Performance of GPT-4 and GPT-3.5 in generating accurate and comprehensive diagnoses across medical subspecialties.

Artificial intelligence has demonstrated a promising potential for diagnosing complex medical cases, with Generative Pre-Trained Transformer 4 (GPT-4) being the most recent advancement in this field. This study evaluated the diagnostic performance of the GPT-4 in comparison with that of its predecessor, GPT-3.5, using 81 complex medical case records from the New England Journal of Medicine . The cases were categorized as cognitive impairment, infectious disease, rheumatology, or drug reactions. The GPT-4 achieved a primary diagnostic accuracy of 38.3%, which improved to 71.6% when differential diagnoses were included. In 84.0% of cases, primary diagnoses were made by conducting investigations suggested by GPT-4. GPT-4 outperformed GPT-3.5 in all subspecialties except for drug reactions. GPT-4 demonstrated the highest performance in infectious diseases and drug reactions, whereas it underperformed in cases of cognitive impairment. These findings indicate that GPT-4 can provide reasonably accurate diagnoses, comprehensive differential diagnoses, and appropriate investigations. However, its performance varies across subspecialties.

Limitations of GPT-4 as a geriatrician in geri-oncology case conference: A case series.

Generative pre-trained transformer 4 (GPT-4) is an artificial intelligence (AI) system with a chat interface. The number of studies testing GPT-4 in clinical applications has been increasing. We hypothesized that GPT-4 would be able to suggest management strategies for medical issues in elderly oncology patients, similar to those provided by geriatricians. We compared the responses of GPT-4 to those of a geriatrician for four oncological patients. After these case conferences, none of the patients required admission for medical consultation. In three out of four scenarios, GPT-4 was able to offer a multidisciplinary approach in the first prompt. In all three scenarios, GPT-4 identified medication-related side effects and suggested appropriate medications in the first prompt. However, GPT-4 was unable to suggest initial dosages of medications to be used in the first prompt and was unable to suggest a more humanistic and non-pharmacological approach to anorexia, even with a follow-up prompt. In conclusion, GPT-4 may be used as a screening tool to provide potential rudimentary directions for management, which can then be reviewed by medical professionals before considering a formal consultation for more tailored and refined opinions from specialists.

Use of GPT-4 to Analyze Medical Records of Patients With Extensive Investigations and Delayed Diagnosis.

This case series investigates whether analysis of clinical history via a language model system improves diagnostic accuracy in patients with complex and delayed diagnoses.

Disrupted Structural White Matter Network in Alzheimer's Disease Continuum, Vascular Dementia, and Mixed Dementia: A Diffusion Tensor Imaging Study.

BACKGROUND: Dementia presents a significant burden to patients and healthcare systems worldwide. Early and accurate diagnosis, as well as differential diagnosis of various types of dementia, are crucial for timely intervention and management. However, there is currently a lack of clinical tools for accurately distinguishing between these types. OBJECTIVE: This study aimed to investigate the differences in the structural white matter (WM) network among different types of cognitive impairment/dementia using diffusion tensor imaging, and to explore the clinical relevance of the structural network. METHODS: A total of 21 normal control, 13 subjective cognitive decline (SCD), 40 mild cognitive impairment (MCI), 22 Alzheimer's disease (AD), 13 mixed dementia (MixD), and 17 vascular dementia (VaD) participants were recruited. Graph theory was utilized to construct the brain network. RESULTS: Our findings revealed a monotonic trend of disruption in the brain WM network (VaD > MixD > AD > MCI > SCD) in terms of decreased global efficiency, local efficiency, and average clustering coefficient, as well as increased characteristic path length. These network measurements were significantly associated with the clinical cognition index in each disease group separately. CONCLUSION: These findings suggest that structural WM network measurements can be utilized to differentiate between different types of cognitive impairment/dementia, and these measurements can provide valuable cognition-related information.

The Use of Diffusion Kurtosis Imaging for the Differential Diagnosis of Alzheimer's Disease Spectrum.

