Bigbird-base-mimic-mortality Open-source Medical Text Analysis Model: Accurately Predict ICU Patients' Mortality

Bigbird Base Mimic Mortality

Developed by mnaylor

A medical text analysis model fine-tuned from Google's BigBird foundation model, specifically designed for predicting ICU patient mortality.

Downloads 67

Release Time : 3/2/2022

Model Overview

This model is fine-tuned on MIMIC admission notes to predict ICU mortality risk by analyzing patients' textual admission information.

Long-text Processing Capability

Optimized for handling long-sequence medical texts (e.g., admission notes) based on BigBird architecture

Clinical Outcome Prediction

Specialized in extracting key prognostic indicators from unstructured medical texts

Transfer Learning Application

Adapted for specific medical prediction tasks through fine-tuning pre-trained models

Medical Text Analysis

Clinical Risk Prediction

Long-document Understanding

Healthcare

ICU Mortality Prediction

Predicts patient mortality risk during ICU stay based on admission notes

Validated effective on MIMIC dataset (specific metrics not provided)

Clinical Decision Support

Provides early risk warning references for physicians

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base