Promotional language and the adoption of innovative ideas in science.

How are the merits of innovative ideas communicated in science? Here, we conduct semantic analyses of grant application success with a focus on scientific promotional language, which may help to convey an innovative idea's originality and significance. Our analysis attempts to surmount the limitations of prior grant studies by examining the full text of tens of thousands of both funded and unfunded grants from three leading public and private funding agencies: the NIH, the NSF, and the Novo Nordisk Foundation, one of the world's largest private science funding foundations. We find a robust association between promotional language and the support and adoption of innovative ideas by funders and other scientists. First, a grant proposal's percentage of promotional language is associated with up to a doubling of the grant's probability of being funded. Second, a grant's promotional language reflects its intrinsic innovativeness. Third, the percentage of promotional language is predictive of the expected citation and productivity impact of publications that are supported by funded grants. Finally, a computer-assisted experiment that manipulates the promotional language in our data demonstrates how promotional language can communicate the merit of ideas through cognitive activation. With the incidence of promotional language in science steeply rising, and the pivotal role of grants in converting promising and aspirational ideas into solutions, our analysis provides empirical evidence that promotional language is associated with effectively communicating the merits of innovative scientific ideas.

People who share encounters with racism are silenced online by humans and machines, but a guideline-reframing intervention holds promise.

Are members of marginalized communities silenced on social media when they share personal experiences of racism? Here, we investigate the role of algorithms, humans, and platform guidelines in suppressing disclosures of racial discrimination. In a field study of actual posts from a neighborhood-based social media platform, we find that when users talk about their experiences as targets of racism, their posts are disproportionately flagged for removal as toxic by five widely used moderation algorithms from major online platforms, including the most recent large language models. We show that human users disproportionately flag these disclosures for removal as well. Next, in a follow-up experiment, we demonstrate that merely witnessing such suppression negatively influences how Black Americans view the community and their place in it. Finally, to address these challenges to equity and inclusion in online spaces, we introduce a mitigation strategy: a guideline-reframing intervention that is effective at reducing silencing behavior across the political spectrum.

Literature-based predictions of Mendelian disease therapies.

In the effort to treat Mendelian disorders, correcting the underlying molecular imbalance may be more effective than symptomatic treatment. Identifying treatments that might accomplish this goal requires extensive and up-to-date knowledge of molecular pathways-including drug-gene and gene-gene relationships. To address this challenge, we present "parsing modifiers via article annotations" (PARMESAN), a computational tool that searches PubMed and PubMed Central for information to assemble these relationships into a central knowledge base. PARMESAN then predicts putatively novel drug-gene relationships, assigning an evidence-based score to each prediction. We compare PARMESAN's drug-gene predictions to all of the drug-gene relationships displayed by the Drug-Gene Interaction Database (DGIdb) and show that higher-scoring relationship predictions are more likely to match the directionality (up- versus down-regulation) indicated by this database. PARMESAN had more than 200,000 drug predictions scoring above 8 (as one example cutoff), for more than 3,700 genes. Among these predicted relationships, 210 were registered in DGIdb and 201 (96%) had matching directionality. This publicly available tool provides an automated way to prioritize drug screens to target the most-promising drugs to test, thereby saving time and resources in the development of therapeutics for genetic disorders.

PIF4 enhances the expression of SAUR genes to promote growth in response to nitrate.

