Tipping point analysis for the between-arm correlation in an arm-based evidence synthesis.

Systematic reviews and meta-analyses are essential tools in contemporary evidence-based medicine, synthesizing evidence from various sources to better inform clinical decision-making. However, the conclusions from different meta-analyses on the same topic can be discrepant, which has raised concerns about their reliability. One reason is that the result of a meta-analysis is sensitive to factors such as study inclusion/exclusion criteria and model assumptions. The arm-based meta-analysis model is growing in importance due to its advantage of including single-arm studies and historical controls with estimation efficiency and its flexibility in drawing conclusions with both marginal and conditional effect measures. Despite its benefits, the inference may heavily depend on the heterogeneity parameters that reflect design and model assumptions. This article aims to evaluate the robustness of meta-analyses using the arm-based model within a Bayesian framework. Specifically, we develop a tipping point analysis of the between-arm correlation parameter to assess the robustness of meta-analysis results. Additionally, we introduce some visualization tools to intuitively display its impact on meta-analysis results. We demonstrate the application of these tools in three real-world meta-analyses, one of which includes single-arm studies.

Design and Synthesis of Aryl Amide Derivatives Containing Thiazole as Type III Secretion System Inhibitors against Pseudomonas aeruginosa.

To identify potent inhibitors of the type III secretion system (T3SS) in the foodborne pathogen Pseudomonas aeruginosa, we synthesized 35 thiazole-containing aryl amides by merging salicylic acid with various heterocycles through active splicing. Screening for exoS promoter activity led to the discovery of a highly effective T3SS inhibitor from these 35 compounds. Through subsequent experiments, it was confirmed that compound II-22 specifically targeted the T3SS of P. aeruginosa. Additionally, compound II-22 inhibited the secretion of the effector protein ExoS by modulating the CyaB-cAMP/Vfr-ExsA and ExsCED-ExsA regulatory pathways. Furthermore, compound II-22 suppressed the transcription of genes involved in the needle complex assembly, leading to reduced bacterial virulence. Further validation through inoculation tests using Galleria mellonella larvae demonstrated the strong in vivo efficacy of compound II-22. The study also revealed that compound II-22 enhanced the bactericidal activity of antibiotics, such as CIP (ciprofloxacin) and TOB (tobramycin). These results could help develop novel antimicrobial drugs to reduce bacterial resistance.

Biochar alters the soil fauna functional traits and community diversity: A quantitative and cascading perspective.

With the widespread use of biochar, the cascading effects of biochar exposure on soil fauna urgently require deeper understanding. A meta-analysis quantified hierarchical changes in functional traits and community diversity of soil fauna under biochar exposure. Antioxidant enzymes (24.1 %) did not fully mitigate the impact of MDA (13.5 %), leading to excessive DNA damage in soil fauna (21.2 %). Concurrently, reproduction, growth, and survival rates decreased by 20.2 %, 8.5 %, and 21.2 %, respectively. Due to a 39.7 % increase in avoidance behavior of soil fauna towards biochar, species richness ultimately increased by 80.2 %. Compared to other feeding habits, biochar posed a greater threat to the survival of herbivores. Additionally, macrofauna were the most sensitive to biochar. The response of soil fauna also depended on the type, size, concentration, and duration of biochar exposure. It should be emphasized that as exposure concentration increased, the damage to soil fauna became more severe. Furthermore, the smaller the biochar sizes, the greater the damage to soil fauna. To mitigate the adverse effects on soil fauna, this study recommens applying biochar at appropriate times and selecting large sizes in low to medium concentrations. These findings confirm the threat of biochar to soil health from the perspective of soil fauna.

CBIL-VHPLI: a model for predicting viral-host protein-lncRNA interactions based on machine learning and transfer learning.