Structural and diffusion kurtosis imaging (DKI) can be used to assess hippocampal macrostructural and microstructural alterations respectively, in Alzheimer's disease (AD) spectrum, spanning from subjective cognitive decline (SCD) to mild cognitive impairment (MCI) and AD. In this study, we explored the diagnostic performance of structural imaging and DKI of the hippocampus in the AD spectrum. Eleven SCD, thirty-seven MCI, sixteen AD, and nineteen age- and sex-matched normal controls (NCs) were included. Bilateral hippocampal volume, mean diffusivity (MD), and mean kurtosis (MK) were obtained. We detected that in AD vs. NCs, the right hippocampal volume showed the most prominent AUC value (AUC = 0.977); in MCI vs. NCs, the right hippocampal MD was the most sensitive discriminator (AUC = 0.819); in SCD vs. NCs, the left hippocampal MK was the most sensitive biomarker (AUC = 0.775). These findings suggest that, in the predementia stage (SCD and MCI), hippocampal microstructural changes are predominant, and the best discriminators are microstructural measurements (left hippocampal MK for SCD and right hippocampal MD for MCI); while in the dementia stage (AD), hippocampal macrostructural alterations are superior, and the best indicator is the macrostructural index (right hippocampal volume).

MicroRNA-128 suppresses tau phosphorylation and reduces amyloid-beta accumulation by inhibiting the expression of GSK3ß, APPBP2, and mTOR in Alzheimer's disease.

INTRODUCTION AND AIMS: Alzheimer's disease (AD) is characterized by the abnormal accumulation of hyperphosphorylated tau proteins and amyloid-beta (Aß) peptides. Recent studies have shown that many microRNAs (miRNAs) are dysregulated in AD, and modulation of these miRNAs can influence the development of tau and Aß pathology. The brain-specific miRNA miR-128, encoded by MIR128-1 and MIR128-2, is important for brain development and dysregulated in AD. In this study, the role of miR-128 in tau and Aß pathology as well as the regulatory mechanism underlying its dysregulation were investigated. METHODS: The effect of miR-128 on tau phosphorylation and Aß accumulation was examined in AD cellular models through miR-128 overexpression and inhibition. The therapeutic potential of miR-128 in AD mouse model was assessed by comparing phenotypes of 5XFAD mice administered with miR-128-expressing AAVs with 5XFAD mice administered with control AAVs. Phenotypes examined included behavior, plaque load, and protein expression. The regulatory factor of miR-128 transcription was identified through luciferase reporter assay and validated by siRNA knockdown and ChIP analysis. RESULTS: Both gain-of-function and loss-of-function studies in AD cellular models reveal that miR-128 represses tau phosphorylation and Aß secretion. Subsequent investigations show that miR-128 directly inhibits the expression of tau phosphorylation kinase GSK3ß and Aß modulators APPBP2 and mTOR. Upregulation of miR-128 in the hippocampus of 5XFAD mice ameliorates learning and memory impairments, decreases plaque deposition, and enhances autophagic flux. We further demonstrated that C/EBPα transactivates MIR128-1 transcription, while both C/EBPα and miR-128 expression are inhibited by Aß. CONCLUSION: Our findings suggest that miR-128 suppresses AD pathogenesis, and could be a promising therapeutic target for AD. We also find a possible mechanism underlying the dysregulation of miR-128 in AD, in which Aß reduces miR-128 expression by inhibiting C/EBPα.

Level of Amyloid-ß (Aß) Binding Leading to Differential Effects on Resting State Functional Connectivity in Major Brain Networks.

INTRODUCTION: Amyloid-ß protein (Aß) is one of the biomarkers for Alzheimer's disease (AD). The recent application of interhemispheric functional connectivity (IFC) in resting-state fMRI has been used as a non-invasive diagnostic tool for early dementia. In this study, we focused on the level of Aß accumulated and its effects on the major functional networks, including default mode network (DMN), central executive network (CEN), salience network (SN), self-referential network (SRN) and sensory motor network (SMN). METHODS: 58 participants (27 Hi Aß (HiAmy) and 31 low Aß (LowAmy)) and 25 healthy controls (HC) were recruited. [18F]flutemetamol PET/CT was performed for diseased groups, and MRI scanning was done for all participants. Voxel-by-voxel correlation analysis was done for both groups in all networks. RESULTS: In HiAmy, IFC was reduced in all networks except SN. A negative correlation in DMN, CEN, SRN and SMN suggests high Aß related to IFC reduction; However, a positive correlation in SN suggests high Aß related to an increase in IFC. In LowAmy, IFC increased in CEN, SMN, SN and SRN. Positive correlation in all major brain networks. CONCLUSION: The level of Aß accumulated demonstrated differential effects on IFC in various brain networks. As the treatment to reduce Aß plaque deposition is available in the market, it may be an option for the HiAmy group to improve their IFC in major brain networks.

