbionts, like Acinetobacter baumannii, exhibited significant variations in response to the fungal infection. The decrease in these symbionts had a considerable impact on the insect host’s survival. Interestingly, reintroducing A. baumannii enhanced the host’s resistance to M. anisopliae, emphasizing its role in pathogen defense. Additionally, A. baumannii stimulated host immune responses, as evidenced by increased expression of immune genes after reintroduction. Overall, our findings highlight the significance of preserving a stable gut microbial community for the survival of insects. In specific conditions, the symbiotic microorganism A. baumannii can enhance the host’s ability to resist entomopathogenic pathogens through immune regulation.

Abstract With generative artificial intelligence (GenAI), particularly large language models (LLMs), continuing to make inroads in healthcare, assessing LLMs with human evaluations is essential to assuring safety and effectiveness. This study reviews existing literature on human evaluation methodologies for LLMs in healthcare across various medical specialties and addresses factors such as evaluation dimensions, sample types and sizes, selection, and recruitment of evaluators, frameworks and metrics, evaluation process, and statistical analysis type. Our literature review of 142 studies shows gaps in reliability, generalizability, and applicability of current human evaluation practices. To overcome such significant obstacles to healthcare LLM developments and deployments, we propose QUEST, a comprehensive and practical framework for human evaluation of LLMs covering three phases of workflow: Planning, Implementation and Adjudication, and Scoring and Review. QUEST is designed with five proposed evaluation principles: Quality of Information, Understanding and Reasoning, Expression Style and Persona, Safety and Harm, and Trust and Confidence.

Abstract Large language models (LLMs) hold great promise in summarizing medical evidence. Most recent studies focus on the application of proprietary LLMs. Using proprietary LLMs introduces multiple risk factors, including a lack of transparency and vendor dependency. While open-source LLMs allow better transparency and customization, their performance falls short compared to the proprietary ones. In this study, we investigated to what extent fine-tuning open-source LLMs can further improve their performance. Utilizing a benchmark dataset, MedReview, consisting of 8161 pairs of systematic reviews and summaries, we fine-tuned three broadly-used, open-sourced LLMs, namely PRIMERA, LongT5, and Llama-2. Overall, the performance of open-source models was all improved after fine-tuning. The performance of fine-tuned LongT5 is close to GPT-3.5 with zero-shot settings. Furthermore, smaller fine-tuned models sometimes even demonstrated superior performance compared to larger zero-shot models. The above trends of improvement were manifested in both a human evaluation and a larger-scale GPT4-simulated evaluation.

Abstract Deep learning has enabled breakthroughs in automated diagnosis from medical imaging, with many successful applications in ophthalmology. However, standard medical image classification approaches only assess disease presence at the time of acquisition, neglecting the common clinical setting of longitudinal imaging. For slow, progressive eye diseases like age-related macular degeneration (AMD) and primary open-angle glaucoma (POAG), patients undergo repeated imaging over time to track disease progression and forecasting the future risk of developing a disease is critical to properly plan treatment. Our proposed Longitudinal Transformer for Survival Analysis (LTSA) enables dynamic disease prognosis from longitudinal medical imaging, modeling the time to disease from sequences of fundus photography images captured over long, irregular time periods. Using longitudinal imaging data from the Age-Related Eye Disease Study (AREDS) and Ocular Hypertension Treatment Study (OHTS), LTSA significantly outperformed a single-image baseline in 19/20 head-to-head comparisons on late AMD prognosis and 18/20 comparisons on POAG prognosis. A temporal attention analysis also suggested that, while the most recent image is typically the most influential, prior imaging still provides additional prognostic value.

Abstract Recent studies indicate that Generative Pre-trained Transformer 4 with Vision (GPT-4V) outperforms human physicians in medical challenge tasks. However, these evaluations primarily focused on the accuracy of multi-choice questions alone. Our study extends the current scope by conducting a comprehensive analysis of GPT-4V’s rationales of image comprehension, recall of medical knowledge, and step-by-step multimodal reasoning when solving New England Journal of Medicine (NEJM) Image Challenges—an imaging quiz designed to test the knowledge and diagnostic capabilities of medical professionals. Evaluation results confirmed that GPT-4V performs comparatively to human physicians regarding multi-choice accuracy (81.6% vs. 77.8%). GPT-4V also performs well in cases where physicians incorrectly answer, with over 78% accuracy. However, we discovered that GPT-4V frequently presents flawed rationales in cases where it makes the correct final choices (35.5%), most prominent in image comprehension (27.2%). Regardless of GPT-4V’s high accuracy in multi-choice questions, our findings emphasize the necessity for further in-depth evaluations of its rationales before integrating such multimodal AI models into clinical workflows.

Abstract Background CFAP65 (cilia and flagella associated protein 65) is a fundamental protein in the development and formation of ciliated flagella, but few studies have focused on its role in cancer. This study aimed to investigate the prognostic significance of CFAP65 in colon cancer. Methods The functionally enriched genes related to CFAP65 were analyzed through the Gene Ontology (GO) database. Subsequently, CFAP65 expression levels in colon cancer were evaluated by reverse transcription and quantitative polymerase chain reaction (RT-qPCR) and immunoblotting in 20 pairs of frozen samples, including tumors and their matched paratumor tissue. Furthermore, protein expression of CFAP65 in 189 colon cancer patients were assessed via immunohistochemical staining. The correlations between CFAP65 expression and clinical features as well as long-term survival were statistically analyzed. Results CFAP65-related genes are significantly enriched on cellular processes of cell motility, ion channels, and GTPase-associated signaling. The expression of CFAP65 was significantly higher in colon cancer tissue compared to paratumor tissue. The proportion of high expression and low expression of CFAP65 in the clinical samples of colon cancer were 61.9% and 38.1%, respectively, and its expression level was not associated with the clinical parameters including gender, age, tumor location, histological differentiation, tumor stage, vascular invasion and mismatch repair deficiency. The five-year disease-free survival rate of the patients with CFAP65 low expression tumors was significantly lower than that those with high expression tumors (56.9% vs. 72.6%, P = 0.03), but the overall survival rate has no significant difference (69% vs. 78.6%, P = 0.171). The cox hazard regression analysis model showed that CFAP65 expression, tumor stage and tumor location were independent prognostic factors. Conclusions In conclusion, we demonstrate CFAP65 is a potential predictive marker for tumor progression in colon cancer.