Virus‒host protein‒lncRNA interaction (VHPLI) predictions are critical for decoding the molecular mechanisms of viral pathogens and host immune processes. Although VHPLI interactions have been predicted in both plants and animals, they have not been extensively studied in viruses. For the first time, we propose a new deep learning-based approach that consists mainly of a convolutional neural network and bidirectional long and short-term memory network modules in combination with transfer learning named CBIL‒VHPLI to predict viral-host protein‒lncRNA interactions. The models were first trained on large and diverse datasets (including plants, animals, etc.). Protein sequence features were extracted using a k-mer method combined with the one-hot encoding and composition-transition-distribution (CTD) methods, and lncRNA sequence features were extracted using a k-mer method combined with the one-hot encoding and Z curve methods. The results obtained on three independent external validation datasets showed that the pre-trained CBIL‒VHPLI model performed the best with an accuracy of approximately 0.9. Pretraining was followed by conducting transfer learning on a viral protein-human lncRNA dataset, and the fine-tuning results showed that the accuracy of CBIL‒VHPLI was 0.946, which was significantly greater than that of the previous models. The final case study results showed that CBIL‒VHPLI achieved a prediction reproducibility rate of 91.6% for the RIP-Seq experimental screening results. This model was then used to predict the interactions between human lncRNA PIK3CD-AS2 and the nonstructural protein 1 (NS1) of the H5N1 virus, and RNA pull-down experiments were used to prove the prediction readiness of the model in terms of prediction. The source code of CBIL‒VHPLI and the datasets used in this work are available at https://github.com/Liu-Lab-Lnu/CBIL-VHPLI for academic usage.

Regulation of cancer cell ferroptosis by PTRF/Cavin-1.

Ovarian cancer, marked by high rate of recurrence, novel therapeutic strategies are needed to improve patient outcome. One of the potential strategies is inducing ferroptosis in ovarian cancer cells. Ferroptosis is an iron-dependent, lipid peroxidation-driven mode of cell death primarily occurring on the cell membrane. PTRF, an integral component of the caveolae structures located on the cell membrane, is involved in a multitude of physiological processes, including but not limited to, endocytosis, signal transduction, and lipid metabolism. This study elucidates the relationship between PTRF and ferroptosis in ovarian cancer, offering a fresh perspective for the development of new therapeutic strategies. We knocked down PTRF employing siRNA in the ovarian cancer cell lines HEY and SKOV3, following which we stimulated ferroptosis with Erastin (Era). Our research indicates that the lack of PTRF sensitizes cancer cells to ferroptosis, likely by altering membrane stability and tension, thereby affecting signal pathways related to ferroptosis, such as lipid and atherosclerosis, fluid shear stress, and atherosclerosis. Our findings provide new insights for developing new treatments for ovarian cancer.

Epidermolysis Bullosa Pruriginosa treated with baricitinib: A case report.

INTRODUCTION: Epidermolysis Bullosa Pruriginosa (EBP) is a persistent, recurring disease that seriously affects quality of life. Fewer than 100 cases of EBP have been reported to date. Numerous inflammatory dermatoses are driven by soluble inflammatory mediators, which rely on Janus kinase-signal transducer and activator of transcription (JAK-STAT) signaling, and inhibition of this pathway using Janus kinase (JAK) inhibitors might be a useful therapeutic strategy for these diseases. PATIENT CONCERNS: A male patient, 28 years of age, was admitted to our hospital because of recurrent papules, nodules, and intense itching on the trunk and extremities for 12 years. Repeated large and intense itching has seriously affected the patient normal life. DIAGNOSIS: The patient was diagnosed with EBP based on examination results. INTERVENTIONS: Oral baricitinib tablets (2 mg, once a day) + Oral desloratadine citrate disodium tablets (8.8 mg, once a day) combined with topical compound flumethasone ointment and Fucidin cream. OUTCOMES: The patient skin rashes had subsided and flattened remarkable, and his itching was markedly relieved. The visual analogue scale (VAS) itching score of the patient gradually declined from 8 to 9 points to 2 to 3 points. CONCLUSION: This study confirms that baricitinib is effective and feasible in treating EBP, especially in remarkable relieving itching, which rendered new ideas for therapeutic approaches for EBP in the future.

Stratifying heart failure patients with graph neural network and transformer using Electronic Health Records to optimize drug response prediction.

