Communication FZJ-2022-02472

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png
NIMG-17. DISCRIMINATION BETWEEN RADIATION INJURY AND BRAIN METASTASIS RECURRENCE BASED ON TEXTURAL FEATURE ANALYSIS OF FET PET – SUPERIOR TO STANDARD METHODS?

 ;  ;  ;  ;  ;  ;  ;

2016

Abstract: BACKGROUND:Since the differentiation of radiation injury and tumor recurrence using standard MRI alone is difficult, we investigated the potential of textural parameters of O-(2-[18F]fluoroethyl)-L-tyrosine (FET) PET for this issue.METHODS:Forty-seven patients (mean age, 55 ± 11 y) with single or multiple contrast-enhancing brain lesions (n=54) on MRI after radiotherapy of brain metastases (predominantly stereotactic radiosurgery) underwent FET PET in an ECAT EXACT HR+ scanner. VOIs were defined on summed images 20-40 min post-injection by a 3-dimensional auto-contouring process using a tumor-to-brain ratio (TBR) of 1.6. For each lesion, TBRs and time-activity curves of FET uptake and 59 textural parameters were determined using the CGITA toolbox (Fang et al., 2014 Biomed Res Int). Diagnostic accuracy of TBRs, kinetic patterns, time-to-peak, textural parameters and combinations thereof to discriminate between brain metastasis recurrence and radiation injury was evaluated by ROC analyses using histological results (21 lesions, 20 patients) or clinical course (33 lesions, 27 patients) as reference.RESULTS:The diagnostic accuracy increased from 81% for TBRmean alone to 85% when combined with the textural parameter Coarseness (sensitivity, 73%; specificity, 100%) or Short-zone emphasis (sensitivity, 77%; specificity, 96%). The accuracy of TBRmax alone was 83% and increased to 85% after combination with the textural parameters Coarseness (sensitivity, 73%; specificity, 100%), Short-zone emphasis (sensitivity, 77%; specificity, 96%) or Correlation (sensitivity, 77%; specificity, 96%). The analysis of the time-activity curves resulted in a diagnostic accuracy of 70% for the kinetic pattern alone and increased to 83% when combined with TBRmax. The accuracy of the time-to-peak was 67% and could not be further improved by combination with other parameters.CONCLUSIONS:Textural analysis in combination with TBRs may have the potential to increase the diagnostic accuracy for discrimination between brain metastasis recurrence and radiation injury, without the need for acquiring and analyzing more time-consuming dynamic FET PET scans.

Classification:

Contributing Institute(s):
  1. Physik der Medizinischen Bildgebung (INM-4)
  2. Jara-Institut Quantum Information (INM-11)
  3. Jülich-Aachen Research Alliance - Translational Brain Medicine (JARA-BRAIN)
Research Program(s):
  1. 5253 - Neuroimaging (POF4-525) (POF4-525)

Click to display QR Code for this record

The record appears in these collections:
Dokumenttypen > Andere > Mitteilung
Institutssammlungen > INM > INM-11
Institutssammlungen > INM > INM-4
Workflowsammlungen > Öffentliche Einträge
Publikationsdatenbank

 Datensatz erzeugt am 2022-06-21, letzte Änderung am 2022-06-22



Dieses Dokument bewerten:

Rate this document:
1
2
3
 
(Bisher nicht rezensiert)