Digital Twin Healthcare Dataset - EHR, Imaging, and IoT Integration

Dataset Overview

This groundbreaking Digital Twin Healthcare dataset represents the convergence of multiple medical data sources, published on Kaggle in April 2025. It enables research into holistic patient digital twins combining clinical records, diagnostic imaging, and continuous IoT monitoring for precision medicine applications.

Three-Pillar Data Architecture

1. Electronic Health Records (EHRs)

Comprehensive patient information including demographics, medical history, diagnoses, treatment plans, medication records, and clinical outcomes. Structured data enables longitudinal patient journey analysis.

2. Medical Imaging - CT and MRI

Extensive collection of 5,008 brain scan files including:

• CT Scans: Computed Tomography images for rapid diagnosis of acute conditions
• MRI Scans: Magnetic Resonance Imaging for detailed soft tissue visualization
• Cross-Modality Analysis: Both imaging types for comprehensive diagnostic assessment

3. Wearable IoT Sensor Data

Real-time physiological monitoring from connected devices tracking:

• Heart Rate: Continuous cardiac monitoring with 58 columns of time-series data
• Activity Levels: Physical activity patterns and mobility metrics
• Mental Health Indicators: Stress levels, sleep quality, and behavioral patterns from wearables

Integrated Diabetes Analysis

The dataset includes UCI diabetes data with 20 records and 16 binary attributes (0 or 1) representing presence or absence of diabetes risk factors, enabling correlation studies between metabolic conditions and imaging/sensor data.

Dataset Scale

• Total Size: 196.89 MB
• Imaging Files: 5,008 brain scans
• Tabular Data: Multiple CSV files (EHR.csv, UCI_Diabetes.csv, mental_health_wearable_data.csv)
• Features: 58 columns of IoT sensor measurements

Research Applications

Digital Twin Development

Create comprehensive virtual patient models combining static medical records with dynamic real-time sensor feeds and periodic imaging updates for predictive healthcare.

Personalized Treatment Optimization

Leverage multi-modal data to tailor treatments based on individual patient responses captured through continuous monitoring and imaging assessments.

Multi-Modal Machine Learning

• Fusion Models: Combine imaging features (CNN), EHR data (structured ML), and IoT time-series (LSTM) for enhanced predictions
• Transfer Learning: Apply pre-trained imaging models with patient-specific sensor data fine-tuning
• Ensemble Methods: Integrate predictions from multiple data sources for robust clinical decision support

Neurological Research

Correlate brain imaging patterns with wearable sensor data to study relationships between structural brain changes and functional impairments in cognitive or motor activities.

Chronic Disease Management

Monitor diabetes patients using integrated EHR, continuous glucose data from IoT devices, and imaging to assess complications like neuropathy or cardiovascular changes.

Innovation in Healthcare AI

This dataset enables next-generation healthcare AI systems that move beyond single-source analysis to holistic patient assessment, mirroring how human physicians integrate diverse information sources for clinical decision-making.