Evaluation of ChatGPT-generated medical responses: A systematic review and meta-analysis.

OBJECTIVE: Large language models (LLMs) such as ChatGPT are increasingly explored in medical domains. However, the absence of standard guidelines for performance evaluation has led to methodological inconsistencies. This study aims to summarize the available evidence on evaluating ChatGPT's performance in answering medical questions and provide direction for future research. METHODS: An extensive literature search was conducted on June 15, 2023, across ten medical databases. The keyword used was "ChatGPT," without restrictions on publication type, language, or date. Studies evaluating ChatGPT's performance in answering medical questions were included. Exclusions comprised review articles, comments, patents, non-medical evaluations of ChatGPT, and preprint studies. Data was extracted on general study characteristics, question sources, conversation processes, assessment metrics, and performance of ChatGPT. An evaluation framework for LLM in medical inquiries was proposed by integrating insights from selected literature. This study is registered with PROSPERO, CRD42023456327. RESULTS: A total of 3520 articles were identified, of which 60 were reviewed and summarized in this paper and 17 were included in the meta-analysis. ChatGPT displayed an overall integrated accuracy of 56 % (95 % CI: 51 %-60 %, I2 = 87 %) in addressing medical queries. However, the studies varied in question resource, question-asking process, and evaluation metrics. As per our proposed evaluation framework, many studies failed to report methodological details, such as the date of inquiry, version of ChatGPT, and inter-rater consistency. CONCLUSION: This review reveals ChatGPT's potential in addressing medical inquiries, but the heterogeneity of the study design and insufficient reporting might affect the results' reliability. Our proposed evaluation framework provides insights for the future study design and transparent reporting of LLM in responding to medical questions.

Clinical efficacy of cyclosporin and natamycin for fungal keratitis.

Objectives: To investigate the clinical efficacy of cyclosporin (CYSP) and natamycin (NAT) as a combination therapy in patients with fungal keratitis. Methods: This is a retrospective study. A total of 64 patients (64 eyes) with fungal keratitis treated by Baoding No.1 Central Hospital between December 2018 and May 2022 according to their treatment methods were divided into a monotherapy (MT) group receiving NAT eye drops solely and a combination therapy (CT) group given CYSP eye drops in addition to the exact treatment provided for the MT group. The clinical responses, visual acuity changes, severity of eye symptoms, and adverse reactions were compared between the two groups. Results: At two and four weeks post-treatment, the CT group had an overall response rate (ORR) significantly higher than that of the MT group (P< 0.05, respectively); both groups showed improved visual acuity and eye symptoms compared with the pre-treatment condition, and these improvements were more pronounced in the CT group (P < 0.05, respectively). Compared with the MT group, the CT group experienced a significantly shorter duration of eye symptoms (P < 0.05). The adverse reaction rate(ARR) was 9.38% in the CT group and 6.25% in the MT group, and the difference was not significant (P > 0.05). Conclusion: Using CYSP and NAT as a combination therapy for fungal keratitis can substantially heighten the therapeutic effects, promote visual acuity recovery, and induce rapid remission of eye symptoms without increasing the risk of adverse reactions.

[Serum folate and vitamin B12 levels and their association with neurodevelopmental features in preschool children with autism spectrum disorder]. / å­¦é¾„å‰å­¤ç‹¬ç—‡è°±ç³»éšœç¢å„¿ç«¥è¡€æ¸ å¶é ¸ã€ç»´ç”Ÿç´ B12æ°´å¹³åŠå ¶ä¸Žç¥žç»å‘è‚²ç‰¹å¾çš„å ³è”.

OBJECTIVES: To investigate the levels of serum folate and vitamin B12 (VB12) and their association with the level of neurodevelopment in preschool children with autism spectrum disorder (ASD). METHODS: A total of 324 ASD children aged 2-6 years and 318 healthy children aged 2-6 years were recruited. Serum levels of folate and VB12 were measured using chemiluminescent immunoassay. The Social Responsiveness Scale and the Childhood Autism Rating Scale were used to assess the core symptoms of ASD children, and the Gesell Developmental Schedule was employed to evaluate the level of neurodevelopment. RESULTS: The levels of serum folate and VB12 in ASD children were significantly lower than those in healthy children (P<0.05). Serum folate levels in ASD children were positively correlated with gross and fine motor developmental quotients (P<0.05), and serum VB12 levels were positively correlated with adaptive behavior, fine motor, and language developmental quotients (P<0.05). In ASD children aged 2 to <4 years, serum folate levels were positively correlated with developmental quotients in all domains (P<0.05), and serum VB12 levels were positively correlated with language developmental quotient (P<0.05). In male ASD children, serum VB12 levels were positively correlated with language and personal-social developmental quotients (P<0.05). CONCLUSIONS: Serum folate and VB12 levels in preschool ASD children are lower than those in healthy children and are associated with neurodevelopmental levels, especially in ASD children under 4 years of age. Therefore, maintaining normal serum folate and VB12 levels may be beneficial for the neurodevelopment of ASD children, especially in ASD children under 4 years of age.