Protease-activated receptor-2 (PAR2) is a class-A G protein-coupled receptor (GPCR) activated by serine proteases and is expressed by multiple tissues, including the skin. PAR2 is involved in the skin inflammatory response, promoting Th2 inflammation, delaying skin barrier repair, and affecting the differentiation of keratinocytes. It also participates in the transmission of itch and pain sensations in the skin. Increasing evidence indicates that PAR2 plays an important role in the pathogenesis of inflammatory skin diseases such as acne vulgaris, rosacea, psoriasis, and atopic dermatitis. Additional focus will be placed on potential targeted therapies based on PAR2. The Goal of this review is to outline the emerging effects of PAR2 activation in inflammatory skin disease and highlight the promise of PAR2 modulators.

This paper will focus on analyzing the argument with appealing to nature against synthetic biology and provide a counter-argument against it through demonstrating the ambiguity of the concept of nature, denying the existence of a morally significant line between natural and non/unnatural, and disproving the allegations against synthetic biology raised by the argument appealing to nature. The paper consists of two parts following a brief introduction. The first part will describe the argument appealing to nature against synthetic biology, and identify the deficiencies of the argument per se, e.g., the ambiguity of the concept ‘nature’; and the problems in the morally significant line between the natural and the non/unnatural. The second part will discuss the allegations to synthetic biology stemming from this argument, e.g., committing metaphysical and ethical mistakes, and doing possible harms to the environment.

Abstract In 2020, the U.S. Department of Defense officially disclosed a set of ethical principles to guide the use of Artificial Intelligence (AI) technologies on future battlefields. Despite stark differences, there are core similarities between the military and medical service. Warriors on battlefields often face life-altering circumstances that require quick decision-making. Medical providers experience similar challenges in a rapidly changing healthcare environment, such as in the emergency department or during surgery treating a life-threatening condition. Generative AI, an emerging technology designed to efficiently generate valuable information, holds great promise. As computing power becomes more accessible and the abundance of health data, such as electronic health records, electrocardiograms, and medical images, increases, it is inevitable that healthcare will be revolutionized by this technology. Recently, generative AI has garnered a lot of attention in the medical research community, leading to debates about its application in the healthcare sector, mainly due to concerns about transparency and related issues. Meanwhile, questions around the potential exacerbation of health disparities due to modeling biases have raised notable ethical concerns regarding the use of this technology in healthcare. However, the ethical principles for generative AI in healthcare have been understudied. As a result, there are no clear solutions to address ethical concerns, and decision-makers often neglect to consider the significance of ethical principles before implementing generative AI in clinical practice. In an attempt to address these issues, we explore ethical principles from the military perspective and propose the “GREAT PLEA” ethical principles, namely Governability, Reliability, Equity, Accountability, Traceability, Privacy, Lawfulness, Empathy, and Autonomy for generative AI in healthcare. Furthermore, we introduce a framework for adopting and expanding these ethical principles in a practical way that has been useful in the military and can be applied to healthcare for generative AI, based on contrasting their ethical concerns and risks. Ultimately, we aim to proactively address the ethical dilemmas and challenges posed by the integration of generative AI into healthcare practice.

Abstract Deep learning has become a popular tool for computer-aided diagnosis using medical images, sometimes matching or exceeding the performance of clinicians. However, these models can also reflect and amplify human bias, potentially resulting inaccurate missed diagnoses. Despite this concern, the problem of improving model fairness in medical image classification by deep learning has yet to be fully studied. To address this issue, we propose an algorithm that leverages the marginal pairwise equal opportunity to reduce bias in medical image classification. Our evaluations across four tasks using four independent large-scale cohorts demonstrate that our proposed algorithm not only improves fairness in individual and intersectional subgroups but also maintains overall performance. Specifically, the relative change in pairwise fairness difference between our proposed model and the baseline model was reduced by over 35%, while the relative change in AUC value was typically within 1%. By reducing the bias generated by deep learning models, our proposed approach can potentially alleviate concerns about the fairness and reliability of image-based computer-aided diagnosis.

Abstract Recent advances in large language models (LLMs) have demonstrated remarkable successes in zero- and few-shot performance on various downstream tasks, paving the way for applications in high-stakes domains. In this study, we systematically examine the capabilities and limitations of LLMs, specifically GPT-3.5 and ChatGPT, in performing zero-shot medical evidence summarization across six clinical domains. We conduct both automatic and human evaluations, covering several dimensions of summary quality. Our study demonstrates that automatic metrics often do not strongly correlate with the quality of summaries. Furthermore, informed by our human evaluations, we define a terminology of error types for medical evidence summarization. Our findings reveal that LLMs could be susceptible to generating factually inconsistent summaries and making overly convincing or uncertain statements, leading to potential harm due to misinformation. Moreover, we find that models struggle to identify the salient information and are more error-prone when summarizing over longer textual contexts.

