Related Experiment Video
Updated: Jul 9, 2025
Genetic Analysis of Hereditary Transthyretin Ala97Ser Related Amyloidosis
Published on: June 9, 2018
EstimATTR: A Simplified, Machine-Learning-Based Tool to Predict the Risk of Wild-Type Transthyretin Amyloid
Adam Castaño1, Stephen B Heitner2, Ahmad Masri2
1Pfizer Inc, New York, New York.
A new machine learning model accurately predicts wild-type transthyretin amyloid cardiomyopathy (ATTRwt-CM) risk in heart failure patients. This tool aids early diagnosis and clinical assessment for this often-overlooked condition.
Area of Science:
- Cardiology
- Medical Informatics
- Machine Learning
Background:
- Wild-type transthyretin amyloid cardiomyopathy (ATTRwt-CM) is an underdiagnosed cause of heart failure.
- Early detection is crucial as ATTRwt-CM can progress rapidly without treatment.
Purpose of the Study:
- To simplify a machine learning algorithm for predicting ATTRwt-CM risk.
- To develop an easily implementable tool for clinical risk assessment.
Main Methods:
- A random forest model was developed using 11 phenotypes predictive of ATTRwt-CM.
- The model was trained and validated on U.S. medical claims and electronic health records.
Main Results:
- The simplified model achieved 74% accuracy, 77% sensitivity, and 72% specificity in identifying ATTRwt-CM.
- A robust performance was observed in validation cohorts, with an AUC of 0.82.
Conclusions:
- The machine learning model accurately estimates ATTRwt-CM probability in administrative datasets.
- This tool can facilitate earlier clinical assessment and diagnosis of ATTRwt-CM.
More Related Videos
16:02Demonstration of the Sequence Alignment to Predict Across Species Susceptibility Tool for Rapid Assessment of Protein Conservation
Published on: February 10, 2023
04:09Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma
Published on: October 10, 2018