The fractional amplitude of low-frequency fluctuations signals related to amyloid uptake in high-risk populations-A pilot fMRI study.

Background: Patients with type 2 diabetes mellitus (T2DM) and subjective cognitive decline (SCD) have a higher risk to develop Alzheimer's Disease (AD). Resting-state-functional magnetic resonance imaging (rs-fMRI) was used to document neurological involvement in the two groups from the aspect of brain dysfunction. Accumulation of amyloid-ß (Aß) starts decades ago before the onset of clinical symptoms and may already have been associated with brain function in high-risk populations. However, this study aims to compare the patterns of fractional amplitude of low-frequency fluctuations (fALFF) maps between cognitively normal high-risk groups (SCD and T2DM) and healthy elderly and evaluate the association between regional amyloid deposition and local fALFF signals in certain cortical regions. Materials and methods: A total of 18 T2DM, 11 SCD, and 18 healthy elderlies were included in this study. The differences in the fALFF maps were compared between HC and high-risk groups. Regional amyloid deposition and local fALFF signals were obtained and further correlated in two high-risk groups. Results: Compared to HC, the altered fALFF signals of regions were shown in SCD such as the left posterior cerebellum, left putamen, and cingulate gyrus. The T2DM group illustrated altered neural activity in the superior temporal gyrus, supplementary motor area, and precentral gyrus. The correlation between fALFF signals and amyloid deposition was negative in the left anterior cingulate cortex for both groups. In the T2DM group, a positive correlation was shown in the right occipital lobe and left mesial temporal lobe. Conclusion: The altered fALFF signals were demonstrated in high-risk groups compared to HC. Very early amyloid deposition in SCD and T2DM groups was observed to affect the neural activity mainly involved in the default mode network (DMN).

Incremental diagnostic value of 18F-Fluetemetamol PET in differential diagnoses of Alzheimer's Disease-related neurodegenerative diseases from an unselected memory clinic cohort.

To evaluate the incremental diagnostic value of 18F-Flutemetamol PET following MRI measurements on an unselected prospective cohort collected from a memory clinic. A total of 84 participants was included in this study. A stepwise study design was performed including initial analysis (based on clinical assessments), interim analysis (revision of initial analysis post-MRI) and final analysis (revision of interim analysis post-18F-Flutemetamol PET). At each time of evaluation, every participant was categorized into SCD, MCI or dementia syndromal group and further into AD-related, non-AD related or non-specific type etiological subgroup. Post 18F-Flutemetamol PET, the significant changes were seen in the syndromal MCI group (57%, p < 0.001) involving the following etiological subgroups: AD-related MCI (57%, p < 0.01) and non-specific MCI (100%, p < 0.0001); and syndromal dementia group (61%, p < 0.0001) consisting of non-specific dementia subgroup (100%, p < 0.0001). In the binary regression model, amyloid status significantly influenced the diagnostic results of interim analysis (p < 0.01). 18F-Flutemetamol PET can have incremental value following MRI measurements, particularly reflected in the change of diagnosis of individuals with unclear etiology and AD-related-suspected patients due to the role in complementing AD-related pathological information.

Reduced Pneumonia Risk in Advanced Dementia Patients on Careful Hand Feeding Compared With Nasogastric Tube Feeding.