Machine learning based on eye-tracking data to identify Autism Spectrum Disorder: A systematic review and meta-analysis.

BACKGROUND: Machine learning has been widely used to identify Autism Spectrum Disorder (ASD) based on eye-tracking, but its accuracy is uncertain. We aimed to summarize the available evidence on the performances of machine learning algorithms in classifying ASD and typically developing (TD) individuals based on eye-tracking data. METHODS: We searched Medline, Embase, Web of Science, Scopus, Cochrane Library, IEEE Xplore Digital Library, Wan Fang Database, China National Knowledge Infrastructure, Chinese BioMedical Literature Database, VIP Database for Chinese Technical Periodicals, from database inception to December 24, 2021. Studies using machine learning methods to classify ASD and TD individuals based on eye-tracking technologies were included. We extracted the data on study population, model performances, algorithms of machine learning, and paradigms of eye-tracking. This study is registered with PROSPERO, CRD42022296037. RESULTS: 261 articles were identified, of which 24 studies with sample sizes ranging from 28 to 141 were included (n = 1396 individuals). Machine learning based on eye-tracking yielded the pooled classified accuracy of 81 % (I2 = 73 %), specificity of 79 % (I2 = 61 %), and sensitivity of 84 % (I2 = 61 %) in classifying ASD and TD individuals. In subgroup analysis, the accuracy was 88 % (95 % CI: 85-91 %), 79 % (95 % CI: 72-84 %), 71 % (95 % CI: 59-91 %) for preschool-aged, school-aged, and adolescent-adult group. Eye-tracking stimuli and machine learning algorithms varied widely across studies, with social, static, and active stimuli and Support Vector Machine and Random Forest most commonly reported. Regarding the model performance evaluation, 15 studies reported their final results on validation datasets, four based on testing datasets, and five did not report whether they used validation datasets. Most studies failed to report the information on eye-tracking hardware and the implementation process. CONCLUSION: Using eye-tracking data, machine learning has shown potential in identifying ASD individuals with high accuracy, especially in preschool-aged children. However, the heterogeneity between studies, the absence of test set-based performance evaluations, the small sample size, and the non-standardized implementation of eye-tracking might deteriorate the reliability of results. Further well-designed and well-executed studies with comprehensive and transparent reporting are needed to determine the optimal eye-tracking paradigms and machine learning algorithms.

Natamycin versus natamycin combined with voriconazole in the treatment of fungal keratitis.

Objective: To observe the therapeutic effect of natamycin combined with voriconazole in the treatment of fungal keratitis (FK). Methods: This is a retrospective study. The subjects of this study were 64 patients with FK who were admitted to Baoding No.1 Central Hospital from February 2019 to July 2022. The enrolled patients were divided into control group (n= 32) and study group (n= 32) by the random number table method. The control group was treated with natamycin alone, and the study group was treated with natamycin combined with voriconazole. The total efficacy, time of disappearance of ocular symptoms, visual acuity level, keratitis severity score, corneal ulcer area, tear fungus index and incidence of adverse reactions were compared between the two groups. Results: The total efficacy,of the study group was significantly higher than that of the control group. The disappearance time of corneal ulcer, photophobia, foreign body sensation and hypopyon in the study group was shorter than those in the control group. Keratitis severity score and (1,3)-ß-D-glucan level in the study group were lower than those in the control group. The corneal ulcer area was narrowed in the study group than that of the control group, and the visual acuity level in the former group was higher than that of the latter group. Besides, there was no significant difference in the frequency of adverse reactions between the two groups. Conclusion: Natamycin combined with voriconazole is safe and effective in the treatment of FK.

[Value of autism screening checklists in the early identification of autism spectrum disorder].

