pith. sign in

arxiv: 2604.11176 · v2 · submitted 2026-04-13 · 💻 cs.CV

Precision Synthesis of Multi-Tracer PET via VLM-Modulated Rectified Flow for Stratifying Mild Cognitive Impairment

Tuo Liu , Shuijin Lin , Shaozhen Yan , HaiFeng Wang , Jie Lu , Jianhua Ma , Chunfeng Lian This is my paper

Pith reviewed 2026-05-10 16:06 UTC · model grok-4.3

classification 💻 cs.CV
keywords PET image synthesismild cognitive impairmentrectified flowvision-language modelAlzheimer's diseasemulti-modal neuroimaginggenerative models18F-FDG PET
0
0 comments X

The pith

DIReCT++ synthesizes high-fidelity multi-tracer PET from MRI and clinical data to enable precise personalized stratification of mild cognitive impairment.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces DIReCT++, a generative model that creates synthetic 18F-AV-45 and 18F-FDG PET images from MRI scans plus basic clinical scores. This tackles the practical barriers of real PET imaging, namely high cost and radiation exposure, that limit early Alzheimer's screening. The method combines a 3D rectified flow network with a biomedical vision-language model to produce subject-specific images that preserve disease-related patterns. Tests on multi-center data show the synthetic scans achieve strong visual and quantitative fidelity while supporting better separation of mild cognitive impairment cases when fused with MRI. The result points toward a lower-barrier way to incorporate PET-level information into routine early diagnosis workflows.

Core claim

DIReCT++ integrates a 3D rectified flow architecture with a domain-adapted vision-language model (BiomedCLIP) to synthesize subject-specific 18F-AV-45 and 18F-FDG PET images from MRI and clinical scores. This synthesis achieves superior fidelity and generalizability while recapitulating disease-specific patterns, enabling precise personalized stratification of mild cognitive impairment when combined with MRI.

What carries the argument

Domain-Informed ReCTified flow (DIReCT++) model modulated by BiomedCLIP vision-language model for text-guided, personalized multi-tracer PET synthesis from MRI.

Load-bearing premise

The synthetic PET images accurately recapitulate true disease-specific patterns without introducing artifacts or biases that would affect clinical decisions.

What would settle it

A prospective study in which adding the synthetic PET images to MRI fails to raise MCI stratification accuracy above MRI alone, or where the synthetic images show poor correlation with real PET in amyloid or glucose-uptake patterns.

Figures

Figures reproduced from arXiv: 2604.11176 by Chunfeng Lian, HaiFeng Wang, Jianhua Ma, Jie Lu, Shaozhen Yan, Shuijin Lin, Tuo Liu.

Figure 1
Figure 1. Figure 1: Overview of the DIReCT++ framework for multimodal PET synthesis and downstream analysis. (a) Schematic of the rectified flow (RF) process that synthesizes dual-tracer 18F-AV-45 and 18F-FDG PET images from MRI. (b) Fine-tuning of the pre-trained BiomedCLIP model via prompt learning to encode tracer-specific and subject-level text guidance. (c) Training pipeline of the conditional 3D RF model implemented as … view at source ↗
Figure 2
Figure 2. Figure 2: Comparison of PET reconstruction quality across methods and datasets. (a) Radar plots show normalized quantitative results (SSIM, MSE, MAE, PSNR) for FDG-PET and AV45-PET reconstruction on ADNI, and cross-dataset generalization on OASIS. Notably, each metric is normalized by the best value among all methods. (b) Representative PET synthesis results for cognitively normal (CN, top), mild cognitive impairmen… view at source ↗
Figure 3
Figure 3. Figure 3: Detailed regional visualization of synthetic and real PET images. The figure displays magnified regional views of FDG-PET (left two columns) and AV45-PET (right two columns) for representative AD and CN subjects. For each modality and group, the first sub-column shows real PET, and the second sub-column shows synthetic PET. Key anatomical regions are highlighted in the bottom-right corner of each sub-colum… view at source ↗
Figure 4
Figure 4. Figure 4: Regional comparison of mean uptake between real and MRI-synthetic PET images. (a,b) Violin plots of regional mean uptake across cortical and subcortical regions for AV45 (a) and FDG (b), with statistical comparisons. (c–f) Disease￾related patterns. Violin plots show mean uptake in AD (red) and CN (green). For each region, paired half-violins denote real PET (left, solid) and synthetic PET (right, dashed). … view at source ↗
Figure 5
Figure 5. Figure 5: Diagnostic classification performance using MRI-synthetic PET images across different clinical tasks. (a–c) Bar plots illustrate the classification performance—accuracy (ACC), sensitivity (SEN), specificity (SPE), and area under the ROC curve (AUC)—of eleven input-modality combinations evaluated for three diagnostic tasks: (a) Alzheimer’s disease (AD) vs. cognitively normal (CN), (b) mild cognitive impairm… view at source ↗
read the original abstract