OBJECTIVES: To compare survival and pneumonia risk among hospitalized patients with advanced dementia on nasogastric tube feeding (NGF) vs careful hand feeding (CHF) and to examine outcomes by feeding problem type. DESIGN: Retrospective cohort study. SETTING AND PARTICIPANTS: Advanced dementia patients aged ≥60 years with indication for tube feeding admitted to 2 geriatric convalescent hospitals between January 1, 2015, and June 30, 2019. METHODS: Comparison on the effect of NGF and CHF on survival and pneumonia risk using Kaplan Meier survival analysis and Cox proportional hazards models. RESULTS: Of the 764 patients (mean age 89 years, 61% female, 74% residential care home residents), 464 (61%) were initiated on NGF and 300 (39%) on CHF. The primary feeding problem types were dysphagia (50%), behavioral feeding problem (33%), or both (17%). There was no difference in 1-year survival rate between NGF and CHF groups (36% vs 37%, P = .71) and survival did not differ by feeding problem type. Nasogastric tube feeding was not a significant predictor for survival (adjusted hazard ratio 1.15, 95% CI 0.94-1.39). Among 577 (76%) patients who survived to discharge, pneumonia rates were lower in the CHF group (48% vs 60%, P = .004). After adjusting for cofounders, NGF was a significant risk factor for pneumonia (adjusted hazard ratio 1.41, 95% CI 1.08-1.85). In subgroup analyses, NGF was associated with increased pneumonia risk for patients with both dysphagia and behavioral feeding problem (P = .01) but not in patients with behavioral feeding problem alone (P = .24) or dysphagia alone (P = .30). CONCLUSIONS AND IMPLICATIONS: For advanced dementia patients with feeding problems, there is no difference in survival between NGF and CHF. However, NGF is associated with a higher pneumonia risk, particularly for patients with both dysphagia and behavioral feeding problem. Further research on how the feeding problem type impacts pneumonia risk for patients on NGF is needed.

Prevalence of mild behavioural impairment domains: a meta-analysis.

BACKGROUND: Mild behavioural impairment (MBI) is a neurobehavioural syndrome characterised by later life emergence of persistent neuropsychiatric symptoms. Our previous meta-analysis showed that MBI is prevalent among cognitively normal (CN), subjective cognitive impairment (SCI) and mild cognitive impairment (MCI) subjects. This study is to calculate the pooled prevalence of MBI domains among CN, SCI, and MCI subjects. METHODS: A search of relevant literature published between 1 January 2003 and 6 August 2021 was conducted. Meta-analysis using a random effects model and meta-regression was performed. RESULTS: Ten studies conducted among 12 067 subjects (9758 CN, 1057 SCI and 1252 MCI) with retrievable MBI domains data underwent meta-analysis, revealing pooled prevalence of affective dysregulation (AFD), impulse dyscontrol (IDS), decreased motivation (DMT), social inappropriateness (SIP) and abnormal perception/thought (APT) of 32.84% (95% CI 24.44-42.5%), 26.67% (95% CI 18.24-37.23%), 12.58% (95% CI 6.93-21.75%), 6.05% (95% CI 3.44-10.42%), and 2.81% (95% CI 1.67-4.69%) respectively. AFD and APT domains demonstrated ordinal increase in pooled prevalence from CN, SCI and MCI subgroups, but meta-regression demonstrated no significant difference in MBI domains prevalence among cognitive subgroups (in contrast to the significant increase in MBI prevalence from CN to SCI to MCI). The pooled prevalence of AFD and IDS are greater than that of DMT, SIP and APT among all cognitive subgroups. Several variables were found to explain the high heterogeneity. CONCLUSIONS: AFD and IDS are the two most prevalent MBI domains and remain the same with cognitive deterioration. This finding is potentially relevant to clinical practice.

A systematic review and meta-analysis protocol examining the clinical characteristics and epidemiological features of olfactory dysfunction (OD) in coronavirus disease 2019 (COVID-19).