Nitrate supply is fundamental to support shoot growth and crop performance, but the associated increase in stem height exacerbates the risks of lodging and yield losses. Despite their significance for agriculture, the mechanisms involved in the promotion of stem growth by nitrate remain poorly understood. Here, we show that the elongation of the hypocotyl of Arabidopsis thaliana, used as a model, responds rapidly and persistently to upshifts in nitrate concentration, rather than to the nitrate level itself. The response occurred even in shoots dissected from their roots and required NITRATE TRANSPORTER 1.1 (NRT1.1) in the phosphorylated state (but not NRT1.1 nitrate transport capacity) and NIN-LIKE PROTEIN 7 (NLP7). Nitrate increased PHYTOCHROME INTERACTING FACTOR 4 (PIF4) nuclear abundance by posttranscriptional mechanisms that depended on NRT1.1 and phytochrome B. In response to nitrate, PIF4 enhanced the expression of numerous SMALL AUXIN-UP RNA (SAUR) genes in the hypocotyl. The growth response to nitrate required PIF4, positive and negative regulators of its activity, including AUXIN RESPONSE FACTORs, and SAURs. PIF4 integrates cues from the soil (nitrate) and aerial (shade) environments adjusting plant stature to facilitate access to light.

Escalated police stops of Black men are linguistically and psychologically distinct in their earliest moments.

Across the United States, police chiefs, city officials, and community leaders alike have highlighted the need to de-escalate police encounters with the public. This concern about escalation extends from encounters involving use of force to routine car stops, where Black drivers are disproportionately pulled over. Yet, despite the calls for action, we know little about the trajectory of police stops or how escalation unfolds. In study 1, we use methods from computational linguistics to analyze police body-worn camera footage from 577 stops of Black drivers. We find that stops with escalated outcomes (those ending in arrest, handcuffing, or a search) diverge from stops without these outcomes in their earliest moments-even in the first 45 words spoken by the officer. In stops that result in escalation, officers are more likely to issue commands as their opening words to the driver and less likely to tell drivers the reason why they are being stopped. In study 2, we expose Black males to audio clips of the same stops and find differences in how escalated stops are perceived: Participants report more negative emotion, appraise officers more negatively, worry about force being used, and predict worse outcomes after hearing only the officer's initial words in escalated versus non-escalated stops. Our findings show that car stops that end in escalated outcomes sometimes begin in an escalated fashion, with adverse effects for Black male drivers and, in turn, police-community relations.

A complex receptor locus confers responsiveness to necrosis and ethylene-inducing like peptides in Brassica napus.

Brassica crops are susceptible to diseases which can be mitigated by breeding for resistance. MAMPs (microbe-associated molecular patterns) are conserved molecules of pathogens that elicit host defences known as pattern-triggered immunity (PTI). Necrosis and Ethylene-inducing peptide 1-like proteins (NLPs) are MAMPs found in a wide range of phytopathogens. We studied the response to BcNEP2, a representative NLP from Botrytis cinerea, and showed that it contributes to disease resistance in Brassica napus. To map regions conferring NLP response, we used the production of reactive oxygen species (ROS) induced during PTI across a population of diverse B. napus accessions for associative transcriptomics (AT), and bulk segregant analysis (BSA) on DNA pools created from a cross of NLP-responsive and non-responsive lines. In silico mapping with AT identified two peaks for NLP responsiveness on chromosomes A04 and C05 whereas the BSA identified one peak on A04. BSA delimited the region for NLP-responsiveness to 3 Mbp, containing ~245 genes on the Darmor-bzh reference genome and four co-segregating KASP markers were identified. The same pipeline with the ZS11 genome confirmed the highest-associated region on chromosome A04. Comparative BLAST analysis revealed unannotated clusters of receptor-like protein (RLP) homologues on ZS11 chromosome A04. However, no specific RLP homologue conferring NLP response could be identified. Our results also suggest that BR-SIGNALLING KINASE1 may be involved with modulating the NLP response. Overall, we demonstrate that responsiveness to NLP contributes to disease resistance in B. napus and define the associated genomic location. These results can have practical application in crop improvement.

Explainable AI for Bioinformatics: Methods, Tools and Applications.