Abstract The objective of this study was to investigate the potential association between the use of four frequently prescribed drug classes, namely antihypertensive drugs, statins, selective serotonin reuptake inhibitors, and proton-pump inhibitors, and the likelihood of disease progression from mild cognitive impairment (MCI) to dementia using electronic health records (EHRs). We conducted a retrospective cohort study using observational EHRs from a cohort of approximately 2 million patients seen at a large, multi-specialty urban academic medical center in New York City, USA between 2008 and 2020 to automatically emulate the randomized controlled trials. For each drug class, two exposure groups were identified based on the prescription orders documented in the EHRs following their MCI diagnosis. During follow-up, we measured drug efficacy based on the incidence of dementia and estimated the average treatment effect (ATE) of various drugs. To ensure the robustness of our findings, we confirmed the ATE estimates via bootstrapping and presented associated 95% confidence intervals (CIs). Our analysis identified 14,269 MCI patients, among whom 2501 (17.5%) progressed to dementia. Using average treatment estimation and bootstrapping confirmation, we observed that drugs including rosuvastatin (ATE = − 0.0140 [− 0.0191, − 0.0088], p value

Abstract This research proposes a deep-learning paradigm, termed functional learning (FL), to physically train a loose neuron array, a group of non-handcrafted, non-differentiable, and loosely connected physical neurons whose connections and gradients are beyond explicit expression. The paradigm targets training non-differentiable hardware, and therefore solves many interdisciplinary challenges at once: the precise modeling and control of high-dimensional systems, the on-site calibration of multimodal hardware imperfectness, and the end-to-end training of non-differentiable and modeless physical neurons through implicit gradient propagation. It offers a methodology to build hardware without handcrafted design, strict fabrication, and precise assembling, thus forging paths for hardware design, chip manufacturing, physical neuron training, and system control. In addition, the functional learning paradigm is numerically and physically verified with an original light field neural network (LFNN). It realizes a programmable incoherent optical neural network, a well-known challenge that delivers light-speed, high-bandwidth, and power-efficient neural network inference via processing parallel visible light signals in the free space. As a promising supplement to existing power- and bandwidth-constrained digital neural networks, light field neural network has various potential applications: brain-inspired optical computation, high-bandwidth power-efficient neural network inference, and light-speed programmable lens/displays/detectors that operate in visible light.

ObjectiveEvidence suggests that high-quality health education and effective communication within the framework of social support hold significant potential in preventing postpartum depression. Yet, developing trustworthy and engaging health education and communication materials requires extensive expertise and substantial resources. In light of this, we propose an innovative approach that involves leveraging natural language processing (NLP) to classify publicly accessible lay articles based on their relevance and subject matter to pregnancy and mental health.Materials and methodsWe manually reviewed online lay articles from credible and medically validated sources to create a gold standard corpus. This manual review process categorized the articles based on their pertinence to pregnancy and related subtopics. To streamline and expand the classification procedure for relevance and topics, we employed advanced NLP models such as Random Forest, Bidirectional Encoder Representations from Transformers (BERT), and Generative Pre-trained Transformer model (gpt-3.5-turbo).ResultsThe gold standard corpus included 392 pregnancy-related articles. Our manual review process categorized the reading materials according to lifestyle factors associated with postpartum depression: diet, exercise, mental health, and health literacy. A BERT-based model performed best (F1 = 0.974) in an end-to-end classification of relevance and topics. In a two-step approach, given articles already classified as pregnancy-related, gpt-3.5-turbo performed best (F1 = 0.972) in classifying the above topics.DiscussionUtilizing NLP, we can guide patients to high-quality lay reading materials as cost-effective, readily available health education and communication sources. This approach allows us to scale the information delivery specifically to individuals, enhancing the relevance and impact of the materials provided.

Abstract Background The Pooled Cohort Equations (PCEs) are race- and sex-specific Cox proportional hazards (PH)-based models used for 10-year atherosclerotic cardiovascular disease (ASCVD) risk prediction with acceptable discrimination. In recent years, neural network models have gained increasing popularity with their success in image recognition and text classification. Various survival neural network models have been proposed by combining survival analysis and neural network architecture to take advantage of the strengths from both. However, the performance of these survival neural network models compared to each other and to PCEs in ASCVD prediction is unknown. Methods In this study, we used 6 cohorts from the Lifetime Risk Pooling Project (with 5 cohorts as training/internal validation and one cohort as external validation) and compared the performance of the PCEs in 10-year ASCVD risk prediction with an all two-way interactions Cox PH model (Cox PH-TWI) and three state-of-the-art neural network survival models including Nnet-survival, Deepsurv, and Cox-nnet. For all the models, we used the same 7 covariates as used in the PCEs. We fitted each of the aforementioned models in white females, white males, black females, and black males, respectively. We evaluated models’ internal and external discrimination power and calibration. Results The training/internal validation sample comprised 23216 individuals. The average age at baseline was 57.8 years old (SD = 9.6); 16% developed ASCVD during average follow-up of 10.50 (SD = 3.02) years. Based on 10 × 10 cross-validation, the method that had the highest C-statistics was Deepsurv (0.7371) for white males, Deepsurv and Cox PH-TWI (0.7972) for white females, PCE (0.6981) for black males, and Deepsurv (0.7886) for black females. In the external validation dataset, Deepsurv (0.7032), Cox-nnet (0.7282), PCE (0.6811), and Deepsurv (0.7316) had the highest C-statistics for white male, white female, black male, and black female population, respectively. Calibration plots showed that in 10 × 10 validation, all models had good calibration in all race and sex groups. In external validation, all models overestimated the risk for 10-year ASCVD. Conclusions We demonstrated the use of the state-of-the-art neural network survival models in ASCVD risk prediction. Neural network survival models had similar if not superior discrimination and calibration compared to PCEs.

Abstract Left ventricular ejection fraction (EF) is a key measure in the diagnosis and treatment of heart failure (HF) and many patients experience changes in EF overtime. Large-scale analysis of longitudinal changes in EF using electronic health records (EHRs) is limited. In a multi-site retrospective study using EHR data from three academic medical centers, we investigated longitudinal changes in EF measurements in patients diagnosed with HF. We observed significant variations in baseline characteristics and longitudinal EF change behavior of the HF cohorts from a previous study that is based on HF registry data. Data gathered from this longitudinal study were used to develop multiple machine learning models to predict changes in ejection fraction measurements in HF patients. Across all three sites, we observed higher performance in predicting EF increase over a 1-year duration, with similarly higher performance predicting an EF increase of 30% from baseline compared to lower percentage increases. In predicting EF decrease we found moderate to high performance with low confidence for various models. Among various machine learning models, XGBoost was the best performing model for predicting EF changes. Across the three sites, the XGBoost model had an F1-score of 87.2, 89.9, and 88.6 and AUC of 0.83, 0.87, and 0.90 in predicting a 30% increase in EF, and had an F1-score of 95.0, 90.6, 90.1 and AUC of 0.54, 0.56, 0.68 in predicting a 30% decrease in EF. Among features that contribute to predicting EF changes, baseline ejection fraction measurement, age, gender, and heart diseases were found to be statistically significant.