OBJECTIVES: Heart failure (HF) impacts millions of patients worldwide, yet the variability in treatment responses remains a major challenge for healthcare professionals. The current treatment strategies, largely derived from population based evidence, often fail to consider the unique characteristics of individual patients, resulting in suboptimal outcomes. This study aims to develop computational models that are patient-specific in predicting treatment outcomes, by utilizing a large Electronic Health Records (EHR) database. The goal is to improve drug response predictions by identifying specific HF patient subgroups that are likely to benefit from existing HF medications. MATERIALS AND METHODS: A novel, graph-based model capable of predicting treatment responses, combining Graph Neural Network and Transformer was developed. This method differs from conventional approaches by transforming a patient's EHR data into a graph structure. By defining patient subgroups based on this representation via K-Means Clustering, we were able to enhance the performance of drug response predictions. RESULTS: Leveraging EHR data from 11 627 Mayo Clinic HF patients, our model significantly outperformed traditional models in predicting drug response using NT-proBNP as a HF biomarker across five medication categories (best RMSE of 0.0043). Four distinct patient subgroups were identified with differential characteristics and outcomes, demonstrating superior predictive capabilities over existing HF subtypes (best mean RMSE of 0.0032). DISCUSSION: These results highlight the power of graph-based modeling of EHR in improving HF treatment strategies. The stratification of patients sheds light on particular patient segments that could benefit more significantly from tailored response predictions. CONCLUSIONS: Longitudinal EHR data have the potential to enhance personalized prognostic predictions through the application of graph-based AI techniques.

Induction Immunochemotherapy Yields a Higher Conversion Rate and Better Overall Survival than Chemotherapy in Initially Unresectable Esophageal Squamous Cell Carcinoma.

INTRODUCTION: This study compared the surgical conversion rate and overall survival (OS) between induction chemotherapy (iC) and induction immunochemotherapy (iIC) for patients with initially unresectable esophageal squamous cell carcinoma (iuESCC). METHODS: In this multicenter, retrospective cohort study, patients from four high-volume institutions with unresectable diseases were included. The primary endpoints were the conversion surgery rate and OS. A multivariate Cox regression analysis was used to identify the independent significant prognostic factors associated with OS. The stabilized inverse probability of treatment weighting was applied to confirm the survival comparison between the iIC and iC cohorts. RESULTS: A total of 309 patients (150 in the iIC cohort and 159 in the iC cohort) were included. A significantly higher conversion surgical rate was observed in the iIC cohort (iIC vs. iC: 127/150, 84.7% vs. 79/159, 49.7%, P < 0.001). The pathological complete response rates were 22.0% and 5.1% in the iIC and the iC cohorts, respectively (P = 0.001). A significant difference in the OS was observed between the iIC (not reached) and iC cohorts (median 95% CI 36.3 [range 27.2-45.5]). The stabilized inverse probability of treatment weighting yielded similar results. Regimen (iIC vs. iC, HR 0.215, 95% CI 0.102-0.454, P < 0.001) and operation (yes vs. no, HR 0.262, 95% CI 0.161-0.427, P < 0.001) were the significant prognostic factors for OS. CONCLUSIONS: Immunochemotherapy plus conversion surgery in the induction setting may be a better treatment option to achieve high pathological responses and improve OS in iuESCC patients.

Synoptic Reporting by Summarizing Cancer Pathology Reports using Large Language Models.

Background: Synoptic reporting, the documenting of clinical information in a structured manner, is known to improve patient care by reducing errors, increasing readability, interoperability, and report completeness. Despite its advantages, manually synthesizing synoptic reports from narrative reports is expensive and error prone when the number of structured fields are many. While the recent revolutionary developments in Large Language Models (LLMs) have significantly advanced natural language processing, their potential for innovations in medicine is yet to be fully evaluated. Objectives: In this study, we explore the strengths and challenges of utilizing the state-of-the-art language models in the automatic synthesis of synoptic reports. Materials and Methods: We use a corpus of 7,774 cancer related, narrative pathology reports, which have annotated reference synoptic reports from Mayo Clinic EHR. Using these annotations as a reference, we reconfigure the state-of-the-art large language models, such as LLAMA-2, to generate the synoptic reports. Our annotated reference synoptic reports contain 22 unique data elements. To evaluate the accuracy of the reports generated by the LLMs, we use several metrics including the BERT F1 Score and verify our results by manual validation. Results: We show that using fine-tuned LLAMA-2 models, we can obtain BERT Score F1 of 0.86 or higher across all data elements and BERT F1 scores of 0.94 or higher on over 50% (11 of 22) of the questions. The BERT F1 scores translate to average accuracies of 76% and as high as 81% for short clinical reports. Conclusions: We demonstrate successful automatic synoptic report generation by fine-tuning large language models.