Artificial intelligence (AI) systems utilizing deep neural networks and machine learning (ML) algorithms are widely used for solving critical problems in bioinformatics, biomedical informatics and precision medicine. However, complex ML models that are often perceived as opaque and black-box methods make it difficult to understand the reasoning behind their decisions. This lack of transparency can be a challenge for both end-users and decision-makers, as well as AI developers. In sensitive areas such as healthcare, explainability and accountability are not only desirable properties but also legally required for AI systems that can have a significant impact on human lives. Fairness is another growing concern, as algorithmic decisions should not show bias or discrimination towards certain groups or individuals based on sensitive attributes. Explainable AI (XAI) aims to overcome the opaqueness of black-box models and to provide transparency in how AI systems make decisions. Interpretable ML models can explain how they make predictions and identify factors that influence their outcomes. However, the majority of the state-of-the-art interpretable ML methods are domain-agnostic and have evolved from fields such as computer vision, automated reasoning or statistics, making direct application to bioinformatics problems challenging without customization and domain adaptation. In this paper, we discuss the importance of explainability and algorithmic transparency in the context of bioinformatics. We provide an overview of model-specific and model-agnostic interpretable ML methods and tools and outline their potential limitations. We discuss how existing interpretable ML methods can be customized and fit to bioinformatics research problems. Further, through case studies in bioimaging, cancer genomics and text mining, we demonstrate how XAI methods can improve transparency and decision fairness. Our review aims at providing valuable insights and serving as a starting point for researchers wanting to enhance explainability and decision transparency while solving bioinformatics problems. GitHub: https://github.com/rezacsedu/XAI-for-bioinformatics.

DiMo: discovery of microRNA motifs using deep learning and motif embedding.

MicroRNAs are small regulatory RNAs that decrease gene expression after transcription in various biological disciplines. In bioinformatics, identifying microRNAs and predicting their functionalities is critical. Finding motifs is one of the most well-known and important methods for identifying the functionalities of microRNAs. Several motif discovery techniques have been proposed, some of which rely on artificial intelligence-based techniques. However, in the case of few or no training data, their accuracy is low. In this research, we propose a new computational approach, called DiMo, for identifying motifs in microRNAs and generally macromolecules of small length. We employ word embedding techniques and deep learning models to improve the accuracy of motif discovery results. Also, we rely on transfer learning models to pre-train a model and use it in cases of a lack of (enough) training data. We compare our approach with five state-of-the-art works using three real-world datasets. DiMo outperforms the selected related works in terms of precision, recall, accuracy and f1-score.

Generative models for protein sequence modeling: recent advances and future directions.

The widespread adoption of high-throughput omics technologies has exponentially increased the amount of protein sequence data involved in many salient disease pathways and their respective therapeutics and diagnostics. Despite the availability of large-scale sequence data, the lack of experimental fitness annotations underpins the need for self-supervised and unsupervised machine learning (ML) methods. These techniques leverage the meaningful features encoded in abundant unlabeled sequences to accomplish complex protein engineering tasks. Proficiency in the rapidly evolving fields of protein engineering and generative AI is required to realize the full potential of ML models as a tool for protein fitness landscape navigation. Here, we support this work by (i) providing an overview of the architecture and mathematical details of the most successful ML models applicable to sequence data (e.g. variational autoencoders, autoregressive models, generative adversarial neural networks, and diffusion models), (ii) guiding how to effectively implement these models on protein sequence data to predict fitness or generate high-fitness sequences and (iii) highlighting several successful studies that implement these techniques in protein engineering (from paratope regions and subcellular localization prediction to high-fitness sequences and protein design rules generation). By providing a comprehensive survey of model details, novel architecture developments, comparisons of model applications, and current challenges, this study intends to provide structured guidance and robust framework for delivering a prospective outlook in the ML-driven protein engineering field.

Enhancing Arabidopsis thaliana ubiquitination site prediction through knowledge distillation and natural language processing.