Abstract Understanding media frames and the public resonance during disasters is essential for making inclusive climate change and adaptation policies in the context of increasingly extreme weather events. In this study, we use the extreme weather and flood event that occurred in July 2021 in Zhengzhou, China, as a case study to investigate how official media in China reported this event and how the public responded. Moreover, since one accountability investigation report regarding this disaster was released in January 2022, we also compared these posts between the emergency response period and the post-crisis learning period after the report’s release. Topic modeling using the LDA (Latent Dirichlet Allocation) method and emotion analysis were conducted to analyze the posts from Weibo, China’s primary social media platform. The results demonstrated that the posts from official media and the public comments differed in both topics and emotions, with relatively little coherence. During the emergency response period, the media’s posts focused more on the facts, such as the extreme weather event, the places where it occurred, the impacts, and the search and rescue efforts, while the public comments were more about help appeals from the neglected ones in the rural areas, and emotional expressions such as moral support, condolence or encouragement to the victims and their families. After the accountability investigation in January, the media’s posts primarily covered the investigation process, the punishment, the attribution of disaster consequences, and the lessons learned, while the public’s comments were relatively emotional, praised the good, condoled the victims, and condemned the villains. The dominant emotion from the media’s posts was “like” in July 2021, but it became depression in January 2022. Anger was the prevalent emotion from the public during all the stages. This study provided valuable knowledge to the current understanding of the different patterns and dynamics of official media reports and the public’s resonance in disaster management.

Broken waste rock as backfill materials is filled into the underground goaf and will encounter particle breakage under the overburden pressure. The damage degree of the crushed waste rock materials directly affects the compression properties of the backfill body and the control effect of strata movement and surface subsidence. Fractal theory can be used to quantify the breakage degree of waste rock particles. In this study, compression tests are carried out on the broken waste rock materials with different particle size distributions to investigate the influence of the particle size distribution on the fractal characteristics of waste rock backfill materials under compression. The results show that the waste rock specimens with discontinuous size grading and containing more coarse particles have higher stress when entering the stable deformation stage, larger strain if at the same stress level and larger breakage ratio after compression. The particle size distribution of the crushed waste rock specimens after compressive loading has superior fractal characteristics and the fractal dimension increases with the content of fine particles and decreases significantly under discontinuous size grading. The fractal dimension has significant negative exponential function regression relationship with the strain and the breakage ratio. The strain and the breakage ratio decline with the growing fractal dimension. If the waste rock specimens have continuous particle size gradation and higher content of fine particles, the fractal dimension, the strain and the breakage ratio of the waste rock backfill materials tend to be optimal and stable, proving a desirable initial particle size distribution of the waste rock backfill materials.

To examine the complex relationship between Internet use experience and character spelling performance among Chinese junior high school students, the study explored the multiple mediating roles of Pinyin input proficiency and net-speak experience. A total of 447 Chinese junior high school students aged 12–15 years old completed the Internet Use Experience and Pinyin Input Proficiency Assessment, the Net-speak Experience Questionnaire and the Chinese Spelling Test. The results showed that: (1) All investigated variables were significantly correlated with each other, but there was no direct relationship between Internet use and Chinese spelling performance. (2) Pinyin input proficiency and net-speak experience play a chain mediating role in the relationship between Internet use and Chinese character spelling performance. Teens’ Internet use experience indirectly and positively predicted Chinese character spelling performance through the mediation of Pinyin input method use and net-speak experience. The implication of this study is that Chinese children should be guided to engage in Internet activities that require Pinyin typing and use net-speak creatively in order to promote the traditional Chinese character spelling skills when instructing teenagers to engage in Internet activities.

Compared to longitudinal ventilation, there are few studies on fire source development under semi-transverse ventilation. This work studied the influence of semi-transverse ventilation on the combustion characteristics of fire sources in a scaled tunnel. The burning rate and heat transfer feedback during pool fire combustion were revealed under different longitudinal and transverse ventilation velocities. The results showed that transverse ventilation had little influence on combustion characteristics, and the burning rate was more obviously affected by longitudinal ventilation. The heat convection feedback increased monotonically with the increase of the longitudinal ventilation, which led to the increase of the total heat feedback on the fuel. The heat radiation feedback changed little, and the heat conduction feedback decreased monotonically with the increase of the longitudinal ventilation velocity. By aid of a Fire Dynamics Simulator, it was found that the flame tilted downstream and was in the flow line of the lower cold air flow coming from upstream and the upper hot smoke flow outgoing in the downstream direction. The transverse ventilation of 2 m/s or lower hardly affected the combustion field of the fire source. Therefore, semi-transverse ventilation is preferable to longitudinal ventilation from the point of view of limiting fire expansion.

Abstract Primary open-angle glaucoma (POAG) is a leading cause of irreversible blindness worldwide. Although deep learning methods have been proposed to diagnose POAG, it remains challenging to develop a robust and explainable algorithm to automatically facilitate the downstream diagnostic tasks. In this study, we present an automated classification algorithm, GlaucomaNet, to identify POAG using variable fundus photographs from different populations and settings. GlaucomaNet consists of two convolutional neural networks to simulate the human grading process: learning the discriminative features and fusing the features for grading. We evaluated GlaucomaNet on two datasets: Ocular Hypertension Treatment Study (OHTS) participants and the Large-scale Attention-based Glaucoma (LAG) dataset. GlaucomaNet achieved the highest AUC of 0.904 and 0.997 for POAG diagnosis on OHTS and LAG datasets. An ensemble of network architectures further improved diagnostic accuracy. By simulating the human grading process, GlaucomaNet demonstrated high accuracy with increased transparency in POAG diagnosis (comprehensiveness scores of 97% and 36%). These methods also address two well-known challenges in the field: the need for increased image data diversity and relying heavily on perimetry for POAG diagnosis. These results highlight the potential of deep learning to assist and enhance clinical POAG diagnosis. GlaucomaNet is publicly available on https://github.com/bionlplab/GlaucomaNet .

