Head‐to‐head comparison of [<sup>18</sup>F]Flortaucipir and amyloid PET visual reads for differential diagnosis: An international, multi‐center study
Bibliographic record
Abstract
Abstract Background Quantitative [18F]Flortaucipir (FTP) tau PET analysis has good discriminative accuracy for Alzheimer’s disease (AD), but visual reads are used clinically. We aimed to compare the diagnostic accuracy of FTP‐PET to amyloid PET visual reads for distinguishing AD dementia and amyloid‐positive mild cognitive impairment (MCI) from other neurodegenerative diseases and cognitively normal (CN) controls. Method We included 661 participants studied with FTP and amyloid PET at UCSF, Lund University, and Gangnam Severance Hospital (Table 1). Amyloid PET tracers were site‐specific (UCSF, [11C]PIB; Lund, [18F]Flutemetamol; Gangnam, [18F]Florbetaben). Three visual raters interpreted amyloid PET using tracer‐specific criteria and FTP‐PET using a recently proposed method (Figure 1; Sonni et al. Alzheimer's Dement. 2020, 12:e12133. doi: 10.1002/dad2.12133). Amyloid and FTP‐PET were read in separate, randomized batches, blinded to the reads of the other physicians. The majority interpretation of each scan as amyloid positive/negative and tau positive/negative was used in the primary analysis. Diagnostic performance of FTP and amyloid PET visual reads, including sensitivity and specificity, were calculated. Kappa statistics for intra‐ and inter‐rater reliability were calculated. Result Amyloid and FTP‐PET visual reads had similar sensitivity (92.6%, 95%‐confidence interval [88.7‐95.5%], 92.2% [88.2‐95.2%], respectively) and specificity (82.9% [78.9‐86.5%], 78.2% [73.9‐82.1%], respectively) for distinguishing AD and amyloid‐positive MCI from all other dementias and CN controls (Table 2). Amyloid PET and FTP‐PET visual reads had similar diagnostic performance for distinguishing AD or amyloid‐positive MCI from non‐AD neurodegenerative disorders or CN controls (Table 3). FTP‐PET intra‐rater reliability ranged from κ=0.80‐1.00 (p<0.0001), and inter‐rater reliability between reader pairs ranged from κ=0.78‐0.85 (p<0.0001). Amyloid PET intra‐rater reliability was κ=0.97 (p<0.0001) for each reader, and inter‐rater reliability between reader pairs ranged from κ=0.82‐0.88 (p<0.0001). Conclusion FTP and amyloid PET visual reads have similar diagnostic performance and high intra‐ and inter‐rater reliability. Further analysis will be performed to determine if patient characteristics can predict when one PET modality will outperform the other for AD diagnosis.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.007 | 0.012 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.002 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".