Age-related trajectories of the development of social cognition.

Age-related trajectories of intrinsic functional connectivity (iFC), which represent the interconnections between discrete regions of the human brain, for processes related to social cognition (SC) provide evidence for social development through neural imaging and can guide clinical interventions when such development is atypical. However, due to the lack of studies investigating brain development over a wide range of ages, the neural mechanisms of SC remain poorly understood, although considerable behavior-related evidence is available. The present study mapped vortex-wise iFC features between SC networks and the entire cerebral cortex by using common functional networks, creating the corresponding age-related trajectories. Three networks [moral cognition, theory of mind (ToM), and empathy] were selected as representative SC networks. The Enhanced Nathan Kline Institute-Rockland Sample (NKI-RS, N = 316, ages 8-83 years old) was employed delineate iFC characteristics and construct trajectories. The results showed that the SC networks display unique and overlapping iFC profiles. The iFC of the empathy network, an age-sensitive network, with dorsal attention network was found to exhibit a linear increasing pattern, that of the ventral attention network was observed to exhibit a linear decreasing pattern, and that of the somatomotor and dorsal attention networks was noted to exhibit a quadric-concave iFC pattern. Additionally, a sex-specific effect was observed for the empathy network as it exhibits linear and quadric sex-based differences in iFC with the frontoparietal and vision networks, respectively. The iFC of the ToM network with the ventral attention network exhibits a pronounced quadric-convex (inverted U-shape) trajectory. No linear or quadratic trajectories were noted in the iFC of the moral cognition network. These findings indicate that SC networks exhibit iFC with both low-level (somatomotor, vision) and high-level (attention and control) networks along specific developmental trajectories. The age-related trajectories determined in this study advance our understanding of the neural mechanisms of SC, providing valuable references for identification and intervention in cases of development of atypical SC.

Hot spots and trends in microbial disease research on cultural heritage: a bibliometric analysis.

This study is to understand and analyze the development history, research hotspots, and research trends in the study of microbial diseases of cultural heritage through bibliometric analyses in order to fill the current gap of no literature review in this research field and to make certain contributions to the research in this field and the protection of cultural heritage. Bibliometric and visual analyses of the literature on cultural heritage microbial diseases in the Web of Science (WoS) core collection were carried out using VOSviewer and R-bibliometrix, choosing the two main literature types of papers and reviews. The emphasis was placed on analyzing and summarizing core research strengths, hotspots, and trends. Six hundred sixty-seven documents (573 articles and 94 reviews) were retrieved. αIn the WoS core collection, the first literature on cultural heritage microbial disease research was published in January 2000, and the annual number of publications from 2000 to 2009 did not exceed one; the annual number of publications from 2010 onwards increased rapidly, and after 2018, the number of publications per year exceeded 60, reaching 94 in 2020, which indicates that cultural heritage microbial disease research is booming. Our research showed that Italy, the USA, and China were the leading research countries, and Univ Milan was the institution with the most publications. International Biodeterioration &Biodegradation was the most published and co-cited journal, and Gu JD was the most prolific author. The research hotspots in the study of microbial diseases of cultural heritage mainly include biological degradation of cultural heritage; identification of diseased microorganisms and disease mechanisms; cultural heritage microbial disease prevention and control methods; monitoring, prevention, and control of diseased microorganisms in indoor air; antibacterial agents, especially essential oils, nanoparticles, and other safe and efficient antibacterial products research and development; and exploration of the mechanisms of biofilm protection of cultural heritage on cultural heritage surfaces. Monitoring and identifying cultural heritage microbial communities, identifying disease mechanisms, and researching safe and efficient bacteriostatic products such as essential oils and nanoparticles will be the main research directions in the field of cultural heritage microbial disease prevention and control in the future.

