Evaluating the Reliability and Precision of 6° and 16° Flip Angles Compared to Inversion Recovery When Generating T1 Relaxation Maps in MRI on a 3.0T Scanner in Relation to Cerebral Cancer.
Translated title
Evaluering af Reliabilitet og Præcision for 6 og 16 Graders Flip Angles Sammenlignet med Inversion Recovery ved Generering af T1 Relaxations Maps i MR på en 3.0T skanner i Relation til Cerebral Cancer
Author
Lund, Kristian
Term
4. term
Education
Publication year
2014
Submitted on
2014-06-03
Abstract
Glioblastoma multiforme er en aggressiv hjernetumor, hvor vurdering af behandlingseffekt med DCE‑MRI ofte hæmmes af vanskeligheder ved kvantificering. Dette projekt havde til formål at vurdere pålidelighed og præcision af en foruddefineret Variable Flip Angle‑protokol med 6° og 16° sammenlignet med inversion recovery (IR) som gylden standard til generering af T1‑relaxationsmaps på en 3,0T skanner. Seks fantomer med destilleret vand og varierende koncentrationer af Dotarem blev skannet tre gange med både VFA og IR. T1‑maps blev genereret og analyseret i MatLab og Imalytics; variation og reproducerbarhed blev vurderet med varianskoefficient (CV) og intraklassekorrelation (ICC), visualiseret med grafer/Bland‑Altman‑plots, og ICC blev beregnet i SPSS. Resultaterne viste lavere variation og højere overensstemmelse for IR (CV = 0,82 %; ICC = 1, konfidensinterval 1/1) end for VFA (CV = 12,15 %; ICC = 0,723, konfidensinterval 0,266/0,950). En mulig outlier kan have påvirket datasættet; den blev ikke ekskluderet, da den ikke med sikkerhed kunne identificeres som en outlier, men en ekstra analyse uden den blev også udført. Samlet indikerer fundene, at IR under disse betingelser er markant mere pålidelig og præcis end den testede 6°/16° VFA‑protokol til T1‑mapping, hvilket er relevant for korrektion af DCE‑MRI i forskning vedrørende cerebral cancer.
Glioblastoma multiforme is an aggressive brain tumor, and treatment monitoring with DCE‑MRI is often limited by challenges in quantification. This project aimed to evaluate the reliability and precision of a predefined Variable Flip Angle protocol using 6° and 16° compared with inversion recovery (IR), the gold standard, for generating T1 relaxation maps on a 3.0T scanner. Six phantoms containing distilled water and varying concentrations of the contrast agent Dotarem were scanned three times with both VFA and IR. T1 maps were generated and analyzed in MatLab and Imalytics; variability and reproducibility were assessed using the coefficient of variation (CV) and intraclass correlation coefficient (ICC), visualized with graphs/Bland‑Altman plots, and ICC was computed in SPSS. Results showed lower variability and higher agreement for IR (CV = 0.82%; ICC = 1, confidence interval 1/1) than for VFA (CV = 12.15%; ICC = 0.723, confidence interval 0.266/0.950). A potential outlier may have influenced the dataset; it was not excluded as it could not be confirmed, though an additional analysis excluding it was performed. Overall, the findings indicate that IR, under these conditions, is markedly more reliable and precise than the tested 6°/16° VFA protocol for T1 mapping, which is pertinent for correcting DCE‑MRI in cerebral cancer research.
[This summary has been generated with the help of AI directly from the project (PDF)]
Documents