Protein ubiquitination is a critical post-translational modification (PTM) involved in diverse biological processes and plays a pivotal role in regulating physiological mechanisms and disease states. Despite various efforts to develop ubiquitination site prediction tools across species, these tools mainly rely on predefined sequence features and machine learning algorithms, with species-specific variations in ubiquitination patterns remaining poorly understood. This study introduces a novel approach for predicting Arabidopsis thaliana ubiquitination sites using a neural network model based on knowledge distillation and natural language processing (NLP) of protein sequences. Our framework employs a multi-species "Teacher model" to guide a more compact, species-specific "Student model", with the "Teacher" generating pseudo-labels that enhance the "Student" learning and prediction robustness. Cross-validation results demonstrate that our model achieves superior performance, with an accuracy of 86.3 % and an area under the curve (AUC) of 0.926, while independent testing confirmed these results with an accuracy of 86.3 % and an AUC of 0.923. Comparative analysis with established predictors further highlights the model's superiority, emphasizing the effectiveness of integrating knowledge distillation and NLP in ubiquitination prediction tasks. This study presents a promising and efficient approach for ubiquitination site prediction, offering valuable insights for researchers in related fields. The code and resources are available on GitHub: https://github.com/nuinvtnu/KD_ArapUbi.

GPAD: a natural language processing-based application to extract the gene-disease association discovery information from OMIM.

BACKGROUND: Thousands of genes have been associated with different Mendelian conditions. One of the valuable sources to track these gene-disease associations (GDAs) is the Online Mendelian Inheritance in Man (OMIM) database. However, most of the information in OMIM is textual, and heterogeneous (e.g. summarized by different experts), which complicates automated reading and understanding of the data. Here, we used Natural Language Processing (NLP) to make a tool (Gene-Phenotype Association Discovery (GPAD)) that could syntactically process OMIM text and extract the data of interest. RESULTS: GPAD applies a series of language-based techniques to the text obtained from OMIM API to extract GDA discovery-related information. GPAD can inform when a particular gene was associated with a specific phenotype, as well as the type of validation-whether through model organisms or cohort-based patient-matching approaches-for such an association. GPAD extracted data was validated with published reports and was compared with large language model. Utilizing GPAD's extracted data, we analysed trends in GDA discoveries, noting a significant increase in their rate after the introduction of exome sequencing, rising from an average of about 150-250 discoveries each year. Contrary to hopes of resolving most GDAs for Mendelian disorders by now, our data indicate a substantial decline in discovery rates over the past five years (2017-2022). This decline appears to be linked to the increasing necessity for larger cohorts to substantiate GDAs. The rising use of zebrafish and Drosophila as model organisms in providing evidential support for GDAs is also observed. CONCLUSIONS: GPAD's real-time analyzing capacity offers an up-to-date view of GDA discovery and could help in planning and managing the research strategies. In future, this solution can be extended or modified to capture other information in OMIM and scientific literature.

Cotton peroxisome-localized lysophospholipase counteracts the toxic effects of Verticillium dahliae NLP1 and confers wilt resistance.

Plasma membrane represents a critical battleground between plants and attacking microbes. Necrosis-and-ethylene-inducing peptide 1 (Nep1)-like proteins (NLPs), cytolytic toxins produced by some bacterial, fungal and oomycete species, are able to target on lipid membranes by binding eudicot plant-specific sphingolipids (glycosylinositol phosphorylceramide) and form transient small pores, causing membrane leakage and subsequent cell death. NLP-producing phytopathogens are a big threat to agriculture worldwide. However, whether there are R proteins/enzymes that counteract the toxicity of NLPs in plants remains largely unknown. Here we show that cotton produces a peroxisome-localized enzyme lysophospholipase, GhLPL2. Upon Verticillium dahliae attack, GhLPL2 accumulates on the membrane and binds to V. dahliae secreted NLP, VdNLP1, to block its contribution to virulence. A higher level of lysophospholipase in cells is required to neutralize VdNLP1 toxicity and induce immunity-related genes expression, meanwhile maintaining normal growth of cotton plants, revealing the role of GhLPL2 protein in balancing resistance to V. dahliae and growth. Intriguingly, GhLPL2 silencing cotton plants also display high resistance to V. dahliae, but show severe dwarfing phenotype and developmental defects, suggesting GhLPL2 is an essential gene in cotton. GhLPL2 silencing results in lysophosphatidylinositol over-accumulation and decreased glycometabolism, leading to a lack of carbon sources required for plants and pathogens to survive. Furthermore, lysophospholipases from several other crops also interact with VdNLP1, implying that blocking NLP virulence by lysophospholipase may be a common strategy in plants. Our work demonstrates that overexpressing lysophospholipase encoding genes have great potential for breeding crops with high resistance against NLP-producing microbial pathogens.