Best acupuncture method for mammary gland hyperplasia: Evaluation of randomized controlled trials and Bayesian network meta-analysis.

Objective: To evaluate the effectiveness of different acupuncture treatments for mammary gland hyperplasia (MGH) using a network meta-analysis. Methods: Several databases were searched without language restrictions from 2000 to February 2023, including PubMed, Embase, Web of Science, Cochrane Library, China Science and Technology Journal Database, China Biology Medicine Database, Wanfang Database, China National Knowledge Infrastructure Database, and other professional websites and gray literature. Inclusion criteria were adult women diagnosed with MGH; intervention measures included acupuncture and related therapies; the control group was treated with simple drugs; and the research type was a randomized controlled trial (RCT). The primary outcomes were treatment effectiveness and estradiol and progesterone levels. Secondary outcomes were breast lump size and visual analog scale (VAS) score of breast pain. Exclusion criteria were studies unrelated to MGH, incorrect study populations, control measures or interventions, incomplete data, non-RCTs, case reports, and animal experiments. Cochrane tools were used to assess the risk of bias. The R software (x64 version 4.2.1), Review Manager 5.3 software and STATA 16.0 software were used for data analysis. Results: Following a rigorous screening process, data extraction, and quality assessment, 48 eligible RCTs encompassing 4,500 patients with MGH and 16 interventions were included. The results indicated that acupuncture, alone or in combination with traditional Chinese or Western medicine, had better therapeutic effects than conventional therapy. In terms of effectiveness, warm needle acupuncture was the best choice (94.6%). Bloodletting pricking was the most effective method (85.7%) for lowering progesterone levels. Bloodletting pricking was the most effective method (98.3%) for lowering estradiol levels. Manual acupuncture combined with traditional Chinese medicine was the most effective (74.5%) treatment to improve the size of the breast lump. Warm needle acupuncture was the most effective (69.8%) in improving the VAS score. Conclusion: Acupuncture therapy was more effective in treating MGH than drug therapy alone, and warm needle acupuncture and bloodletting pricking were the two best options. However, larger sample sizes and high-quality RCTs are required.

Discovery of novel amide derivatives as potent quorum sensing inhibitors of Pseudomonas aeruginosa.

With the increasing reports of antibiotic resistance in this species, Pseudomonas aeruginosa is a common human pathogen with important implications for public health. Bacterial quorum sensing (QS) systems are potentially broad and versatile targets for developing new antimicrobial compounds. While previous reports have demonstrated that certain amide compounds can inhibit bacterial growth, there are few reports on the specific inhibitory effects of these compounds on bacterial quorum sensing systems. In this study, thirty-one amide derivatives were synthesized. The results of the biological activity assessment indicated that A9 and B6 could significantly inhibit the expression of lasB, rhlA, and pqsA, effectively reducing several virulence factors regulated by the QS systems of PAO1. Additionally, compound A9 attenuated the pathogenicity of PAO1 to Galleria mellonella larvae. Meanwhile, RT-qPCR, SPR, and molecular docking studies were conducted to explore the mechanism of these compounds, which suggests that compound A9 inhibited the QS systems by binding with LasR and PqsR, especially PqsR. In conclusion, amide derivatives A9 and B6 exhibit promising potential for further development as novel QS inhibitors in P. aeruginosa.

Metal-Graphene Hybrid Terahertz Metasurfaces for Circulating Tumor DNA Detection Based on Dual Signal Amplification.

