This repository contains the code used for Temporal Pointwise Convolutional Networks for Length of Stay Prediction in the Intensive Care Unit (https://dl.acm.org/doi/10.1145/3450439.3451860).

A Multimodal Transformer: Fusing Clinical Notes With Structured EHR Data for Interpretable In-Hospital Mortality Prediction

Implementation of Deep Patient Representation of Clinical Notes at Intensive Care Unit for Multi-Task Prediction

Time-Sensitive Deep Learning for ICU Outcome Prediction Without Variable Selection or Cleaning.

FDSML Course Project 2020/21

Real-time AI model for predicting emergency room mortality in trauma patients using only prehospital data.

Standard tools to compare and evaluate mortality forecasting methods

A Machine Learning Approach for Predicting the Death Time and Mortality

Repo of the Fleming project (dynamic prediction of mortality threat)

Continuous patient state attention model for addressing irregularity in electronic health records

Code and Datasets for the paper "An Interpretable Risk Prediction Model for Healthcare with Pattern Attention", published on BMC Medical Informatics and Decision Making.

Transformer model for Biomarkers prediction: Evaluating the impact of ECDF normalization on model robustness in clinical data

COVID-19 Critical Event Prediction Tools

A RNN Predictive Model for COVID-19 mortality prediction.

Predicting Mortality among a Cohort of patients with Heart Failure

This is implementation of my Masters thesis work and aims at predicting the patient mortality from the MIMIC-III v1.4 Database

Code and Description of AI-based encoding of ventricular action potential morphology changes to predict mortality (all-cause and arrhythmic) in patients with sudden death.

Stata PRISM Score Calculation

本專案基於 MIMIC-III C Clinical Database (MIMIC3C)，系統性地完成醫療大數據的預測建模流程。專案分為兩大階段： 階段一(資料工程)： 針對高維度、多缺失的真實醫療資料進行品質檢查、特徵工程與前處理，產出可建模之資料表（Analysis-Ready Table）。 階段二(模型實驗)： 針對「分層抽樣（Stratified Sampling）」對模型表現的影響進行實證研究，橫跨 7 種機器學習模型（含隨機森林、神經網路等）與不同資料規模，驗證其在非平衡醫療資料下的統計顯著性。