Multimodal ChatGPT-4V for Electrocardiogram Interpretation: Promise and Limitations.
Ontology highlight
ABSTRACT:
SUBMITTER: Zhu L
PROVIDER: S-EPMC11237788 | biostudies-literature | 2024 Jun
REPOSITORIES: biostudies-literature
Publications
Multimodal ChatGPT-4V for Electrocardiogram Interpretation: Promise and Limitations.
Zhu Lingxuan L Mou Weiming W Wu Keren K Lai Yancheng Y Lin Anqi A Yang Tao T Zhang Jian J Luo Peng P
Journal of medical Internet research 20240626
This study evaluated the capabilities of the newly released ChatGPT-4V, a large language model with visual recognition abilities, in interpreting electrocardiogram waveforms and answering related multiple-choice questions for assisting with cardiovascular care. ...[more]
Similar Datasets
Project description:OverviewThis study evaluates the diagnostic accuracy of a multimodal large language model (LLM), ChatGPT-4, in recognizing glaucoma using color fundus photographs (CFPs) with a benchmark dataset and without prior training or fine tuning.MethodsThe publicly accessible Retinal Fundus Glaucoma Challenge "REFUGE" dataset was utilized for analyses. The input data consisted of the entire 400 image testing set. The task involved classifying fundus images into either 'Likely Glaucomatous' or 'Likely Non-Glaucomatous'. We constructed a confusion matrix to visualize the results of predictions from ChatGPT-4, focusing on accuracy of binary classifications (glaucoma vs non-glaucoma).ResultsChatGPT-4 demonstrated an accuracy of 90% with a 95% confidence interval (CI) of 87.06%-92.94%. The sensitivity was found to be 50% (95% CI: 34.51%-65.49%), while the specificity was 94.44% (95% CI: 92.08%-96.81%). The precision was recorded at 50% (95% CI: 34.51%-65.49%), and the F1 Score was 0.50.ConclusionChatGPT-4 achieved relatively high diagnostic accuracy without prior fine tuning on CFPs. Considering the scarcity of data in specialized medical fields, including ophthalmology, the use of advanced AI techniques, such as LLMs, might require less data for training compared to other forms of AI with potential savings in time and financial resources. It may also pave the way for the development of innovative tools to support specialized medical care, particularly those dependent on multimodal data for diagnosis and follow-up, irrespective of resource constraints.
| S-EPMC11182172 | biostudies-literature
Project description: Not available
| S-EPMC11254196 | biostudies-literature
Project description:BackgroundAutomated computerized electrocardiogram (ECG) interpretation algorithms are designed to enhance physician ECG interpretation, minimize medical error, and expedite clinical workflow. However, the performance of current computer algorithms is notoriously inconsistent. We aimed to develop and validate an artificial intelligence-enabled ECG (AI-ECG) algorithm capable of comprehensive 12-lead ECG interpretation with accuracy comparable to practicing cardiologists.MethodsWe developed an AI-ECG algorithm using a convolutional neural network as a multilabel classifier capable of assessing 66 discrete, structured diagnostic ECG codes using the cardiologist's final annotation as the gold-standard interpretation. We included 2,499,522 ECGs from 720,978 patients ≥18 years of age with a standard 12-lead ECG obtained at the Mayo Clinic ECG laboratory between 1993 and 2017. The total sample was randomly divided into training (n = 1,749,654), validation (n = 249,951), and testing (n = 499,917) datasets with a similar distribution of codes. We compared the AI-ECG algorithm's performance to the cardiologist's interpretation in the testing dataset using receiver operating characteristic (ROC) and precision recall (PR) curves.ResultsThe model performed well for various rhythm, conduction, ischemia, waveform morphology, and secondary diagnoses codes with an area under the ROC curve of ≥0.98 for 62 of the 66 codes. PR metrics were used to assess model performance accounting for category imbalance and demonstrated a sensitivity ≥95% for all codes.ConclusionsAn AI-ECG algorithm demonstrates high diagnostic performance in comparison to reference cardiologist interpretation of a standard 12-lead ECG. The use of AI-ECG reading tools may permit scalability as ECG acquisition becomes more ubiquitous.