Purpose: We compared the long-term outcome of the watch and wait (WW) strategy and surgery in patients with locally advanced rectal cancer. Patients and methods: This prospective cohort study included 84 patients who achieved clinical complete response (cCR) after neoadjuvant chemoradiotherapy (NCRT). They were divided into the WW group (n = 58) and surgery group (SG, n = 26). Patients in the SG underwent total mesorectal excision. The study site was the Peking University Cancer Hospital. Results: Eighty-four patients were included (58 and 26 in the WW group and SG, respectively). A total of 76·9% of the patients in the SG achieved pathological complete response (pCR) and 23·1% of the patients had a residual tumor. The total recurrence and metastasis rate was 15·4% (4/26) in the SG and 18·9% (11/58) in the WW group. There was no significant difference in the recurrence and metastasis rate between the two groups. In the WW group, 9 cases developed tumor regrowth during follow-up and underwent salvage surgery. The overall survival rate of the WW group (96·6% vs 92·3%) was not significantly different from that of the SG (P > 0·05). The WW patients also retained their anal sphincter function and avoided surgery-related complications. Conclusion: The WW strategy is a feasible treatment option in patients with cCR after NCRT. Surgery may not bring benefits to these cCR patients.

BackgroundFew studies have evaluated the significance of sarcopenia in predicting the outcomes of patients with adenocarcinoma of the esophagogastric junction (AEG), especially those who received neoadjuvant chemoradiotherapy (NCRT). We aimed to identify the sarcopenic status and its impact on the outcomes of patients with locally advanced AEG who received NCRT followed by radical surgery or systemic therapy.Materials and methodsPatients with T3-4N+M0 AEG with accessible abdominal computed tomography (CT) before and after NCRT were retrospectively analyzed. Body composition parameters, particularly the skeletal muscle index (SMI), were assessed using a CT-based method, and sarcopenia was defined using a predetermined SMI cutoff value. Survival analysis was conducted using the Kaplan–Meier method. A Cox proportional hazards regression model was used to identify independent prognostic factors. Receiver operating characteristic curve analysis was carried out, and the area under the curve (AUC) was calculated to test the prognostic accuracy of different factors.ResultsA total of 63 patients were enrolled, 65.1 and 79.4% of whom developed pre- and post-NCRT sarcopenia, respectively. Patients with pre-NCRT sarcopenia had lower radical surgery rates (70.7 vs. 95.5%, p = 0.047) than those without sarcopenia; however, sarcopenic status did not affect other short-term outcomes, including treatment-related toxicity and efficacy. Pre-NCRT sarcopenia was identified as an independent predictive factor for poor overall survival (OS) [adjusted hazard ratio (HR), 6.053; p = 0.002] and progression-free survival (PFS) (adjusted HR, 2.873; p = 0.031). Compared with nutritional indices such as the Nutritional Risk Screening 2002, weight loss during NCRT, and post-NCRT sarcopenia, pre-NCRT sarcopenia was regarded as the best predictive index for the 5-year OS (AUC = 0.735) and PFS rates (AUC = 0.770).ConclusionPre-NCRT sarcopenia may be an independent predictive factor for OS and PFS rates in patients with locally advanced AEG receiving multimodal treatment.

Purpose: Primary open-angle glaucoma (POAG) is one of the leading causes of irreversible blindness in the United States and worldwide. Although deep learning methods have been proposed to diagnose POAG, these methods all used a single image as input. Contrastingly, glaucoma specialists typically compare the follow-up image with the baseline image to diagnose incident glaucoma. To simulate this process, we proposed a Siamese neural network, POAGNet, to detect POAG from optic disc photographs. Design: The POAGNet, an algorithm for glaucoma diagnosis, is developed using optic disc photographs. Participants: The POAGNet was trained and evaluated on 2 data sets: (1) 37 339 optic disc photographs from 1636 Ocular Hypertension Treatment Study (OHTS) participants and (2) 3684 optic disc photographs from the Sequential fundus Images for Glaucoma (SIG) data set. Gold standard labels were obtained using reading center grades. Methods: We proposed a Siamese network model, POAGNet, to simulate the clinical process of identifying POAG from optic disc photographs. The POAGNet consists of 2 side outputs for deep supervision and uses convolution to measure the similarity between 2 networks. Main Outcome Measures: The main outcome measures are the area under the receiver operating characteristic curve, accuracy, sensitivity, and specificity. Results: In POAG diagnosis, extensive experiments show that POAGNet performed better than the best state-of-the-art model on the OHTS test set (area under the curve [AUC] 0.9587 versus 0.8750). It also outperformed the baseline models on the SIG test set (AUC 0.7518 versus 0.6434). To assess the transferability of POAGNet, we also validated the impact of cross-data set variability on our model. The model trained on OHTS achieved an AUC of 0.7490 on SIG, comparable to the previous model trained on the same data set. When using the combination of SIG and OHTS for training, our model achieved superior AUC to the single-data model (AUC 0.8165 versus 0.7518). These demonstrate the relative generalizability of POAGNet. Conclusions: By simulating the clinical grading process, POAGNet demonstrated high accuracy in POAG diagnosis. These results highlight the potential of deep learning to assist and enhance clinical POAG diagnosis. The POAGNet is publicly available on https://github.com/bionlplab/poagnet.

Imaging systems are widely applied in harsh environments where the performance of shallow‐designed systems may deviate from expectation. As a representative scenario, environmental temperature variation may degrade image quality due to thermal defocus and sensor response, resulting in blur and noise. However, extensive athermalization in optics usually requires a complex design process and is limited by materials. Herein, a multibranch computational imaging scheme is developed, using emerging generative adversarial networks as the postprocessing to compensate for degradation of all kinds caused by thermal defocus and noise. In addition, a temperature controllable data acquisition, division, and mixture scheme is described to facilitate effective datasets for model robustness. Experiments on a vehicle lens and a mobile phone lens reveal that the proposed multibranch learned strategy notably increases image quality in the temperature range of 0–80 °C, and outperforms conventional athermalization in most instances, which is beneficial to lowering the design and manufacturing costs of imaging systems.