MHCRoBERTa: pan-specific peptide-MHC class I binding prediction through transfer learning with label-agnostic protein sequences.

Predicting the binding of peptide and major histocompatibility complex (MHC) plays a vital role in immunotherapy for cancer. The success of Alphafold of applying natural language processing (NLP) algorithms in protein secondary struction prediction has inspired us to explore the possibility of NLP methods in predicting peptide-MHC class I binding. Based on the above motivations, we propose the MHCRoBERTa method, RoBERTa pre-training approach, for predicting the binding affinity between type I MHC and peptides. Analysis of the results on benchmark dataset demonstrates that MHCRoBERTa can outperform other state-of-art prediction methods with an increase of the Spearman rank correlation coefficient (SRCC) value. Notably, our model gave a significant improvement on IC50 value. Our method has achieved SRCC value and AUC value as 0.785 and 0.817, respectively. Our SRCC value is 14.3% higher than NetMHCpan3.0 (the second highest SRCC value on pan-specific) and is 3% higher than MHCflurry (the second highest SRCC value on all methods). The AUC value is also better than any other pan-specific methods. Moreover, we visualize the multi-head self-attention for the token representation across the layers and heads by this method. Through the analysis of the representation of each layer and head, we can show whether the model has learned the syntax and semantics necessary to perform the prediction task well. All these results demonstrate that our model can accurately predict the peptide-MHC class I binding affinity and that MHCRoBERTa is a powerful tool for screening potential neoantigens for cancer immunotherapy. MHCRoBERTa is available as an open source software at github (https://github.com/FuxuWang/MHCRoBERTa).

Diversified molecular adaptations of inorganic nitrogen assimilation and signaling machineries in plants.

Plants evolved sophisticated machineries to monitor levels of external nitrogen supply, respond to nitrogen demand from different tissues and integrate this information for coordinating its assimilation. Although roles of inorganic nitrogen in orchestrating developments have been studied in model plants and crops, systematic understanding of the origin and evolution of its assimilation and signaling machineries remains largely unknown. We expanded taxon samplings of algae and early-diverging land plants, covering all main lineages of Archaeplastida, and reconstructed the evolutionary history of core components involved in inorganic nitrogen assimilation and signaling. Most components associated with inorganic nitrogen assimilation were derived from the ancestral Archaeplastida. Improvements of assimilation machineries by gene duplications and horizontal gene transfers were evident during plant terrestrialization. Clusterization of genes encoding nitrate assimilation proteins might be an adaptive strategy for algae to cope with changeable nitrate availability in different habitats. Green plants evolved complex nitrate signaling machinery that was stepwise improved by domains shuffling and regulation co-option. Our study highlights innovations in inorganic nitrogen assimilation and signaling machineries, ranging from molecular modifications of proteins to genomic rearrangements, which shaped developmental and metabolic adaptations of plants to changeable nutrient availability in environments.

Gut neuroendocrine signaling regulates synaptic assembly in C. elegans.

