Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

On prognostic models, artificial intelligence and censored observations.

S S Anand1, P W Hamilton, J G Hughes

  • 1School of Information and Software Engineering, University of Ulster at Jordanstown, Northern Ireland. ss.anand@ulst.ac.uk

Methods of Information in Medicine
|April 20, 2001
PubMed
Summary

Developing new prognostic models requires an evaluation-centric approach. Censored k-nearest neighbour (Ck-NN) offers enhancements for handling censored data, improving model utility in clinical practice.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

A global initiative to deliver precision health in diabetes.

Nature medicine·2024
Same author

Implementing simple algorithms to improve glucose and lipid management in people with diabetes and acute coronary syndrome.

Diabetic medicine : a journal of the British Diabetic Association·2019
Same author

The influence of maternal and infant nutrition on cardiometabolic traits: novel findings and future research directions from four Canadian birth cohort studies.

The Proceedings of the Nutrition Society·2019
Same author

Patients with type 1 diabetes in a tertiary setting do not attain recommended lipid targets.

Diabetes & metabolism·2019
Same author

Anti-thrombotic options for secondary prevention in patients with chronic atherosclerotic vascular disease: what does COMPASS add?

European heart journal·2018
Same author

FIBRODYSPLASIA OSSIFICANS PROGRESSIVA: A Case Report.

Medical journal, Armed Forces India·2017

Area of Science:

  • Medical Informatics
  • Machine Learning
  • Statistics

Background:

  • Prognostic model development for clinical decision support faces challenges in meeting desirable characteristics.
  • Current statistical and artificial intelligence models often lack practical utility, limiting real-world deployment.
  • Academic interest in modeling techniques often outweighs practical application due to inadequate evaluation.

Purpose of the Study:

  • To introduce an evaluation-centric approach for developing advanced prognostic models.
  • To enhance the k-nearest neighbour (k-NN) algorithm for improved clinical decision-making.
  • To address the limitations of existing models in handling censored observations.

Main Methods:

  • Developed extensions to the basic k-nearest neighbour (k-NN) paradigm.

Related Experiment Videos

  • Employed standard statistical techniques to refine the distance metric.
  • Utilized a framework based on evidence theory for prediction from retrieved exemplars.
  • Introduced the Censored k-NN (Ck-NN) algorithm.

    • Main Results:

    • The Censored k-NN (Ck-NN) algorithm effectively handles censored observations within the k-NN framework.
    • Enhanced distance metrics and evidence theory improve prediction accuracy.
    • The evaluation-centric approach facilitates the development of more practical prognostic models.

    Conclusions:

    • The proposed Censored k-NN (Ck-NN) algorithm offers a significant advancement over traditional k-NN methods.
    • An evaluation-centric development strategy is crucial for creating clinically useful prognostic models.
    • Ck-NN demonstrates potential for wider adoption in medical practice due to its ability to handle censored data.