Measurement(s) chemical entity • textual entity • Annotation Technology Type(s) expert manual annotation • digital curation • machine learning Machine-accessible metadata file describing the reported data: https://doi.org/10.6084/m9.figshare.13486839

Abstract By 2040, age-related macular degeneration (AMD) will affect ~288 million people worldwide. Identifying individuals at high risk of progression to late AMD, the sight-threatening stage, is critical for clinical actions, including medical interventions and timely monitoring. Although deep learning has shown promise in diagnosing/screening AMD using color fundus photographs, it remains difficult to predict individuals’ risks of late AMD accurately. For both tasks, these initial deep learning attempts have remained largely unvalidated in independent cohorts. Here, we demonstrate how deep learning and survival analysis can predict the probability of progression to late AMD using 3298 participants (over 80,000 images) from the Age-Related Eye Disease Studies AREDS and AREDS2, the largest longitudinal clinical trials in AMD. When validated against an independent test data set of 601 participants, our model achieved high prognostic accuracy (5-year C-statistic 86.4 (95% confidence interval 86.2–86.6)) that substantially exceeded that of retinal specialists using two existing clinical standards (81.3 (81.1–81.5) and 82.0 (81.8–82.3), respectively). Interestingly, our approach offers additional strengths over the existing clinical standards in AMD prognosis (e.g., risk ascertainment above 50%) and is likely to be highly generalizable, given the breadth of training data from 82 US retinal specialty clinics. Indeed, during external validation through training on AREDS and testing on AREDS2 as an independent cohort, our model retained substantially higher prognostic accuracy than existing clinical standards. These results highlight the potential of deep learning systems to enhance clinical decision-making in AMD patients.

Abstract As one of the most ubiquitous diagnostic imaging tests in medical practice, chest radiography requires timely reporting of potential findings and diagnosis of diseases in the images. Automated, fast, and reliable detection of diseases based on chest radiography is a critical step in radiology workflow. In this work, we developed and evaluated various deep convolutional neural networks (CNN) for differentiating between normal and abnormal frontal chest radiographs, in order to help alert radiologists and clinicians of potential abnormal findings as a means of work list triaging and reporting prioritization. A CNN-based model achieved an AUC of 0.9824 ± 0.0043 (with an accuracy of 94.64 ± 0.45%, a sensitivity of 96.50 ± 0.36% and a specificity of 92.86 ± 0.48%) for normal versus abnormal chest radiograph classification. The CNN model obtained an AUC of 0.9804 ± 0.0032 (with an accuracy of 94.71 ± 0.32%, a sensitivity of 92.20 ± 0.34% and a specificity of 96.34 ± 0.31%) for normal versus lung opacity classification. Classification performance on the external dataset showed that the CNN model is likely to be highly generalizable, with an AUC of 0.9444 ± 0.0029. The CNN model pre-trained on cohorts of adult patients and fine-tuned on pediatric patients achieved an AUC of 0.9851 ± 0.0046 for normal versus pneumonia classification. Pretraining with natural images demonstrates benefit for a moderate-sized training image set of about 8500 images. The remarkable performance in diagnostic accuracy observed in this study shows that deep CNNs can accurately and effectively differentiate normal and abnormal chest radiographs, thereby providing potential benefits to radiology workflow and patient care.

Background. There is an increasing trend to represent domain knowledge in structured graphs, which provide efficient knowledge representations for many downstream tasks. Knowledge graphs are widely used to model prior knowledge in the form of nodes and edges to represent semantically connected knowledge entities, which several works have adopted into different medical imaging applications. Methods. We systematically searched over five databases to find relevant articles that applied knowledge graphs to medical imaging analysis. After screening, evaluating, and reviewing the selected articles, we performed a systematic analysis. Results. We looked at four applications in medical imaging analysis, including disease classification, disease localization and segmentation, report generation, and image retrieval. We also identified limitations of current work, such as the limited amount of available annotated data and weak generalizability to other tasks. We further identified the potential future directions according to the identified limitations, including employing semisupervised frameworks to alleviate the need for annotated data and exploring task-agnostic models to provide better generalizability. Conclusions. We hope that our article will provide the readers with aggregated documentation of the state-of-the-art knowledge graph applications for medical imaging to encourage future research.

BackgroundSemantic textual similarity (STS) measures the degree of relatedness between sentence pairs. The Open Health Natural Language Processing (OHNLP) Consortium released an expertly annotated STS data set and called for the National Natural Language Processing Clinical Challenges. This work describes our entry, an ensemble model that leverages a range of deep learning (DL) models. Our team from the National Library of Medicine obtained a Pearson correlation of 0.8967 in an official test set during 2019 National Natural Language Processing Clinical Challenges/Open Health Natural Language Processing shared task and achieved a second rank. ObjectiveAlthough our models strongly correlate with manual annotations, annotator-level correlation was only moderate (weighted Cohen κ=0.60). We are cautious of the potential use of DL models in production systems and argue that it is more critical to evaluate the models in-depth, especially those with extremely high correlations. In this study, we benchmark the effectiveness and efficiency of top-ranked DL models. We quantify their robustness and inference times to validate their usefulness in real-time applications. MethodsWe benchmarked five DL models, which are the top-ranked systems for STS tasks: Convolutional Neural Network, BioSentVec, BioBERT, BlueBERT, and ClinicalBERT. We evaluated a random forest model as an additional baseline. For each model, we repeated the experiment 10 times, using the official training and testing sets. We reported 95% CI of the Wilcoxon rank-sum test on the average Pearson correlation (official evaluation metric) and running time. We further evaluated Spearman correlation, R², and mean squared error as additional measures. ResultsUsing only the official training set, all models obtained highly effective results. BioSentVec and BioBERT achieved the highest average Pearson correlations (0.8497 and 0.8481, respectively). BioSentVec also had the highest results in 3 of 4 effectiveness measures, followed by BioBERT. However, their robustness to sentence pairs of different similarity levels varies significantly. A particular observation is that BERT models made the most errors (a mean squared error of over 2.5) on highly similar sentence pairs. They cannot capture highly similar sentence pairs effectively when they have different negation terms or word orders. In addition, time efficiency is dramatically different from the effectiveness results. On average, the BERT models were approximately 20 times and 50 times slower than the Convolutional Neural Network and BioSentVec models, respectively. This results in challenges for real-time applications. ConclusionsDespite the excitement of further improving Pearson correlations in this data set, our results highlight that evaluations of the effectiveness and efficiency of STS models are critical. In future, we suggest more evaluations on the generalization capability and user-level testing of the models. We call for community efforts to create more biomedical and clinical STS data sets from different perspectives to reflect the multifaceted notion of sentence-relatedness.