Synaptic connections are essential to build a functional brain. How synapses are formed during development is a fundamental question in neuroscience. Recent studies provided evidence that the gut plays an important role in neuronal development through processing signals derived from gut microbes or nutrients. Defects in gut-brain communication can lead to various neurological disorders. Although the roles of the gut in communicating signals from its internal environment to the brain are well known, it remains unclear whether the gut plays a genetically encoded role in neuronal development. Using C. elegans as a model, we uncover that a Wnt-endocrine signaling pathway in the gut regulates synaptic development in the brain. A canonical Wnt signaling pathway promotes synapse formation through regulating the expression of the neuropeptides encoding gene nlp-40 in the gut, which functions through the neuronally expressed GPCR/AEX-2 receptor during development. Wnt-NLP-40-AEX-2 signaling likely acts to modulate neuronal activity. Our study reveals a genetic role of the gut in synaptic development and identifies a novel contribution of the gut-brain axis.

EHR-BERT: A BERT-based model for effective anomaly detection in electronic health records.

OBJECTIVE: Physicians and clinicians rely on data contained in electronic health records (EHRs), as recorded by health information technology (HIT), to make informed decisions about their patients. The reliability of HIT systems in this regard is critical to patient safety. Consequently, better tools are needed to monitor the performance of HIT systems for potential hazards that could compromise the collected EHRs, which in turn could affect patient safety. In this paper, we propose a new framework for detecting anomalies in EHRs using sequence of clinical events. This new framework, EHR-Bidirectional Encoder Representations from Transformers (BERT), is motivated by the gaps in the existing deep-learning related methods, including high false negatives, sub-optimal accuracy, higher computational cost, and the risk of information loss. EHR-BERT is an innovative framework rooted in the BERT architecture, meticulously tailored to navigate the hurdles in the contemporary BERT method; thus, enhancing anomaly detection in EHRs for healthcare applications. METHODS: The EHR-BERT framework was designed using the Sequential Masked Token Prediction (SMTP) method. This approach treats EHRs as natural language sentences and iteratively masks input tokens during both training and prediction stages. This method facilitates the learning of EHR sequence patterns in both directions for each event and identifies anomalies based on deviations from the normal execution models trained on EHR sequences. RESULTS: Extensive experiments on large EHR datasets across various medical domains demonstrate that EHR-BERT markedly improves upon existing models. It significantly reduces the number of false positives and enhances the detection rate, thus bolstering the reliability of anomaly detection in electronic health records. This improvement is attributed to the model's ability to minimize information loss and maximize data utilization effectively. CONCLUSION: EHR-BERT showcases immense potential in decreasing medical errors related to anomalous clinical events, positioning itself as an indispensable asset for enhancing patient safety and the overall standard of healthcare services. The framework effectively overcomes the drawbacks of earlier models, making it a promising solution for healthcare professionals to ensure the reliability and quality of health data.

Artificial intelligence-powered pharmacovigilance: A review of machine and deep learning in clinical text-based adverse drug event detection for benchmark datasets.

OBJECTIVE: The primary objective of this review is to investigate the effectiveness of machine learning and deep learning methodologies in the context of extracting adverse drug events (ADEs) from clinical benchmark datasets. We conduct an in-depth analysis, aiming to compare the merits and drawbacks of both machine learning and deep learning techniques, particularly within the framework of named-entity recognition (NER) and relation classification (RC) tasks related to ADE extraction. Additionally, our focus extends to the examination of specific features and their impact on the overall performance of these methodologies. In a broader perspective, our research extends to ADE extraction from various sources, including biomedical literature, social media data, and drug labels, removing the limitation to exclusively machine learning or deep learning methods. METHODS: We conducted an extensive literature review on PubMed using the query "(((machine learning [Medical Subject Headings (MeSH) Terms]) OR (deep learning [MeSH Terms])) AND (adverse drug event [MeSH Terms])) AND (extraction)", and supplemented this with a snowballing approach to review 275 references sourced from retrieved articles. RESULTS: In our analysis, we included twelve articles for review. For the NER task, deep learning models outperformed machine learning models. In the RC task, gradient Boosting, multilayer perceptron and random forest models excelled. The Bidirectional Encoder Representations from Transformers (BERT) model consistently achieved the best performance in the end-to-end task. Future efforts in the end-to-end task should prioritize improving NER accuracy, especially for 'ADE' and 'Reason'. CONCLUSION: These findings hold significant implications for advancing the field of ADE extraction and pharmacovigilance, ultimately contributing to improved drug safety monitoring and healthcare outcomes.

