Search

Objective: This study investigated the ability of Large Language Models (LLMs) to provide accurate and consistent answers by focusing on their performance in complex gynecologic cancer cases. Background: LLMs are advancing rapidly and require a thorough evaluation to ensure that they can be safely and effectively used in clinical decision-making. Such evaluations are essential for confirming LLM reliability and accuracy in supporting medical professionals in casework. Study design: We assessed three prominent LLMs—ChatGPT-4 (CG-4), Gemini Advanced (GemAdv), and Copilot—evaluating their accuracy, consistency, and overall performance. Fifteen clinical vignettes of varying difficulty and five open-ended questions based on real patient cases were used. The responses were coded, randomized, and evaluated blindly by six expert gynecologic oncologists using a 5-point Likert scale for relevance, clarity, depth, focus, and coherence. Results: GemAdv demonstrated superior accuracy (81.87 %) compared to both CG-4 (61.60 %) and Copilot (70.67 %) across all difficulty levels. GemAdv consistently provided correct answers more frequently (>60 % every day during the testing period). Although CG-4 showed a slight advantage in adhering to the National Comprehensive Cancer Network (NCCN) treatment guidelines, GemAdv excelled in the depth and focus of the answers provided, which are crucial aspects of clinical decision-making. Conclusion: LLMs, especially GemAdv, show potential in supporting clinical practice by providing accurate, consistent, and relevant information for gynecologic cancer. However, further refinement is needed for more complex scenarios. This study highlights the promise of LLMs in gynecologic oncology, emphasizing the need for ongoing development and rigorous evaluation to maximize their clinical utility and reliability.

Abstract This study explores disparities and opportunities in healthcare information provided by AI chatbots. We focused on recommendations for adjuvant therapy in endometrial cancer, analyzing responses across four regions (Indonesia, Nigeria, Taiwan, USA) and three platforms (Bard, Bing, ChatGPT-3.5). Utilizing previously published cases, we asked identical questions to chatbots from each location within a 24-h window. Responses were evaluated in a double-blinded manner on relevance, clarity, depth, focus, and coherence by ten experts in endometrial cancer. Our analysis revealed significant variations across different countries/regions (p

Autoantibodies against apolipoprotein A-I (ApoA-I) are associated with cardiovascular disease risks. We aimed to examine the 4-hydroxy-2-nonenal (HNE) modification of ApoA-I in coronary artery disease (CAD) and evaluate the potential risk of autoantibodies against their unmodified and HNE-modified peptides. We assessed plasma levels of ApoA-I, HNE-protein adducts, and autoantibodies against unmodified and HNE-peptide adducts, and significant correlations and odds ratios (ORs) were examined. Two novel CAD-specific HNE-peptide adducts, ApoA-I251–262 and ApoA-I70–83, were identified. Notably, immunoglobulin G (IgG) anti-ApoA-I251–262 HNE, IgM anti-ApoA-I70–83 HNE, IgG anti-ApoA-I251–262, IgG anti-ApoA-I70–83, and HNE-protein adducts were significantly correlated with triglycerides, creatinine, or high-density lipoprotein in CAD with various degrees of stenosis (70%). The HNE-protein adduct (OR = 2.208-fold, p = 0.020) and IgM anti-ApoA-I251–262 HNE (2.046-fold, p = 0.035) showed an increased risk of progression from >30% stenosis in CAD. HNE-protein adducts and IgM anti-ApoA-I251–262 HNE may increase the severity of CAD at high and low levels, respectively.

Abstract Background Individual organisms may exhibit phenotypic plasticity when they acclimate to different conditions. Such plastic responses may facilitate or constrain the adaptation of their descendant populations to new environments, complicating their evolutionary trajectories beyond the genetic blueprint. Intriguingly, phenotypic plasticity itself can evolve in terms of its direction and magnitude during adaptation. However, we know little about what determines the evolution of phenotypic plasticity, including gene expression plasticity. Recent laboratory-based studies suggest dominance of reversing gene expression plasticity—plastic responses that move the levels of gene expression away from the new optima. Nevertheless, evidence from natural populations is still limited. Results Here, we studied gene expression plasticity and its evolution in the montane and lowland populations of an elevationally widespread songbird—the Rufous-capped Babbler (Cyanoderma ruficeps)—with reciprocal transplant experiments and transcriptomic analyses; we set common gardens at altitudes close to these populations’ native ranges. We confirmed the prevalence of reversing plasticity in genes associated with altitudinal adaptation. Interestingly, we found a positive relationship between magnitude and degree of evolution in gene expression plasticity, which was pertinent to not only adaptation-associated genes but also the whole transcriptomes from multiple tissues. Furthermore, we revealed that genes with weaker expressional interactions with other genes tended to exhibit stronger plasticity and higher degree of plasticity evolution, which explains the positive magnitude-evolution relationship. Conclusions Our experimental evidence demonstrates that species may initiate their adaptation to new habitats with genes exhibiting strong expression plasticity. We also highlight the role of expression interdependence among genes in regulating the magnitude and evolution of expression plasticity. This study illuminates how the evolution of phenotypic plasticity in gene expression facilitates the adaptation of species to challenging environments in nature.

Abstract Counselling sessions have multiple stages, each with its themes and objectives. This study aimed to apply Hidden Markov Models (HMMs) to analyse counselling sessions from Open Up, an online text-based counselling platform in Hong Kong. The focus was on inferring latent stages over word distributions and identifying distinctive patterns of progression in more versus less satisfying sessions. Transcripts from 2589 sessions were categorized into more satisfying sessions ( $$n=\mathrm{1993}$$ n = 1993 ) and less satisfying sessions ( $$n=596$$ n = 596 ) based on post-session surveys. A message-level HMM identified five distinct stages: Rapport-building, Problem-identification, Problem-exploration, Problem-solving, and Wrap-up. Compared with less satisfying sessions, more satisfying sessions saw significantly more efficient initial rapport building (7.5% of session duration), problem introduction (20.2%), problem exploration (28.5%), elaborated solution development (46.6%), and concise conclusion (8.2%). This study offers insights for improving the efficiency and satisfaction of text-based counselling services through efficient initial engagement, thorough issue exploration, and focused problem-solving.

BackgroundSentiment 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. ObjectiveThis 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. MethodsWe 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. ResultsOur 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). ConclusionsAmong 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.