BACKGROUND: The coronavirus disease 2019 (COVID-19) pandemic has caused recurring and major outbreaks in multiple human populations around the world. The plethora of clinical presentations of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has been described extensively, of which olfactory dysfunction (OD) was established as an important and common extrapulmonary manifestation of COVID-19 infection. The aim of this protocol is to conduct a systematic review and meta-analysis on peer-reviewed articles which described clinical data of OD in COVID-19 patients. METHODS: This research protocol has been prospectively registered with the Prospective Register of Systematic Reviews (PROSPERO; CRD42020196202). CINAHL, ClinicalTrials.gov, Cochrane Central, EMBASE, MEDLINE and PubMed, as well as Chinese medical databases China National Knowledge Infrastructure (CNKI), VIP and WANFANG, will be searched using keywords including 'COVID-19', 'coronavirus disease', '2019-nCoV', 'SARS-CoV-2', 'novel coronavirus', 'anosmia', 'hyposmia', 'loss of smell', and 'olfactory dysfunction'. Systematic review and meta-analysis will be conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) and the Meta-analyses Of Observational Studies in Epidemiology (MOOSE) guidelines. Articles will be screened according to pre-specified inclusion and exclusion criteria to extract studies that include new clinical data investigating the effect of COVID-19 on olfactory dysfunction. Included articles will be reviewed in full; data including patient demographics, clinical characteristics of COVID-19-related OD, methods of olfactory assessment and relevant clinical outcomes will be extracted. Statistical analyses will be performed using the Comprehensive Meta-Analysis version 3. DISCUSSION: This systematic review and meta-analysis protocol will aim to collate and synthesise all available clinical evidence regarding COVID-19-related OD as an important neurosensory dysfunction of COVID-19 infection. A comprehensive search strategy and screening process will be conducted to incorporate broad clinical data for robust statistical analyses and representation. The outcome of the systematic review and meta-analysis will aim to improve our understanding of the symptomatology and clinical characteristics of COVID-19-related OD and identify knowledge gaps in its disease process, which will guide future research in this specific neurosensory defect. SYSTEMATIC REVIEW REGISTRATION: PROSPERO registration number: CRD42020196202.

A systematic review of atypical Alzheimer's disease including behavioural and psychological symptoms.

Alzheimer's disease (AD) is the commonest cause of dementia, characterized by the clinical presentation of progressive anterograde episodic memory impairment. However, atypical presentation of patients is increasingly recognized. These atypical AD include logopenic aphasia, behavioural variant AD, posterior cortical atrophy, and corticobasal syndrome. These atypical AD are more common in patients with young onset AD before the age of 65 years old. Since medical needs (including the behavioural and psychological symptoms of dementia) of atypical AD patients could be different from typical AD patients, it is important for clinicians to be aware of these atypical forms of AD. In addition, disease modifying treatment may be available in the future. This review aims at providing an update on various important subtypes of atypical AD including behavioural and psychological symptoms.

Heterogeneity of Amyloid Binding in Cognitively Impaired Patients Consecutively Recruited from a Memory Clinic: Evaluating the Utility of Quantitative 18F-Flutemetamol PET-CT in Discrimination of Mild Cognitive Impairment from Alzheimer's Disease and Other Dementias.

BACKGROUND: With the more widespread use of 18F-radioligand-based amyloid-ß (Aß) PET-CT imaging, we evaluated Aß binding and the utility of neocortical 18F-Flutemetamol standardized uptake value ratio (SUVR) as a biomarker. OBJECTIVE: 18F-Flutemetamol SUVR was used to differentiate 1) mild cognitive impairment (MCI) from Alzheimer's disease (AD), and 2) MCI from other non-AD dementias (OD). METHODS: 109 patients consecutively recruited from a University memory clinic underwent clinical evaluation, neuropsychological test, MRI and 18F-Flutemetamol PET-CT. The diagnosis was made by consensus of a panel consisting of 1 neuroradiologist and 2 geriatricians. The final cohort included 13 subjective cognitive decline (SCD), 22 AD, 39 MCI, and 35 OD. Quantitative analysis of 16 region-of-interests made by Cortex ID software (GE Healthcare). RESULTS: The global mean 18F-Flutemetamol SUVR in SCD, MCI, AD, and OD were 0.50 (SD-0.08), 0.53 (SD-0.16), 0.76 (SD-0.10), and 0.56 (SD-0.16), respectively, with SUVR in SCD and MCI and OD being significantly lower than AD. Aß binding in SCD, MCI, and OD was heterogeneous, being 23%, 38.5%, and 42.9% respectively, as compared to 100% amyloid positivity in AD. Using global SUVR, ROC analysis showed AUC of 0.868 and 0.588 in differentiating MCI from AD and MCI from OD respectively. CONCLUSION: 18F-Flutemetamol SUVR differentiated MCI from AD with high efficacy (high negative predictive value), but much lower efficacy from OD. The major benefit of the test was to differentiate cognitively impaired patients (either SCD, MCI, or OD) without AD-related-amyloid-pathology from AD in the clinical setting, which was under-emphasized in the current guidelines proposed by Amyloid Imaging Task Force.