Terahertz (THz) spectroscopy has impressive capability for label-free biosensing, but its utility in clinical laboratories is rarely reported due to often unsatisfactory detection performances. Here, we fabricated metal-graphene hybrid THz metasurfaces (MSs) for the sensitive and enzyme-free detection of circulating tumor DNA (ctDNA) in pancreatic cancer plasma samples. The feasibility and mechanism of the enhanced effects of a graphene bridge across the MS and amplified by gold nanoparticles (AuNPs) were investigated experimentally and theoretically. The AuNPs serve to boost charge injection in the graphene film and result in producing a remarkable change in the graded transmissivity index to THz radiation of the MS resonators. Assay design utilizes this feature and a cascade hybridization chain reaction initiated on magnetic beads in the presence of target ctDNA to achieve dual signal amplification (chemical and optical). In addition to demonstrating subfemtomolar detection sensitivity and single-nucleotide mismatch selectivity, the proposed method showed remarkable capability to discriminate between pancreatic cancer patients and healthy individuals by recognizing and quantifying targeted ctDNAs. The introduction of graphene to the metasurface produces an improved sensitivity of 2 orders of magnitude for ctDNA detection. This is the first study to report the combined application of graphene and AuNPs in biosensing by THz spectroscopic resonators and provides a combined identification scheme to detect and discriminate different biological analytes, including nucleic acids, proteins, and various biomarkers.

Quality of Answers of Generative Large Language Models Versus Peer Users for Interpreting Laboratory Test Results for Lay Patients: Evaluation Study.

BACKGROUND: Although patients have easy access to their electronic health records and laboratory test result data through patient portals, laboratory test results are often confusing and hard to understand. Many patients turn to web-based forums or question-and-answer (Q&A) sites to seek advice from their peers. The quality of answers from social Q&A sites on health-related questions varies significantly, and not all responses are accurate or reliable. Large language models (LLMs) such as ChatGPT have opened a promising avenue for patients to have their questions answered. OBJECTIVE: We aimed to assess the feasibility of using LLMs to generate relevant, accurate, helpful, and unharmful responses to laboratory test-related questions asked by patients and identify potential issues that can be mitigated using augmentation approaches. METHODS: We collected laboratory test result-related Q&A data from Yahoo! Answers and selected 53 Q&A pairs for this study. Using the LangChain framework and ChatGPT web portal, we generated responses to the 53 questions from 5 LLMs: GPT-4, GPT-3.5, LLaMA 2, MedAlpaca, and ORCA_mini. We assessed the similarity of their answers using standard Q&A similarity-based evaluation metrics, including Recall-Oriented Understudy for Gisting Evaluation, Bilingual Evaluation Understudy, Metric for Evaluation of Translation With Explicit Ordering, and Bidirectional Encoder Representations from Transformers Score. We used an LLM-based evaluator to judge whether a target model had higher quality in terms of relevance, correctness, helpfulness, and safety than the baseline model. We performed a manual evaluation with medical experts for all the responses to 7 selected questions on the same 4 aspects. RESULTS: Regarding the similarity of the responses from 4 LLMs; the GPT-4 output was used as the reference answer, the responses from GPT-3.5 were the most similar, followed by those from LLaMA 2, ORCA_mini, and MedAlpaca. Human answers from Yahoo data were scored the lowest and, thus, as the least similar to GPT-4-generated answers. The results of the win rate and medical expert evaluation both showed that GPT-4's responses achieved better scores than all the other LLM responses and human responses on all 4 aspects (relevance, correctness, helpfulness, and safety). LLM responses occasionally also suffered from lack of interpretation in one's medical context, incorrect statements, and lack of references. CONCLUSIONS: By evaluating LLMs in generating responses to patients' laboratory test result-related questions, we found that, compared to other 4 LLMs and human answers from a Q&A website, GPT-4's responses were more accurate, helpful, relevant, and safer. There were cases in which GPT-4 responses were inaccurate and not individualized. We identified a number of ways to improve the quality of LLM responses, including prompt engineering, prompt augmentation, retrieval-augmented generation, and response evaluation.

Assessing the Seasonality of Lab Tests Among Patients with Alzheimer's Disease and Related Dementias in OneFlorida Data Trust.