Coal gangue, as a solid waste produced in the coal mining process, can be disposed by being prepared into backfill materials and then filled in underground goafs, thus controlling strata movement and surface subsidence. However, gangue backfill materials are non-continuous; therefore, research into the surface deformation induced by backfill mining should consider the creep compression behavior of gangue backfill materials. The research took a backfill panel in Tangshan Coal Mine (Tangshan City, Hebei Province, China) as the background. In addition, broken coal gangue was collected in the field to prepare specimens of gangue backfill materials, and their creep compression properties were measured. The corresponding constitutive equation of creep compression was then established and embedded in the numerical software, FLAC3D. By building the numerical model for surface deformation induced by backfill mining, the surface deformation above the backfill panel under conditions of different creep durations of backfill materials was simulated and evaluated. In addition, two measuring lines were arranged on the surface to monitor changes in surface subsidence. After surface subsidence stabilized, the maximum surface subsidence was 163.4 mm, which satisfied the fortification criterion of surface buildings. This means the backfill mining did not affect nearby buildings. The results provide a theoretical basis for predicting surface deformation induced by backfill mining and its effective control.

Summary: Single-cell RNA sequencing (scRNA-seq) has become a revolutionary technology to characterize cells under different biological conditions. Unlike bulk RNA-seq, gene expression from scRNA-seq is highly sparse due to limited sequencing depth per cell. This is worsened by tossing away a significant portion of reads that attribute to gene quantification. To overcome data sparsity and fully utilize original reads, we propose scSimClassify, a reference-free and alignment-free approach to classify cell types with k-mer level features. The compressed k-mer groups (CKGs), identified by the simhash method, contain k-mers with similar abundance profiles and serve as the cells’ features. Our experiments demonstrate that CKG features lend themselves to better performance than gene expression features in scRNA-seq classification accuracy in the majority of experimental cases. Because CKGs are derived from raw reads without alignment to reference genome, scSimClassify offers an effective alternative to existing methods especially when reference genome is incomplete or insufficient to represent subject genomes.

Abstract Background The aim of this study was to explore the impact of including or excluding the ischiorectal fossa (IRF) within the clinical target volume during neoadjuvant chemoradiotherapy (NCRT) using intensity modulated radiotherapy, in locally advanced lower rectal cancer (LALRC). Methods We retrospectively analysed the data of 220 LALRC patients who received NCRT followed by abdominoperineal resection between January 2009 and January 2015. Six patients were excluded because of loss to follow-up, 90 patients received IRF irradiation (IRF group) while 124 patients did not (NIRF group). Survival, patterns of recurrence, and treatment toxicities were compared between the two groups. Results Overall, patient/treatment variables were well balanced except for surgical technique. Perineal wound complications in the IRF and NIRF groups, were 40.0 and 24.2%, respectively (p = 0.010); corresponding 3-year perineal recurrence rates, local recurrence free survival, overall survival, and distant relapse free survival were 4.4% vs. 2.4% (p = 0.670), 88.1% vs. 95.0% (p = 0.079), 82.6% vs. 88.4% (p = 0.087), and 61.9% vs. 81.0% (p = 0.026), respectively. Multivariate analyses demonstrated the following factors to be significantly related to perineal wound complications: irradiation of the IRF (odds ratio [OR] 2.892, p = 0.002), anaemia (OR 3.776, p = 0.010), operation duration > 180 min (OR 2.486, p = 0.007), and interval between radiotherapy and surgery > 8 weeks (OR 2.400, p = 0.010). Conclusions Exclusion of the IRF from the clinical target volume during NCRT using intensity-modulated radiotherapy in LALRC could lower the incidence of perineal wound complications after abdominoperineal resection, without compromising oncological outcomes.

The properties of optical vortex provide new insights into a wide range of optical and physical phenomena, giving rise to numerous applications in multiple disciplines. Notably, as much as the associated technologies have become increasingly flexible, it is still challenging to simultaneously generate multiple longitudinal polarization optical vortices with random topological charges (l) and tunable intensities at arbitrary positions. Herein, a novel approach that, by manipulating phase and amplitude at the back pupil of a high numerical‐aperture (NA) objective, realizes arbitrary longitudinal polarization vortices within the tightly focused light field is developed. Building upon the same model, the impacts of polarization on the optical vortex field and its potential of tuning the ratio of the longitudinal polarization component are further investigated.

Metastasis, the leading cause of death in cancer patients, requires the invasion of tumor cells through the stroma in response to migratory cues, in part provided by the extracellular matrix (ECM). Recent advances in proteomics have led to the identification of hundreds of ECM proteins, which are more abundant in tumors relative to healthy tissue. Our goal was to develop a pipeline to easily predict which ECM proteins are more likely to have an effect on cancer invasion and metastasis. We evaluated the effect of four ECM proteins upregulated in breast tumor tissue in multiple human breast cancer cell lines in three assays. There was no linear relationship between cell adhesion to ECM proteins and ECM-driven 2D cell migration speed, persistence, or 3D invasion. We then used classifiers and partial-least squares regression analysis to identify which metrics best predicted ECM-driven 2D migration and 3D invasion responses. We find that ECM-driven 2D cell migration speed or persistence did not predict 3D invasion in response to the same cue. However, cell adhesion, and in particular cell elongation and shape irregularity, accurately predicted the magnitude of ECM-driven 2D migration and 3D invasion. Our models successfully predicted the effect of novel ECM proteins in a cell-line specific manner. Overall, our studies identify the cell morphological features that determine 3D invasion responses to individual ECM proteins. This platform will help provide insight into the functional role of ECM proteins abundant in tumor tissue and help prioritize strategies for targeting tumor-ECM interactions to treat metastasis.

