1 / 6

Aim :

Comparison of Modelling Technique to Predict Clinical Outcome using Routinely Collected Data. Aim : Develop prediction model that can be used to facilitate clinicians in targeting patients at high or low risk of mortality. Method : Logistic Regression Decision Tree Clustering KMeans

zayit
Download Presentation

Aim :

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Comparison of Modelling Technique to Predict Clinical Outcome using Routinely Collected Data • Aim : • Develop prediction model that can be used to facilitate clinicians in targeting patients at high or low risk of mortality. • Method : • Logistic Regression • Decision Tree • Clustering KMeans • Neural Network • Hybrid => Neuro Fuzzy, Fuzzy Subtractive Clustering, etc

  2. BHOM Dataset • Model was built from BHOM dataset, during 12-month study period => 17,417 patients, quarters 1, 2, 3 and 4 (q1,q2,q3,q4). • q1 as data training (n1=2257), q2, q3, and q4 are data testing (n2=2335, n3=2361, n4=2544) • The fields are : • death - at discharge  - F=alive, T =dead (class attribute) • age at admission • mode of admission (mostly emergency, but some elective) • gender • haemoglobin • white cell count • urea • serum sodium • serum potassium • creatinine • urea / creatinine

  3. Predicted Risk of Mortality for quarter q2,q3 and q4using Logistic Regression Q2 Q3 Q4

  4. Predicted Risk of Mortality for quarter q2,q3 and q4using Inducer Classifier Q2 Q3 Q4

  5. Predicted Risk of Mortality for quarter q2,q3 and q4using KMeans Clustering Q2 Q3 Q4

  6. The Result and Work Plan • Logistic regression is an acceptable method by assessing model performance using techniques designed to test both calibration and discrimination • This involved use of the χ2 test to compare frequency tables • using the c-index (equivalent to the area under the ROC curve). • Investigating methods at Model Assessment for Inducer Classifier and KMeansClustering. • Investigating How to Compare the Different Methods

More Related