3M Pipeline Project

This project contains various notebooks exploring different aspects of the MIMIC-IV database to demonstrate data science modeling capabilities.

In a broader perspective, this pipeline could facilitate advanced monitoring/alerting systems, extract insights for clinical decision making, and enable predictive modeling for patient outcomes.

Pipeline Overview

• Vitals Analysis: Time series analysis and feature engineering of patient vital signs
• Medications Analysis: Clustering and pattern mining of medication administration data (todo)
• Labs Analysis: Predictive modeling using laboratory results (todo)
• Clinical Notes: NLP and text mining of clinical documentation (todo)
• Procedures: Sequential pattern mining of medical procedures (todo)
• Outcomes Analysis: Survival analysis and risk modeling (todo)

Vitals Analysis Notebook (00_vitals_analysis.ipynb)

Purpose

Analysis of patient vital signs data to identify patterns and relationships between measurements through feature engineering and statistical analysis.

Key Components

• Temporal features (hour, day, etc.)
• Vital sign ratios (shock index, MAP, etc.)
• Statistical metrics (mean, std, min/max)
• Variability measures (changes between measurements)

Models & Analysis Methods

• K-means clustering for patient grouping
• Time series decomposition
• Principal Component Analysis (PCA) for dimension reduction
• Correlation analysis and heatmapping
• Distribution analysis with KDE
• Pattern detection using statistical tests
• Anomaly detection with Isolation Forest