Different glycaemia-related risk factors for incident Alzheimer's disease in men and women with type 2 diabetes-A sex-specific analysis of the Hong Kong diabetes database.

AIMS: Sexual dimorphism has been reported in the epidemiology, neurobiologic susceptibility and clinical presentation of Alzheimer's disease (AD). As poor glycaemic control is associated with increased risks of AD, we aimed to investigate whether glycaemia-related risk factors also differ between men and women, using a retrospective, sex-specific analysis of a large Chinese cohort with diabetes. MATERIALS & METHODS: A total of 85,514 Chinese individuals with type 2 diabetes (T2D; 46,783 women and 38,731 men), aged ≥60 years, were identified from electronic health records and observed for incident AD. Multivariable Cox regression analysis was used to evaluate the associations with incident AD of several glycaemia-related risk factors, including severe hypoglycaemia, mean HbA1c and indices of HbA1c variability, in men and women separately. RESULTS: Over a median follow-up of 6 years, women had a higher incidence of AD than men (2.3% vs. 1.2%, p < 0.001). Both men and women shared the same independent non-glycaemic clinical predictors, which included older age, lower body mass index and longer duration of diabetes. However, for glycaemia-related risk factors, we observed that severe hypoglycaemia and indices of HbA1c variability were independent predictors of incident AD in women but not in men, and the associations were irrespective of their baseline glycaemic control and duration of diabetes. CONCLUSIONS: Our findings highlighted that glycaemia-related risk factors for incident AD differ between men and women with T2D. Strategies to maintain glycaemic stability and avoid severe hypoglycaemia might be especially important to preserve healthy cognition in older women with diabetes.

Prevalence of mild behavioural impairment: a systematic review and meta-analysis.

AIM: Mild behavioural impairment (MBI) is a neurobehavioural syndrome characterized by emergent neuropsychiatric symptoms in later life. There has been no systematic review or meta-analysis on the prevalence of MBI. The main aim of the study is to calculate the pooled prevalence of MBI. METHODS: A search of the literature on MBI in mild cognitive impairment (MCI), cognitively normal (CN), and subjective cognitive impairment (SCI) and CN but at risk (CN-AR) subjects published between 1 January 2003 and 28 September 2020 was conducted. Meta-analysis using a random effects model was performed to determine the pooled estimate of the prevalence of MBI. Meta-regression was performed to identify factors contributing to the variance of prevalence rate. A systematic review was also performed to study the impact of MBI in cognitive outcomes and its correlation to the pathology and genetics of Alzheimer's disease. RESULTS: Eleven studies conducted among 15 689 subjects underwent meta-analysis, revealing the pooled prevalence of MBI to be 33.5% (95% confidence interval (CI): 22.6%-46.6%). Seven studies conducted among 1358 MCI subjects underwent meta-analysis, revealing the pooled prevalence to be 45.5% (95%CI: 36.1%-55.3%). Four studies conducted among 13 153 CN subjects underwent meta-analysis, revealing the pooled prevalence to be 17.0% (95%CI: 7.2%-34.9%). Five studies conducted among 1158 SCI or CN-AR subjects underwent meta-analysis, revealing the pooled prevalence to be 35.8% (95%CI: 21.4%-53.2%). A systematic review of 13 studies showed that MBI has a significant impact on cognitive deterioration and is associated with the pathology and genetics of Alzheimer's disease. CONCLUSIONS: In MCI, CN, and SCI and CN-AR subjects, MBI is common. Our finding is potentially useful in planning future clinical trials.