Abstract:Introduction: Laboratory value represents a cornerstone of medical diagnostics, but suffers from slow turnaround times, and high costs and only provides information about a single point in time. The continuous estimation of laboratory values from non-invasive data such as electrocardiogram (ECG) would therefore mark a significant frontier in healthcare monitoring. Despite its transformative potential, this domain remains relatively underexplored within the medical community. Methods: In this preliminary study, we used a publicly available dataset (MIMIC-IV-ECG) to investigate the feasibility of inferring laboratory values from ECG features and patient demographics using tree-based models (XGBoost). We define the prediction task as a binary prediction problem of predicting whether the lab value falls into low or high abnormalities. The model performance can then be assessed using AUROC. Results: Our findings demonstrate promising results in the estimation of laboratory values related to different organ systems based on a small yet comprehensive set of features. While further research and validation are warranted to fully assess the clinical utility and generalizability of ECG-based estimation in healthcare monitoring, our findings lay the groundwork for future investigations into approaches to laboratory value estimation using ECG data. Such advancements hold promise for revolutionizing predictive healthcare applications, offering faster, non-invasive, and more affordable means of patient monitoring.

What problem does this paper attempt to address?

### What problem does this paper attempt to solve? This paper aims to explore the feasibility of using electrocardiogram (ECG) features and patient demographic data to estimate laboratory index values. Specifically, the researchers hope to extract features from electrocardiograms in a non - invasive way and combine other easily obtainable data (such as age, gender, etc.) to predict whether the patient's laboratory indices are abnormal. #### Main problem background: 1. **Limitations of laboratory tests**: - Laboratory tests are crucial in medical diagnosis, but they have problems such as long turnaround times, high costs, and can only provide information at a certain point in time. - Frequent invasive operations such as blood sampling are not only resource - intensive but may also lead to delays or a lack of laboratory personnel at night, thus limiting the ability of real - time monitoring. 2. **Need for continuous monitoring**: - Continuous monitoring of laboratory indices can significantly improve the effectiveness of diagnosis and treatment, especially for patients in the intensive care unit (ICU). - Non - invasive continuous monitoring means (such as electrocardiogram) can provide a faster, more economical and more convenient monitoring method. 3. **Deficiencies in existing research**: - Although previous studies have shown an association between electrocardiogram features and some laboratory indices, there are still relatively few studies on accurately estimating multiple laboratory indices using electrocardiogram data. #### Research objectives: - **Verify feasibility**: By using the public data set (MIMIC - IV - ECG), the researchers hope to verify the feasibility of inferring laboratory index values from electrocardiogram features and patient demographic data. - **Model performance evaluation**: Use a tree - based model (XGBoost) for binary classification prediction tasks and evaluate the performance of the model on laboratory indices related to different organ systems. - **Potential for clinical application**: Explore the potential application value of this electrocardiogram - based estimation method in actual medical treatment and lay the foundation for future in - depth research. #### Method overview: - **Data sources**: Use the MIMIC - IV and MIMIC - IV - ECG data sets, which include electrocardiogram features, demographic information, and basic vital signs. - **Model selection**: Use the XGBoost model for binary classification prediction, and define an abnormal situation as a laboratory index lower or higher than the patient - specific median low or high threshold. - **Performance evaluation**: Evaluate the model performance through the area under the receiver operating characteristic curve (AUROC) and its 95% confidence interval. #### Results and significance: - **Preliminary results**: The study shows that this method performs well in predicting multiple laboratory indices related to different organ systems and has a high AUROC value. - **Clinical significance**: This method is expected to achieve faster, more economical and non - invasive patient monitoring, especially in resource - limited environments, which helps to improve diagnostic accuracy and optimize the clinical work flow. In summary, the main purpose of this paper is to explore a new non - invasive method, based on electrocardiogram features and combined with other easily obtainable data, to continuously estimate laboratory index values, thereby improving the effectiveness and efficiency of patient monitoring and management.

CardioLab: Laboratory Values Estimation from Electrocardiogram Features -- An Exploratory Study

CardioLab: Laboratory Values Estimation and Monitoring from Electrocardiogram Signals -- A Multimodal Deep Learning Approach

Estimation of Cardiac and Non-cardiac Diagnosis from Electrocardiogram Features

Automated and Interpretable Patient ECG Profiles for Disease Detection, Tracking, and Discovery

Electrocardiogram-based diagnosis of liver diseases: an externally validated and explainable machine learning approach

Prospects for artificial intelligence-enhanced electrocardiogram as a unified screening tool for cardiac and non-cardiac conditions: an explorative study in emergency care

Benchmarking mortality risk prediction from electrocardiograms

Development and validation of machine learning algorithms based on electrocardiograms for cardiovascular diagnoses at the population level

ECG-Based Electrolyte Prediction: Evaluating Regression and Probabilistic Methods

Prospects for AI-Enhanced ECG as a Unified Screening Tool for Cardiac and Non-Cardiac Conditions -- An Explorative Study in Emergency Care

Interpretable estimation of the risk of heart failure hospitalization from a 30-second electrocardiogram

Predicting Abnormalities in Laboratory Values of Patients in the Intensive Care Unit Using Different Deep Learning Models: Comparative Study

Exploring a new frontier in cardiac diagnosis: ECG analysis enhanced by machine learning and parametric quartic spline modeling

An Electrocardiogram Foundation Model Built on over 10 Million Recordings with External Evaluation across Multiple Domains

Estimating Critical Values from Electrocardiogram Using a Deep Ordinal Convolutional Neural Network

Decoding 2.3 million ECGs: interpretable deep learning for advancing cardiovascular diagnosis and mortality risk stratification

Towards artificial intelligence-based learning health system for population-level mortality prediction using electrocardiograms

Automated Cardiovascular Record Retrieval by Multimodal Learning between Electrocardiogram and Clinical Report

ECG-CL: A Comprehensive Electrocardiogram Interpretation Method Based on Continual Learning

Estimating ECG Intervals from Lead-I Alone: External Validation of Supervised Models

Predicting Ejection Fraction from Electrocardiogram Signals using a Multi-task Learning Model