Summary: Obesity commonly co-exists with fatty liver disease with increasing health burden worldwide. Family with Sequence Similarity 13, Member A (FAM13A) has been associated with lipid levels and fat mass by genome-wide association studies (GWAS). However, the function of FAM13A in maintaining metabolic homeostasis in vivo remains unclear. Here, we demonstrated that rs2276936 in this locus has allelic-enhancer activity in massively parallel reporter assays (MPRA) and reporter assay. The DNA region containing rs2276936 regulates expression of endogenous FAM13A in HepG2 cells. In vivo, Fam13a−/− mice are protected from high-fat diet (HFD)-induced fatty liver accompanied by increased insulin sensitivity and reduced glucose production in liver. Mechanistically, loss of Fam13a led to the activation of AMP-activated protein kinase (AMPK) and increased mitochondrial respiration in primary hepatocytes. These findings demonstrate that FAM13A mediates obesity-related dysregulation of lipid and glucose homeostasis. Targeting FAM13A might be a promising treatment of obesity and fatty liver disease. : Biological Sciences; Cell Biology; Functional Aspects of Cell Biology Subject Areas: Biological Sciences, Cell Biology, Functional Aspects of Cell Biology

The resistance of the notorious rice pest Nilaparvata lugens to many insecticides has caused significant concerns. Our previous study demonstrated that the fungus Metarhizium anisopliae CQMa421 shows great potential for the control of this pest, but the interactions between them are still unclear. Thus, we further investigated fungal infection-related microRNAs (miRNAs) in N. lugens during M. anisopliae CQMa421 challenge using Illumina sequencing. In this study, we constructed twenty-four small RNA libraries over different time courses (i.e., 4 h, 8 h, 16 h, and 24 h). A total of 478.62 M clean reads were collected, with each sample producing more than 13.37 M reads, after the removal of low-quality reads. We identified 2324 miRNAs and their 11,076 target genes within the twenty-four libraries by bioinformatics analysis. Differentially expressed miRNAs (DEmiRNAs), including 58 (32 upregulated vs. 26 downregulated), 62 (30 upregulated vs. 32 downregulated), 126 (71 upregulated vs. 55 downregulated), and 109 (40 upregulated vs. 69 downregulated) DEmiRNAs were identified at 4 h, 8 h, 16 h, and 24 h post-infection, respectively. We further conducted Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis to predict the functions of all target genes of DEmiRNAs. These DEmiRNAs targets identified during 24 h of infection were primarily involved in energy metabolism, lysine degradation, the FoxO signaling pathway, ubiquitin-mediated proteolysis, the mRNA surveillance pathway, and the MAPK signaling pathway. Taken together, our results provide essential information for further study of the interactions between the entomopathogenic fungus M. anisopliae and N. lugens at the posttranscriptional level.

Nilaparvata lugens is one of the major pests of rice and results in substantial yield loss every year. Our previous study found that the entomopathogenic fungus Metarhizium anisopliae showed effective potential for controlling this pest. However, the mechanisms underlying M. anisopliae infection of N. lugens are not well known. In the present study, we further examined the transcriptome of N. lugens at 4 h, 8 h, 16 h, and 24 h after M. anisopliae infection by Illumina deep sequencing. In total, 174.17 Gb of data was collected after sequencing, from which 23,398 unigenes were annotated by various databases, including 3694 newly annotated genes. The results showed that there were 246 vs 75, 275 vs 586, 378 vs 1055, and 638 vs 182 up- and downregulated differentially expressed genes (DEGs) at 4 h, 8 h, 16 h, and 24 h after M. anisopliae infection, respectively. The biological functions and associated metabolic processes of these genes were determined with the Clusters of Orthologous Groups (COG), Gene Ontology (GO), and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases. The DEGs data were verified using RT-qPCR. These results indicated that the DEGs during the initial fungal infection appropriately reflected the time course of the response to the fungal infection. Taken together, the results of this study provide new insights into the molecular mechanisms underlying the insect host response to fungal infection, especially during the initial stage of infection, and may improve the potential control strategies for N. lugens.

Entomopathogenic fungi are the key regulators of insect populations and some of them are important biological agents used in integrated pest management strategies. Compared with their ability to become resistant to insecticides, insect pests do not easily become resistant to the infection by entomopathogenic fungi. In this study, we evaluated the mortality and immune response of the serious crop pest Locusta migratoria manilensis after exposure to a new entomopathogenic fungus strain, Metarhizium anisopliae CQMa421. M. anisopliae CQMa421 could effectively infect and kill the L. migratoria adults and nymphs. The locust LT50 under 1 × 108 conidia/mL concentration of M. anisopliae was much lower than that under conidial concentration 1 × 105 conidia/mL (i.e., 6.0 vs. 11.2 and 5.0 vs. 13.8 for adults and nymphs, respectively). The LC50 (log10) of M. anisopliae against locust adults and nymphs after 10 days was 5.2 and 5.6, respectively. Although the number of hemocytes in L. migratoria after exposure to M. anisopliae did not differ with that in the controls, the enzymatic activity of superoxide dismutase (SOD) and prophenoloxidase (ProPO) did differ between the two treatments. The activities of both SOD and ProPO under the M. anisopliae treatment were lower than that in the controls, except for the ProPO activity at 72 h and the SOD activity at 96 h. Further, the expression of the L. migratoria immune-related genes defensin, spaetzle, and attacin differed after exposure to M. anisopliae for 24 h to 96 h. Taken together, this study indicated that infection with M. anisopliae CQMa421 could cause the death of L. migratoria by interacting with the immune responses of the host, demonstrating that this fungal strain of M. anisopliae can be an efficient biocontrol agent against L. migratoria.