edoc

Use of decrease in apparent diffusion coefficient values to predict infarct age

Provenzale, James M. and Stinnett, Sandra S. and Engelter, Stefan T.. (2010) Use of decrease in apparent diffusion coefficient values to predict infarct age. Emergency radiology, Vol. 17, H. 5. pp. 391-395.

Full text not available from this repository.

Official URL: http://edoc.unibas.ch/dok/A6004452

Downloads: Statistics Overview

Abstract

The purpose of the study was to determine the ability to predict infarct age based on decrease in apparent diffusion coefficient (ADC) values. We retrospectively identified 94 individuals (age range 16 years to 91 years; mean 63.7 + 14.1 years) who underwent magnetic resonance diffusion-weighted imaging at our institution over a course of 14 months whose infarct age could be reliably determined within 1 h. A single observer measured ADC values within the center of the infarct and compared them against values in contralateral normal tissue. We examined the ability of four ADC decrease thresholds (i.e., <50%, <40%, <30%, and <20%) to predict infarct age of >24 and >48 h. Levels of ADC decrease in infarcts were as follows: >20%, n = 9; 20-29%, n = 21; 30-39%, n = 25; 40-49%, n = 23; <50%, n = 16. For prediction of infarct age >24 h, sensitivity for the four ADC decrease thresholds ranged from 25% to 94%, specificity ranged from 10% to 85%, positive predictive value (PPV) ranged from 18% to 25%, and negative predictive value (NPV) ranged from 85% to 89%. For prediction of infarct age >48 h, sensitivity ranged from 23% to 98%, specificity ranged from 15% to 87%, PPV ranged from 46% to 56%, and NPV ranged from 60% to 89%. Test performance characteristics for predicting infarct age of >24 and >48 h were relatively poor. In particular, PPV was very low for predicting infarcts >24 h old.
Faculties and Departments:03 Faculty of Medicine > Bereich Medizinische Fächer (Klinik) > Neurologie
03 Faculty of Medicine > Departement Klinische Forschung > Bereich Medizinische Fächer (Klinik) > Neurologie
UniBasel Contributors:Engelter, Stefan
Item Type:Article, refereed
Article Subtype:Research Article
Publisher:Springer
ISSN:1070-3004
Note:Publication type according to Uni Basel Research Database: Journal article
Related URLs:
Identification Number:
Last Modified:07 Dec 2012 13:04
Deposited On:07 Dec 2012 13:02

Repository Staff Only: item control page