Accuracy of natural language processors for patients seeking inguinal hernia information.

BACKGROUND: NLPs such as ChatGPT are novel sources of online healthcare information that are readily accessible and integrated into internet search tools. The accuracy of NLP-generated responses to health information questions is unknown. METHODS: We queried four NLPs (ChatGPT 3.5 and 4, Bard, and Claude 2.0) for responses to simulated patient questions about inguinal hernias and their management. Responses were graded on a Likert scale (1 poor to 5 excellent) for relevance, completeness, and accuracy. Responses were compiled and scored collectively for readability using the Flesch-Kincaid score and for educational quality using the DISCERN instrument, a validated tool for evaluating patient information materials. Responses were also compared to two gold-standard educational materials provided by SAGES and the ACS. Evaluations were performed by six hernia surgeons. RESULTS: The average NLP response scores for relevance, completeness, and accuracy were 4.76 (95% CI 4.70-4.80), 4.11 (95% CI 4.02-4.20), and 4.14 (95% CI 4.03-4.24), respectively. ChatGPT4 received higher accuracy scores (mean 4.43 [95% CI 4.37-4.50]) than Bard (mean 4.06 [95% CI 3.88-4.26]) and Claude 2.0 (mean 3.85 [95% CI 3.63-4.08]). The ACS document received the best scores for reading ease (55.2) and grade level (9.2); however, none of the documents achieved the readibility thresholds recommended by the American Medical Association. The ACS document also received the highest DISCERN score of 63.5 (57.0-70.1), and this was significantly higher compared to ChatGPT 4 (50.8 [95% CI 46.2-55.4]) and Claude 2.0 (48 [95% CI 41.6-54.4]). CONCLUSIONS: The evaluated NLPs provided relevant responses of reasonable accuracy to questions about inguinal hernia. Compiled NLP responses received relatively low readability and DISCERN scores, although results may improve as NLPs evolve or with adjustments in question wording. As surgical patients expand their use of NLPs for healthcare information, surgeons should be aware of the benefits and limitations of NLPs as patient education tools.

Efficacy of ChatGPT in Cantonese Sentiment Analysis: Comparative Study.

BACKGROUND: Sentiment analysis is a significant yet difficult task in natural language processing. The linguistic peculiarities of Cantonese, including its high similarity with Standard Chinese, its grammatical and lexical uniqueness, and its colloquialism and multilingualism, make it different from other languages and pose additional challenges to sentiment analysis. Recent advances in models such as ChatGPT offer potential viable solutions. OBJECTIVE: This study investigated the efficacy of GPT-3.5 and GPT-4 in Cantonese sentiment analysis in the context of web-based counseling and compared their performance with other mainstream methods, including lexicon-based methods and machine learning approaches. METHODS: We analyzed transcripts from a web-based, text-based counseling service in Hong Kong, including a total of 131 individual counseling sessions and 6169 messages between counselors and help-seekers. First, a codebook was developed for human annotation. A simple prompt ("Is the sentiment of this Cantonese text positive, neutral, or negative? Respond with the sentiment label only.") was then given to GPT-3.5 and GPT-4 to label each message's sentiment. GPT-3.5 and GPT-4's performance was compared with a lexicon-based method and 3 state-of-the-art models, including linear regression, support vector machines, and long short-term memory neural networks. RESULTS: Our findings revealed ChatGPT's remarkable accuracy in sentiment classification, with GPT-3.5 and GPT-4, respectively, achieving 92.1% (5682/6169) and 95.3% (5880/6169) accuracy in identifying positive, neutral, and negative sentiment, thereby outperforming the traditional lexicon-based method, which had an accuracy of 37.2% (2295/6169), and the 3 machine learning models, which had accuracies ranging from 66% (4072/6169) to 70.9% (4374/6169). CONCLUSIONS: Among many text analysis techniques, ChatGPT demonstrates superior accuracy and emerges as a promising tool for Cantonese sentiment analysis. This study also highlights ChatGPT's applicability in real-world scenarios, such as monitoring the quality of text-based counseling services and detecting message-level sentiments in vivo. The insights derived from this study pave the way for further exploration into the capabilities of ChatGPT in the context of underresourced languages and specialized domains like psychotherapy and natural language processing.

