AutoScore

"AutoScore" is a machine learning-based automatic clinical score generator to aid in clinical risk stratification and the efficient allocation of medical resources. Recognizing the significance of point-based scores in clinical decision-making due to their understandability and explainability, AutoScore simplifies the development of these models, which has traditionally been a complex and underexplored process.

Key Features and Functionalities:

- Six-Module Framework: AutoScore is built around a comprehensive framework comprising six modules: variable ranking, variable transformation, score derivation, model selection, score fine-tuning, and model evaluation. This structured approach facilitates the systematic development of interpretable point-based scores.

- Machine Learning-Based Score Generation: Utilizing advanced machine learning techniques, AutoScore automates creating clinical scores, making them accessible for future users across various clinical applications without requiring in-depth statistical knowledge.

Potential for Wide Application in Medical Fields: AutoScore's ability to create interpretable, point-based scoring models efficiently holds significant potential as a scoring tool in various medical applications, from risk stratification to resource allocation.

Topic

Machine learning;Preclinical and clinical studies;Surgery

Detail

  • Operation: Regression analysis;Feature selection;Standardisation and normalisation

  • Software interface: Library

  • Language: R

  • License: GNU General Public License >= version 2

  • Cost: Free with restrictions

  • Version name: 1.0.0

  • Credit: The Singapore National Medical Research Council under the PULSES Center Grant.

  • Input: -

  • Output: -

  • Contact: Nan Liu liu.nan@duke-nus.edu.sg

  • Collection: -

  • Maturity: -

Publications

  • AutoScore: A Machine Learning-Based Automatic Clinical Score Generator and Its Application to Mortality Prediction Using Electronic Health Records.
  • Xie F, et al. AutoScore: A Machine Learning-Based Automatic Clinical Score Generator and Its Application to Mortality Prediction Using Electronic Health Records. AutoScore: A Machine Learning-Based Automatic Clinical Score Generator and Its Application to Mortality Prediction Using Electronic Health Records. 2020; 8:e21798. doi: 10.2196/21798
  • https://doi.org/10.2196/21798
  • PMID: 33084589
  • PMC: PMC7641783

Download and documentation


< Back to DB search