About 1 in 9 older adults over 65 has Alzheimer's disease (AD), many of whom also have multiple other chronic conditions such as hypertension and diabetes, necessitating careful monitoring through laboratory tests. Understanding the patterns of laboratory tests in this population aids our understanding and management of these chronic conditions along with AD. In this study, we used an unimodal cosinor model to assess the seasonality of lab tests using electronic health record (EHR) data from 34,303 AD patients from the OneFlorida+ Clinical Research Consortium. We observed significant seasonal fluctuations-higher in winter in lab tests such as glucose, neutrophils per 100 white blood cells (WBC), and WBC. Notably, certain leukocyte types like eosinophils, lymphocytes, and monocytes are elevated during summer, likely reflecting seasonal respiratory diseases and allergens. Seasonality is more pronounced in older patients and varies by gender. Our findings suggest that recognizing these patterns and adjusting reference intervals for seasonality would allow healthcare providers to enhance diagnostic precision, tailor care, and potentially improve patient outcomes.

Quality of Answers of Generative Large Language Models vs Peer Patients for Interpreting Lab Test Results for Lay Patients: Evaluation Study.

Background: Even though patients have easy access to their electronic health records and lab test results data through patient portals, lab results are often confusing and hard to understand. Many patients turn to online forums or question and answering (Q&A) sites to seek advice from their peers. However, the quality of answers from social Q&A on health-related questions varies significantly, and not all the responses are accurate or reliable. Large language models (LLMs) such as ChatGPT have opened a promising avenue for patients to get their questions answered. Objective: We aim to assess the feasibility of using LLMs to generate relevant, accurate, helpful, and unharmful responses to lab test-related questions asked by patients and to identify potential issues that can be mitigated with augmentation approaches. Methods: We first collected lab test results related question and answer data from Yahoo! Answers and selected 53 Q&A pairs for this study. Using the LangChain framework and ChatGPT web portal, we generated responses to the 53 questions from four LLMs including GPT-4, Meta LLaMA 2, MedAlpaca, and ORCA_mini. We first assessed the similarity of their answers using standard QA similarity-based evaluation metrics including ROUGE, BLEU, METEOR, BERTScore. We also utilized an LLM-based evaluator to judge whether a target model has higher quality in terms of relevance, correctness, helpfulness, and safety than the baseline model. Finally, we performed a manual evaluation with medical experts for all the responses of seven selected questions on the same four aspects. Results: Regarding the similarity of the responses from 4 LLMs, where GPT-4 output was used as the reference answer, the responses from LLaMa 2 are the most similar ones, followed by LLaMa 2, ORCA_mini, and MedAlpaca. Human answers from Yahoo data were scored lowest and thus least similar to GPT-4-generated answers. The results of Win Rate and medical expert evaluation both showed that GPT-4's responses achieved better scores than all the other LLM responses and human responses on all the four aspects (relevance, correctness, helpfulness, and safety). However, LLM responses occasionally also suffer from lack of interpretation in one's medical context, incorrect statements, and lack of references. Conclusions: By evaluating LLMs in generating responses to patients' lab test results related questions, we find that compared to other three LLMs and human answer from the Q&A website, GPT-4's responses are more accurate, helpful, relevant, and safer. However, there are cases that GPT-4 responses are inaccurate and not individualized. We identified a number of ways to improve the quality of LLM responses including prompt engineering, prompt augmentation, retrieval augmented generation, and response evaluation.

Quantum imaging of biological organisms through spatial and polarization entanglement.

Quantum imaging holds potential benefits over classical imaging but has faced challenges such as poor signal-to-noise ratios, low resolvable pixel counts, difficulty in imaging biological organisms, and inability to quantify full birefringence properties. Here, we introduce quantum imaging by coincidence from entanglement (ICE), using spatially and polarization-entangled photon pairs to overcome these challenges. With spatial entanglement, ICE offers higher signal-to-noise ratios, greater resolvable pixel counts, and the ability to image biological organisms. With polarization entanglement, ICE provides quantitative quantum birefringence imaging capability, where both the phase retardation and the principal refractive index axis angle of an object can be remotely and instantly quantified without changing the polarization states of the photons incident on the object. Furthermore, ICE enables 25 times greater suppression of stray light than classical imaging. ICE has the potential to pave the way for quantum imaging in diverse fields, such as life sciences and remote sensing.