OBJECTIVE: To evaluate the value of autism screening checklists in the early identification of autism spectrum disorder (ASD). METHODS: A total of 2 571 children who attended the Children's Hospital of Chongqing Medical University and completed autism screening and diagnostic test were enrolled as subjects, among whom 2 074 were diagnosed with ASD, 261 were diagnosed with global developmental delay (GDD), 206 were diagnosed with developmental language disorder (DLD), and 30 had normal development. The sensitivity, specificity, and optimal threshold value of the Modified Checklist for Autism in Toddlers (M-CHAT) and the Autism Behavior Checklist (ABC) for the early identification of ASD were evaluated by the receiver operating characteristic (ROC) curve. RESULTS: The M-CHAT had a high sensitivity of 88.3% but a low specificity of 36.0% for the identification of ASD. Its sensitivity decreased with age, and was maintained above 80% for children aged 16 to < 48 months. The ABC had a high specificity of 87.3% but a low sensitivity of 27.2%, with an optimal cut-off value of 47.5 based on the ROC curve analysis. The multivariate linear regression model based on a combination of the M-CHAT and ABC for screening of ASD showed a specificity of 85.8% and a sensitivity of 56.6%. CONCLUSIONS: The M-CHAT has a high sensitivity and a low specificity in the identification of ASD, with a better effect in children aged 16 to < 48 months. The ABC has a high specificity and a low sensitivity. The multiple linear regression model method based on the combined M-CHAT and ABC to screen ASD appears to be effective.

A comparative study on early vision quality after implantation of refractive segmental and diffractive multifocal intraocular lens.

OBJECTIVES: To compare early subjective and objective vision quality of postoperative patients undergoing phacoemulsification cataract surgery combined with implantation of refractive segmental multifocal intraocular lens (MIOL) SBL-3 and apodized diffractive MIOL SN6AD1. METHODS: As a prospective study, it enrolled 53 patients (53 eyes) to undergo phacoemulsification cataract surgery combined with MIOL implantation. According to differences in MIOL implanted, patients were divided into a SBL-3 group (25 eyes) and a SN6AD1 group (28 eyes). Ophthalmological evaluation included uncorrected (UDVA) and corrected (CDVA) distance visual acuities, uncorrected intermediate (UIVA) and near (UNVA) visual acuities, distance-corrected intermediate (DCIVA) and near (DCNVA)visual acuities and corrected near(CNVA) visual acuity, contrast sensitivity, modulation transfer function (MTF) and high order aberration (4 mm pupil diameter) at three months postoperatively. Moreover, a questionnaire survey was carried out to assess near spectacle independence, patient satisfaction and symptoms of visual disturbance. RESULTS: At three months after surgery, UIVA and UNVA in the SBL-3 group are statistically significantly superior to those of the SN6AD1 group (P>0.05). There was statistical difference in contrast sensitivity at four spatial frequencies (3, 6, 12, 18cycles/degree) under mesopic conditions and mesopic conditions with glare (P>0.05). The total ocular high order aberration, coma and trefoil were statistically significantly larger in the SBL-3 group than in the SN6AD1 group with 4.0 mm pupil diameters (P>0.05). Statistical differences were found in the MTF at spatial frequencies of 5, 10 and 15 cycles/degree between the groups. There were no significant differences in spectacle independence, patient satisfaction and visual disturbance between the groups (P>0.05). CONCLUSIONS: Both the two multifocal intraocular lens provided an excellent level of quality of vision three months postoperatively. However, the application effect of SBL-3 MIOL is superior to that of SN6AD1 MIOL as far as intermediate vision, near vision and contrast sensitivity are concerned.

HPLC with cellulose Tris (3,5-DimethylPhenylcarbamate) chiral stationary phase: Influence of coating times and coating amount on chiral discrimination.

Coating cellulose tris (3,5-dimethylphenylcarbamate) (CDMPC) on silica gels with large pores have been demonstrated as an efficient way for the preparation of chiral stationary phase (CSP) for high-performance liquid chromatography (HPLC). During the process, a number of parameters, including the type of coating solvent, amount of coating, and the method for subsequent solvent removing, have been proved to affect the performance of the resultant CSPs. Coating times and the concentration of coating solution, however, also makes a difference to CSPs' performance by changing the arrangement of cellulose derivatives while remaining the coating amount constant, have much less been studied before, and thereby, were systematically investigated in this work. Results showed that CSPs with more coating times exhibited higher chiral recognition and column efficiency, suggesting that resolution was determined by column efficiency herein. Afterwards, we also investigated the effect of coating amount on the performance of CSPs, and it was shown that the ability of enantio-recognition did not increase all the time as the coating amount; and four of seven racemates achieved best resolution when the coating amount reached to 18.37%. At the end, the reproducibility of CDMPC-coated CSPs were further confirmed by two methods, ie, reprepared the CSP-0.15-3 and reevaluated the effect of coating times.

