Predictive Analytics
Liver Failure
Model Metrics
100

This machine learning model, used in the study by Sheng et al. (2021), predicts complications like AHE and HRS.

What is the RNN-LR model?

100

This type of liver failure happens suddenly, often from infections or toxins.

What is Acute Liver Failure (ALF)?

100

This word means how well the model finds the right answers.

What is performance?

200

This metric indicates the accuracy of predictive models and had a score of 0.82 for the RNN-LR model.

What is the Area Under the Curve (AUC)?

200

This liver failure happens in patients with a long-term liver problem.

What is Acute-on-Chronic Liver Failure (ACLF)?

200

This word shows how many of the true problems the model finds.

What is recall?

300

The RNN-LR model uses patient data like this to predict problems.  

What are bilirubin, creatinine, and lactate levels?

300

Doctors use this score to check how well the liver and kidneys are working.

What is the SOFA score?

300

This model got a higher score than the others, showing it’s much better.

What is the RNN-LR?

400

The RNN-LR model was trained using patient records from this database.

What is the Chang Gung Research Database (CGRD)?

400

This complication in liver patients causes kidney problems.

What is Hepatorenal Syndrome (HRS)?

400

AUC scores close to this number mean the model is very good.

What is 1?

500

The predictions from the model help create this kind of special care plan.

What is a personalized care plan?

500

This complication affects the brain and is caused by toxins.

What is Acute Hepatic Encephalopathy (AHE)?

500

This word explains how well the model finds both correct and incorrect problems.

What is precision?