Post on 16-Dec-2015
transcript
Agenda:
- Update on master tracers’ segmentation check- Update validation vs pathology- PMT application submission of “Study on the validation of VSRAD” (dr M. Nishikawa)- Benchmark: possible issues for platform
XI PMT meeting – July 3, 2012
Benchmark Images
ADNI scans: 2 x 5 Scheltens’s atrophy score x 2 sides
x 2 magnet strengths (1.5-3T)Total per rater: 40 hippos
5 Master Tracers’ segmentations:
Mapped in overlap on correspondent MRIChecked and reported causes for overlapping
discrepancies at 1.5TImproved Harmonized Protocol in many pointsMasters asked to correct if they agree, or to discuss the
issue if they disagree with the corrections
Benchmark Images
Overlapping agreement Volume ICCs
1.5T images: 0.71
3T images: 0.75
Validation vs Pathology
Validation versus pathology
Tracer (to be defined whether GP will segment all datasets, or if each centre will provide its tracer -
waiting for LA answer on privacy issues)
Segmentation based on Harmonized Protocol:
60 ante mortem46 postmortem MRIs
106 images
Analysis based on the information available for each dataset
Mayo Clinic~50 AD + MCI + CTRL dataset with pathologically confirmed
diagnosis (Braak’s stages),antemortem
left hippocampal measurement(Manual and Freesurfer)
Only volumetric measures to be shared
Mony DeLeon9 AD + 4 CTRL at 1.5T (T1)
13 postmortem and 2 AD antemortem
9 AD + 1 CTRL at 1.5T (PD)10 postmortem and 8 AD antemortem
(Neuron, plaque and tangle counts,Braak’s stages, abeta, tau,
histological volumetry, CA1 definition only on PD),3mm, coronal, no 3D navigation
one side hippocampal measurement (Manual. MIDAS)Images and volumetric measures can be shared
Liana Apostolova(23 hippos at 7T)
CA1 neuronal counts,tau and Abeta
immunoreactivity measures
Validation versus pathology
Originally designated sample: (Bobinski et al., 2000)11 AD + 4 CTRLpostmortem MRI and quantitative histology
de Leon: 9 AD + 4 CTRL at 1.5T (T1) and 9 AD + 1 CTRL at 1.5T (PD)23 postmortem and 10 AD antemortemNeuron, plaque and tangle counts, Braak’s stages, abeta, tau, histological volumetry (CA1 definition only on PD)3mm, coronal, no 3D navigationone side hippocampal measurement (Manual, MIDAS)
Jack: ~50 AD + MCI + CTRL antemortem MRIs with pathologically confirmed diagnosis (Braak’s stages). Volumes of left Hippocampus (Manual and Freesurfer). No post-mortem hippocampal measurement
Liana? 23 (one side only) 7T 60-hour postmortem scans (in progress: CA1 neuronal counts, tau and Abeta immunoreactivity measures)
“Study on the validation of VSRAD”Dr. Masami Nishikawa
VSRAD (Voxel-based Specific Regional analysis system for Alzheimer’s Disease): automatic VBM-based software for hippocampal atrophy.
Project aim: to validate the new version of VSRAD, comparing it versus the Harmonized Protocol as the gold standard method to manually measure hippocampal volume.
Subjects: 22 AD, 19 MCI, 18 controls+ 3 healthy volunteers that served as human phantoms for the pilot E-ADNI project (Frisoni et al., 2008): scanned by the 7 different machines.
Hippocampal segmentations carried out on these subjects will serve not only the aims of the present project, but will also contribute to provide data that will add to the validation of the Harmonized Protocol.
PMT application submission
Benchmark Maps: CSF exclusion
Harmonized Protocol criteria: internal CSF pools must be properly and independently segmented and excluded
MultiTracer: using one label generates higher variability in segmentations, since every tracer will "connect" somehow the internal pool to the external CSF
Solution: To segment the internal pools with an additional label, that will be used to subtract both volume and segmented voxels.
Benchmark Maps: plausible variability
Papers describing the project
Survey of protocols (preliminary phase; published, JAD 2011)Operationalization (preliminary phase; I revision, Alzheimer’s & Dementia, MS n. ADJ-D-12-00094)
Axes check short report (Brescia Team, in progress)Delphi consensus (Brescia Team, in progress)Master tracers’ practice and reliability (Brescia Team, in progr)
Development of certification platform (Duchesne and coll)Validation data and Protocol definition + Protocol (Brescia Team)Validation vs pathology (TBD)
DO
NE
IN P
RO
GR
ES
S P
LA
NN
ED
VALIDATION VS CURRENT PROTOCOLSASSESSMENT OF SOURCES OF
VARIANCE TRAINING SET DEVELOPMENT
VALIDATION VS PATHOLOGY
GOLD STANDARD
Harmonized ProtocolADNI scans: 2 x 5 Scheltens’s
atrophy score x 2 sides x 2 magnet strengths (1.5-3T)
Total per rater: 40 hippos
Harmonized ProtocolADNI scans: 2 sides x 5 Scheltens’s atrophy scores
x 3 time points (bl-1y-2y) x 3 scanners (+ retracing @ bl)
x 2 magnet strengths (1.5-3T)Total per rater: 240 hippos
Assessment of variance due to rater and center
Local ProtocolADNI scans: 2 x 5 Scheltens’s atrophy scores
x 2 sides x 2 magnet strengths (1.5-3T)
Harmonized ProtocolADNI scans: 2 x 5 Scheltens’s atrophy score x
2 sides x 2 magnet strength (1.5-3T)Total per rater: 40 hippos Harmonized Protocol:
Pathological datasets: Mayo Clinic and NYU
Total: about 40 hippos
TrainingADNI scans: 10 at 1.5T x 2 sides x 7
SUs x 2 tracing roundsTotal per rater: 40 hippos
20 naïve tracers 5 master tracers 1 tracer
REFERENCE PROBABILISTIC MASKS
with 95% C.I.
QUALIFICATION QUALIFICATION
Best 5 naïve tracers
Assessment of variance due to side, trace-retrace, atrophy, time, scanner, rater
TRAINING SET
Assessment of agreement with volume on pathology or ex vivo
MRI and correlation with neuronal density
GANTT