Reliability of cortical lesion detection on double inversion recovery MRI applying the MAGNIMS-Criteria in multiple sclerosis patients within a 16-months period
Standard
Reliability of cortical lesion detection on double inversion recovery MRI applying the MAGNIMS-Criteria in multiple sclerosis patients within a 16-months period. / Faizy, Tobias; Thaler, Christian; Ceyrowski, Tim; Broocks, Gabriel; Treffler, Natascha; Sedlacik, Jan; Stürner, Klarissa Hanja; Stellmann, Jan-Patrick; Heesen, Christoph; Fiehler, Jens; Siemonsen, Susanne.
in: PLOS ONE, Jahrgang 12, Nr. 2, 24.02.2017, S. e0172923.Publikationen: SCORING: Beitrag in Fachzeitschrift/Zeitung › SCORING: Zeitschriftenaufsatz › Forschung › Begutachtung
Harvard
APA
Vancouver
Bibtex
}
RIS
TY - JOUR
T1 - Reliability of cortical lesion detection on double inversion recovery MRI applying the MAGNIMS-Criteria in multiple sclerosis patients within a 16-months period
AU - Faizy, Tobias
AU - Thaler, Christian
AU - Ceyrowski, Tim
AU - Broocks, Gabriel
AU - Treffler, Natascha
AU - Sedlacik, Jan
AU - Stürner, Klarissa Hanja
AU - Stellmann, Jan-Patrick
AU - Heesen, Christoph
AU - Fiehler, Jens
AU - Siemonsen, Susanne
PY - 2017/2/24
Y1 - 2017/2/24
N2 - PURPOSE: In patients with multiple sclerosis (MS), Double Inversion Recovery (DIR) magnetic resonance imaging (MRI) can be used to identify cortical lesions (CL). We sought to evaluate the reliability of CL detection on DIR longitudinally at multiple subsequent time-points applying the MAGNIMs scoring criteria for CLs.METHODS: 26 MS patients received a 3T-MRI (Siemens, Skyra) with DIR at 12 time-points (TP) within a 16 months period. Scans were assessed in random order by two different raters. Both raters separately marked all CLs on each scan and total lesion numbers were obtained for each scan-TP and patient. After a retrospective re-evaluation, the number of consensus CLs (conL) was defined as the total number of CLs, which both raters finally agreed on. CLs volumes, relative signal intensities and CLs localizations were determined. Both ratings (conL vs. non-consensus scoring) were compared for further analysis.RESULTS: A total number of n = 334 CLs were identified by both raters in 26 MS patients with a first agreement of both raters on 160 out of 334 of the CLs found (κ = 0.48). After the retrospective re-evaluation, consensus agreement increased to 233 out of 334 CL (κ = 0.69). 93.8% of conL were visible in at least 2 consecutive TP. 74.7% of the conL were visible in all 12 consecutive TP. ConL had greater mean lesion volumes and higher mean signal intensities compared to lesions that were only detected by one of the raters (p<0.05). A higher number of CLs in the frontal, parietal, temporal and occipital lobe were identified by both raters than the number of those only identified by one of the raters (p<0.05).CONCLUSIONS: After a first assessment, slightly less than a half of the CL were considered as reliably detectable on longitudinal DIR images. A retrospective re-evaluation notably increased the consensus agreement. However, this finding is narrowed, considering the fact that retrospective evaluation steps might not be practicable in clinical routine. Lesions that were not reliably identifiable by both raters seem to be characterized by lower signal intensity and smaller size, or located in distinct anatomical brain regions.
AB - PURPOSE: In patients with multiple sclerosis (MS), Double Inversion Recovery (DIR) magnetic resonance imaging (MRI) can be used to identify cortical lesions (CL). We sought to evaluate the reliability of CL detection on DIR longitudinally at multiple subsequent time-points applying the MAGNIMs scoring criteria for CLs.METHODS: 26 MS patients received a 3T-MRI (Siemens, Skyra) with DIR at 12 time-points (TP) within a 16 months period. Scans were assessed in random order by two different raters. Both raters separately marked all CLs on each scan and total lesion numbers were obtained for each scan-TP and patient. After a retrospective re-evaluation, the number of consensus CLs (conL) was defined as the total number of CLs, which both raters finally agreed on. CLs volumes, relative signal intensities and CLs localizations were determined. Both ratings (conL vs. non-consensus scoring) were compared for further analysis.RESULTS: A total number of n = 334 CLs were identified by both raters in 26 MS patients with a first agreement of both raters on 160 out of 334 of the CLs found (κ = 0.48). After the retrospective re-evaluation, consensus agreement increased to 233 out of 334 CL (κ = 0.69). 93.8% of conL were visible in at least 2 consecutive TP. 74.7% of the conL were visible in all 12 consecutive TP. ConL had greater mean lesion volumes and higher mean signal intensities compared to lesions that were only detected by one of the raters (p<0.05). A higher number of CLs in the frontal, parietal, temporal and occipital lobe were identified by both raters than the number of those only identified by one of the raters (p<0.05).CONCLUSIONS: After a first assessment, slightly less than a half of the CL were considered as reliably detectable on longitudinal DIR images. A retrospective re-evaluation notably increased the consensus agreement. However, this finding is narrowed, considering the fact that retrospective evaluation steps might not be practicable in clinical routine. Lesions that were not reliably identifiable by both raters seem to be characterized by lower signal intensity and smaller size, or located in distinct anatomical brain regions.
KW - Adult
KW - Cerebral Cortex/pathology
KW - Female
KW - Humans
KW - Image Processing, Computer-Assisted
KW - Magnetic Resonance Imaging/methods
KW - Male
KW - Middle Aged
KW - Multiple Sclerosis, Relapsing-Remitting/diagnostic imaging
KW - Reproducibility of Results
KW - Retrospective Studies
KW - Young Adult
U2 - 10.1371/journal.pone.0172923
DO - 10.1371/journal.pone.0172923
M3 - SCORING: Journal article
C2 - 28235075
VL - 12
SP - e0172923
JO - PLOS ONE
JF - PLOS ONE
SN - 1932-6203
IS - 2
ER -