Early identification of autism spectrum disorder based on machine learning with eye-tracking data.

BACKGROUND: Early identification of autism spectrum disorder (ASD) improves long-term outcomes, yet significant diagnostic delays persist. METHODS: A retrospective cohort of 449 children (ASD: 246, typically developing [TD]: 203) was used for model development. Eye-movement data were collected from the participants watching videos that featured eye-tracking paradigms for assessing social and non-social cognition. Five machine learning algorithms, namely random forest, support vector machine, logistic regression, artificial neural network, and extreme gradient boosting, were trained to classify children with ASD and TD. The best-performing algorithm was selected to build the final model which was further evaluated in a prospective cohort of 80 children. The Shapley values interpreted important eye-tracking features. RESULTS: Random forest outperformed other algorithms during model development and achieved an area under the curve of 0.849 (< 3 years: 0.832, ≥ 3 years: 0.868) on the external validation set. Of the ten most important eye-tracking features, three measured social cognition, and the rest were related to non-social cognition. A deterioration in model performance was observed using only the social or non-social cognition-related eye-tracking features. LIMITATIONS: The sample size of this study, although larger than that of existing studies of ASD based on eye-tracking data, was still relatively small compared to the number of features. CONCLUSIONS: Machine learning models based on eye-tracking data have the potential to be cost- and time-efficient digital tools for the early identification of ASD. Eye-tracking phenotypes related to social and non-social cognition play an important role in distinguishing children with ASD from TD children.

Association of the retinol to all-trans retinoic acid pathway with autism spectrum disorder.

BACKGROUND: Autism spectrum disorder (ASD) is a complex group of neurodevelopmental disorders. Research has highlighted a close association between the retinoic acid (RA) signaling pathway and ASD. This study investigates alterations in the vitamin A (VA, retinol) to RA metabolic pathway in children with ASD and speculates on the underlying reasons for these changes. We propose a subtype characterized by downregulated RA signaling in ASD, laying the groundwork for precise diagnosis and treatment research. METHODS: We included 489 children with ASD and 280 typically developing (TD) children. Those with ASD underwent evaluations of core symptoms and neuro-developmental levels, which were conducted by professional developmental behavior physicians using assessment scales. Serum VA and all-trans RA (atRA) levels were determined by high-performance liquid chromatography and ultra-high-performance liquid chromatography-tandem mass spectrometry. The expression levels and concentrations of enzyme molecules such as retinol dehydrogenase 10 were assessed using quantitative polymerase chain reaction and enzyme-linked immunosorbent assay. RESULTS: Children with ASD exhibited reduced serum atRA, accompanied by a downregulation of atRA synthesis enzymes. The reduction in serum atRA levels was linked not only to VA levels but also to the aberrant expression of metabolic enzymes responsible for atRA. Furthermore, the serum atRA levels in children with ASD were more strongly correlated with core symptoms and neurodevelopmental levels than VA levels. CONCLUSION: Children with ASD exhibited a dual regulation of reduced serum atRA levels, influenced by both VA levels and abnormal expression of atRA metabolic enzymes.

Predicting autism spectrum disorder using maternal risk factors: A multi-center machine learning study.

Autism spectrum disorder (ASD) is a neurodevelopmental disorder with a complex environmental etiology involving maternal risk factors, which have been combined with machine learning to predict ASD. However, limited studies have considered the factors throughout preconception, perinatal, and postnatal periods, and even fewer have been conducted in multi-center. In this study, five predictive models were developed using 57 maternal risk factors from a cohort across ten cities (ASD:1232, typically developing[TD]: 1090). The extreme gradient boosting model performed best, achieving an accuracy of 66.2 % on the external cohort from three cities (ASD:266, TD:353). The most important risk factors were identified as unstable emotions and lack of multivitamin supplementation using Shapley values. ASD risk scores were calculated based on predicted probabilities from the optimal model and divided into low, medium, and high-risk groups. The logistic analysis indicated that the high-risk group had a significantly increased risk of ASD compared to the low-risk group. Our study demonstrated the potential of machine learning models in predicting the risk for ASD based on maternal factors. The developed model provided insights into the maternal emotion and nutrition factors associated with ASD and highlighted the potential clinical applicability of the developed model in identifying high-risk populations.

Early identification of autism spectrum disorder by multi-instrument fusion: A clinically applicable machine learning approach.