Using Natural Language Processing (GPT-4) for Computed Tomography Image Analysis of Cerebral Hemorrhages in Radiology: Retrospective Analysis.

BACKGROUND: Cerebral hemorrhage is a critical medical condition that necessitates a rapid and precise diagnosis for timely medical intervention, including emergency operation. Computed tomography (CT) is essential for identifying cerebral hemorrhage, but its effectiveness is limited by the availability of experienced radiologists, especially in resource-constrained regions or when shorthanded during holidays or at night. Despite advancements in artificial intelligence-driven diagnostic tools, most require technical expertise. This poses a challenge for widespread adoption in radiological imaging. The introduction of advanced natural language processing (NLP) models such as GPT-4, which can annotate and analyze images without extensive algorithmic training, offers a potential solution. OBJECTIVE: This study investigates GPT-4's capability to identify and annotate cerebral hemorrhages in cranial CT scans. It represents a novel application of NLP models in radiological imaging. METHODS: In this retrospective analysis, we collected 208 CT scans with 6 types of cerebral hemorrhages at Ren Ji Hospital, Shanghai Jiao Tong University School of Medicine, between January and September 2023. All CT images were mixed together and sequentially numbered, so each CT image had its own corresponding number. A random sequence from 1 to 208 was generated, and all CT images were inputted into GPT-4 for analysis in the order of the random sequence. The outputs were subsequently examined using Photoshop and evaluated by experienced radiologists on a 4-point scale to assess identification completeness, accuracy, and success. RESULTS: The overall identification completeness percentage for the 6 types of cerebral hemorrhages was 72.6% (SD 18.6%). Specifically, GPT-4 achieved higher identification completeness in epidural and intraparenchymal hemorrhages (89.0%, SD 19.1% and 86.9%, SD 17.7%, respectively), yet its identification completeness percentage in chronic subdural hemorrhages was very low (37.3%, SD 37.5%). The misidentification percentages for complex hemorrhages (54.0%, SD 28.0%), epidural hemorrhages (50.2%, SD 22.7%), and subarachnoid hemorrhages (50.5%, SD 29.2%) were relatively high, whereas they were relatively low for acute subdural hemorrhages (32.6%, SD 26.3%), chronic subdural hemorrhages (40.3%, SD 27.2%), and intraparenchymal hemorrhages (26.2%, SD 23.8%). The identification completeness percentages in both massive and minor bleeding showed no significant difference (P=.06). However, the misidentification percentage in recognizing massive bleeding was significantly lower than that for minor bleeding (P=.04). The identification completeness percentages and misidentification percentages for cerebral hemorrhages at different locations showed no significant differences (all P>.05). Lastly, radiologists showed relative acceptance regarding identification completeness (3.60, SD 0.54), accuracy (3.30, SD 0.65), and success (3.38, SD 0.64). CONCLUSIONS: GPT-4, a standout among NLP models, exhibits both promising capabilities and certain limitations in the realm of radiological imaging, particularly when it comes to identifying cerebral hemorrhages in CT scans. This opens up new directions and insights for the future development of NLP models in radiology. TRIAL REGISTRATION: ClinicalTrials.gov NCT06230419; https://clinicaltrials.gov/study/NCT06230419.