| S-EPMC8890098 | biostudies-literature
Project description:The widespread use of X- and gamma-rays in a range of sectors including healthcare, security and industrial screening is underpinned by the efficient detection of the ionising radiation. Such detector applications are dominated by indirect detectors in which a scintillating material is combined with a photodetector. Halide perovskites have recently emerged as an interesting class of semiconductors, showing enormous promise in optoelectronic applications including solar cells, light-emitting diodes and photodetectors. Here, we discuss how the same superior semiconducting properties that have catalysed their rapid development in these optoelectronic devices, including high photon attenuation and fast and efficient emission properties, also make them promising scintillator materials. By outlining the key mechanisms of their operation as scintillators, we show why reports of remarkable performance have already emerged, and describe how further learning from other optoelectronic devices will propel forward their applications as scintillators. Finally, we outline where these materials can make the greatest impact in detector applications by maximally exploiting their unique properties, leading to dramatic improvements in existing detection systems or introducing entirely new functionality.
| S-EPMC8444306 | biostudies-literature
Project description:BackgroundThe 12-lead electrocardiogram (ECG) remains a cornerstone of cardiac diagnostics, yet existing artificial intelligence (AI) solutions for automated interpretation often lack generalizability, remain closed-source, and are primarily trained using supervised learning, limiting their adaptability across diverse clinical settings. To address these challenges, we developed and compared two open-source foundational ECG models: DeepECG-SSL, a self-supervised learning model, and DeepECG-SL, a supervised learning model.MethodsBoth models were trained on over 1 million ECGs using a standardized preprocessing pipeline and automated free-text extraction from ECG reports to predict 77 cardiac conditions. DeepECG-SSL was pretrained using self-supervised contrastive learning and masked lead modeling. The models were evaluated on six multilingual private healthcare systems and four public datasets for ECG interpretation across 77 diagnostic categories. Fairness analyses assessed disparities in performance across age and sex groups, while also investigating fairness and resource utilization.ResultsDeepECG-SSL achieved AUROCs of 0.990 (95%CI 0.990, 0.990) on internal dataset, 0.981 (95%CI 0.981, 0.981) on external public datasets, and 0.983 (95%CI 0.983, 0.983) on external private datasets, while DeepECG-SL demonstrated AUROCs of 0.992 (95%CI 0.992, 0.992), 0.980 (95%CI 0.980, 0.980) and 0.983 (95%CI 0.983, 0.983) respectively. Fairness analyses revealed minimal disparities (true positive rate & false positive rate difference<0.010) across age and sex groups. Digital biomarker prediction (Long QT syndrome (LQTS) classification, 5-year atrial fibrillation prediction and left ventricular ejection fraction (LVEF) classification) with limited labeled data, DeepECG-SSL outperformed DeepECG-SL in predicting 5-year atrial fibrillation risk (N=132,050; AUROC 0.742 vs. 0.720; Δ=0.022; P<0.001), identifying reduced LVEF ≤40% (N=25,252; 0.928 vs. 0.900; Δ=0.028; P<0.001), and classifying LQTS syndrome subtypes (N=127; 0.931 vs. 0.853; Δ=0.078; P=0.026).ConclusionBy releasing model weights, preprocessing tools, and validation code, we aim to support robust, data-efficient AI diagnostics across diverse clinical environments. This study establishes self-supervised learning as a promising paradigm for ECG analysis, particularly in settings with limited annotated data, enhancing accessibility, generalizability, and fairness in AI-driven cardiac diagnostics.
| S-EPMC11908279 | biostudies-literature
Project description: Not available
| S-EPMC7891265 | biostudies-literature
Project description:The study of immune phenotypes in wild animals is beset by numerous methodological challenges, with assessment of detailed aspects of phenotype difficult to impossible. This constrains the ability of disease ecologists and ecoimmunologists to describe immune variation and evaluate hypotheses explaining said variation. The development of simple approaches that allow characterization of immune variation across many populations and species would be a significant advance. Here we explore whether serum protein concentrations and coarse-grained white blood cell profiles, immune quantities that can easily be assayed in many species, can predict, and therefore serve as proxies for, lymphocyte composition properties. We do this in rewilded laboratory mice, which combine the benefits of immune phenotyping of lab mice with the natural context and immune variation found in the wild. We find that easily assayed immune quantities are largely ineffective as predictors of lymphocyte composition, either on their own or with other covariates. Immunoglobulin G (IgG) concentration and neutrophil-lymphocyte ratio show the most promise as indicators of other immune traits, but their explanatory power is limited. Our results prescribe caution in inferring immune phenotypes beyond what is directly measured, but they do also highlight some potential paths forward for the development of proxy measures employable by ecoimmunologists.