Autism spectrum disorder (ASD), developmental language disorder (DLD), and global developmental delay (GDD) are common neurodevelopmental disorders in early childhood; however, the differential diagnosis of these disorders is difficult because of overlapping symptoms. Drawing on a cohort of 2004 children with ASD, DLD, or GDD, this study developed machine learning classifiers using decision trees, support vector machines, eXtreme gradient boosting (XGB), logistic regression, and neural networks by combining several easily accessible behavioral and developmental assessment instruments. The best-performing XGB model was further simplified into a two-stage decision model (TS-DM) to achieve better interpretability. Model performance was tested and compared with that of 12 pediatricians on an external dataset of 60 children. The accuracies of the resident pediatricians, senior pediatricians, TS-DM, and XGB were 53.3%, 66.7%, 75.0%, and 78.3%, respectively. Machine learning has the potential to identify these three neurodevelopmental disorders by integrating information from multiple instruments and thereby may increase our understanding of the roles of different behavioral and developmental characteristics in the different diagnoses.

NLRP3 Inflammasome-Mediated Pyroptosis Pathway Contributes to the Pathogenesis of Candida albicans Keratitis.

Purpose: Fungal keratitis is a sight-threatening corneal infection caused by fungal pathogens, and the pathogenic mechanisms have not been fully elucidated. The aim of this study was to determine whether NOD-like receptor family pyrin domain containing 3 (NLRP3) inflammasome-mediated pyroptosis contributes to Candida albicans (C. albicans) keratitis and explore the underlying mechanism. Methods: An in vivo mouse model of C. albicans keratitis and an in vitro culture model of human corneal epithelial cells (HCECs) challenged with heat-killed C. albicans (HKCA) were established in this study. The degree of corneal infection was evaluated by clinical scoring. Gene expression was assessed using reverse transcription-quantitative polymerase chain reaction (RT-qPCR) and western blot analysis or immunofluorescence staining was performed to evaluate protein expression. TdT-mediated dUTP nick end labeling (TUNEL) staining was performed to examine the pyroptotic cell death. A lactate dehydrogenase (LDH) release assay was performed to assess cytotoxicity. Results: Compared with the mock-infected group, we observed that the mRNA levels of NLRP3, caspase-1 (CASP1), interleukin (IL)-1ß and gasdermin-D (GSDMD) in C. albicans-infected mice cornea was significantly increased. Our data also demonstrated that the protein expression of NLRP3 and the pyroptosis-related markers apoptosis-associated speck-like protein containing a CARD (ASC), cleaved CASP1, N-GSDMD, cleaved IL-1ß and cleaved IL-18 as well as pyroptotic cell death were dramatically elevated in the mouse model of C. albicans keratitis. More importantly, NLRP3 knockdown markedly alleviated pyroptosis and consequently reduced corneal inflammatory reaction in C. albicans keratitis. In vitro, the presence of activated NLRP3 inflammasome and pyroptotic cell death were validated in HCECs exposed to HKCA. Furthermore, the potassium (K+) channel inhibitor glyburide decreased LDH release and suppressed NLRP3 inflammasome activation and pyroptosis in HCECs exposed to HKCA. Conclusion: In conclusion, the current study revealed for the first time that NLRP3 inflammasome activation and pyroptosis occur in C. albicans-infected mouse corneas and HCECs. Moreover, NLRP3 inflammasome-mediated pyroptosis signaling is involved in the disease severity of C. albicans keratitis. Therefore, This NLRP3 inflammasome-dependent pathway may be an attractive target for the treatment of fungal keratitis.

[The effect of vitreous surgery combined with intraocular lens implantation in proliferative diabetic retinopathy].

PURPOSE: To evaluate the effecting and complications of the vitreous surgery combined with intraocular lens implantation in proliferative diabetic retinopathy. METHODS: Forty-one eyes of 38 cases of proliferative diabetic retinopathy complicated with serious cataract were treated with phacoemulsification, pars plan vitctomy, peeling, photocoagulation, intraocular lens implantation and gas tampanade. Preoperative and postoperative visual acuity, anterior segment by slit lamp, fundus fluorescein angiography and surgery complications were analyzed. RESULTS: Average follow-up was 9 months. The visual acuity was improved in 36 eyes. The anterior response of operation was slight. The complications included cornea edema in 1 eye, high intraocular press in 1 eye, rubeosis in 1 eye and retinal detachment in 1 eye. CONCLUSIONS: The vitreous surgery combined with intraocular lens implantation was an effective operation for proliferative diabetic retinopathy with serious cataract.