Master's thesis work aims to develop a tool for the analysis and prediction of data from the MIMIC-III database, using sepsis as a case of study.

This project aims to predict sepsis in patients using advanced machine learning models. The workflow encompasses data preprocessing, feature engineering, class imbalance handling, hyperparameter optimization, model training, evaluation, model card generation, and model registry management for reproducibility and scalability.

Single-file PyTorch pipeline for pathogen class prediction on MIMIC-III/IV EHR data. Hybrid Conv1D+BiLSTM+numeric model, streaming CSV ETL without pandas, ECE calibration, ROC/PR-AUC, and subgroup bias checks. Research and education only.

This repository houses a machine learning project focused on the early detection and classification of sepsis, and integrating the model into a web application using FAST API.

Winter 2024 - DSC180B A05-01 GitHub Repository (Click the link to go to the Dashboard)

Machine learning models for early sepsis prediction using EMR data

This repository contains the implementation and evaluation of multiple machine learning models using Jupyter Notebook. A total of 13 models have been tested, with Model 13 achieving the highest accuracy using ensemble methods like RandomForestClassifier and StackingClassifier.

Early sepsis risk prediction pipeline using machine learning on ICU clinical data.

PoC for sepsis screening in ER

Silent guardian. Background monitoring for critical patient conditions. Saves lives quietly.

It analyze clinical data and detect early signs of sepsis, enabling timely treatment and improving patient outcomes.

This work contains an algorithm for predicting neonatal sepsis using EMR data from Mbarara Regional Referral Hospital (MRRH). The proposed algorithm implements SVM, LR, KNN, NB, and DT.

Deep Learning for ICU Clinical Time-Series Anomaly Detection & Early Warning — sepsis prediction, mortality, and alarm fatigue reduction on MIMIC-III

Sepsis prediction ML platform — XGBoost model trained on 20M ICU records using Azure ML, Fabric, and Power BI

A Machine learning approach to predict whether a patient is likely to develop sepsis within the next six hours.

Data Science Project ( DSP ) : Develop a model to predict sepsis at an early stage, before clinical diagnosis. ( Early Prediction of Sepsis )

ML Modeling of PICU Data for Early Sepsis Detection

AI Clinical Intelligence Platform — Medical Scribe, Clinical RAG, Sepsis Prediction & Readmission Risk for Hospitals