| S-EPMC11264049 | biostudies-literature
Project description:BackgroundFrom a healthcare professional's perspective, the use of ChatGPT (Open AI), a large language model (LLM), offers huge potential as a practical and economic digital assistant. However, ChatGPT has not yet been evaluated for the interpretation of polysomnographic results in patients with suspected obstructive sleep apnea (OSA).Aims/objectivesTo evaluate the agreement of polysomnographic result interpretation between ChatGPT-4o and a board-certified sleep physician and to shed light into the role of ChatGPT-4o in the field of medical decision-making in sleep medicine.Material and methodsFor this proof-of-concept study, 40 comprehensive patient profiles were designed, which represent a broad and typical spectrum of cases, ensuring a balanced distribution of demographics and clinical characteristics. After various prompts were tested, one prompt was used for initial diagnosis of OSA and a further for patients with positive airway pressure (PAP) therapy intolerance. Each polysomnographic result was independently evaluated by ChatGPT-4o and a board-certified sleep physician. Diagnosis and therapy suggestions were analyzed for agreement.ResultsChatGPT-4o and the sleep physician showed 97% (29/30) concordance in the diagnosis of the simple cases. For the same cases the two assessment instances unveiled 100% (30/30) concordance regarding therapy suggestions. For cases with intolerance of treatment with positive airway pressure (PAP) ChatGPT-4o and the sleep physician revealed 70% (7/10) concordance in the diagnosis and 44% (22/50) concordance for therapy suggestions.Conclusion and significancePrecise prompting improves the output of ChatGPT-4o and provides sleep physician-like polysomnographic result interpretation. Although ChatGPT shows some shortcomings in offering treatment advice, our results provide evidence for AI assisted automation and economization of polysomnographic interpretation by LLMs. Further research should explore data protection issues and demonstrate reproducibility with real patient data on a larger scale.
| S-EPMC11890353 | biostudies-literature
Project description:AimsTraditional long-term ECG monitoring systems have primarily been used by cardiologist. New remote and wearable easy-to-use devices have led to increased use of ECG recordings also outside cardiology clinics. The aims of this study were to assess the feasibility and diagnostic accuracy of interpretation of the one-lead ECG recordings from a patch ECG device (ECG247 Smart Heart Sensor system) by general practitioners (GP).MethodsNorwegian GPs were invited to digitally assess 10 long-term ECG recordings with different arrhythmias performed by the ECG247 Smart Heart Sensor system. For all ECG examinations, the presence/absence of different arrhythmias was registered.ResultsA total of 40 GPs accepted the invitation and assessed all the 10 long-term ECG recordings. All the tests were assessed as interpretable by all the GPs. Arrhythmias (atrial fibrillation/flutter, supraventricular tachycardia, and ventricular tachycardia) were correctly identified in most cases, with sensitivity of 98% (95% CI 95-99%), specificity of 75% (95% CI 68-82%), and diagnostic accuracy of 89% (85-92%). Incorrect automatic system algorithm interpretations were rarely corrected by the GPs.ConclusionGPs interpreted one-lead recordings by the ECG247 Smart Heart Sensor system with high diagnostic accuracy for common arrhythmias. However, in cases with rare arrhythmias, we recommend consulting a cardiologist to confirm the diagnosis before treatment is initiated. This trial is registered with NCT04700865.
| S-EPMC10904176 | biostudies-literature
Project description:Oxidative stress is a component of many diseases, including atherosclerosis, chronic obstructive pulmonary disease, Alzheimer disease and cancer. Although numerous small molecules evaluated as antioxidants have exhibited therapeutic potential in preclinical studies, clinical trial results have been disappointing. A greater understanding of the mechanisms through which antioxidants act and where and when they are effective may provide a rational approach that leads to greater pharmacological success. Here, we review the relationships between oxidative stress, redox signalling and disease, the mechanisms through which oxidative stress can contribute to pathology, how antioxidant defences work, what limits their effectiveness and how antioxidant defences can be increased through physiological signalling, dietary components and potential pharmaceutical intervention.
| S-EPMC8243062 | biostudies-literature