The biological definition of Alzheimer's disease (AD) relies on multi-modal neuroimaging, yet the clinical utility of positron emission tomography (PET) is limited by cost and radiation exposure, hindering early screening at preclinical or prodromal stages. While generative models offer a promising alternative by synthesizing PET from magnetic resonance imaging (MRI), achieving subject-specific precision remains a primary challenge. Here, we introduce DIReCT$++$, a Domain-Informed ReCTified flow model for synthesizing multi-tracer PET from MRI combined with fundamental clinical information. Our approach integrates a 3D rectified flow architecture to capture complex cross-modal and cross-tracer relationships with a domain-adapted vision-language model (BiomedCLIP) that provides text-guided, personalized generation using clinical scores and imaging knowledge. Extensive evaluations on multi-center datasets demonstrate that DIReCT$++$ not only produces synthetic PET images ($^{18}$F-AV-45 and $^{18}$F-FDG) of superior fidelity and generalizability but also accurately recapitulates disease-specific patterns. Crucially, combining these synthesized PET images with MRI enables precise personalized stratification of mild cognitive impairment (MCI), advancing a scalable, data-efficient tool for the early diagnosis and prognostic prediction of AD. The source code will be released on https://github.com/ladderlab-xjtu/DIReCT-PLUS.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The manuscript introduces DIReCT++, a Domain-Informed ReCTified flow model that combines a 3D rectified flow architecture with a domain-adapted vision-language model (BiomedCLIP) to synthesize multi-tracer PET images (^{18}F-AV-45 and ^{18}F-FDG) from MRI plus clinical information. It claims superior fidelity and generalizability on multi-center datasets, accurate recapitulation of disease-specific patterns, and that the resulting synthetic PET enables precise personalized stratification of mild cognitive impairment when combined with MRI.

Significance. If the quantitative claims hold with rigorous validation, the work has substantial significance for scalable early AD screening by reducing reliance on costly and radioactive PET scans. The VLM-modulated rectified flow for personalized, cross-tracer synthesis is a technically interesting direction, and the planned release of source code would support reproducibility.

major comments (2)
  1. Abstract: The assertions of 'superior fidelity and generalizability' and 'precise personalized stratification of mild cognitive impairment' are presented without any quantitative metrics, error bars, ablation studies, baseline comparisons, or details on how fidelity or stratification performance (e.g., AUC) was measured. Full results sections must supply these to allow assessment of the central claims.
  2. Evaluation/Results section: The claim that synthetic PET accurately recapitulates subject-specific disease patterns (regional AV-45/FDG deviations) and drives improved MCI stratification requires explicit evidence that gains survive ablation of the PET synthesis branch and hold on fully held-out multi-center cohorts with independent clinical labels. Without such controls, improvements could arise from joint-training artifacts or BiomedCLIP conditioning leakage rather than genuine multi-tracer fidelity.
minor comments (1)
  1. Abstract: The model name appears with a LaTeX artifact ('DIReCT$++$'); render consistently as DIReCT++ throughout.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the thorough and constructive review. The comments highlight important areas for improving the clarity and rigor of our claims. We address each major comment point by point below and will incorporate revisions to strengthen the manuscript.

read point-by-point responses

  1. Referee: Abstract: The assertions of 'superior fidelity and generalizability' and 'precise personalized stratification of mild cognitive impairment' are presented without any quantitative metrics, error bars, ablation studies, baseline comparisons, or details on how fidelity or stratification performance (e.g., AUC) was measured. Full results sections must supply these to allow assessment of the central claims.

    Authors: We agree that the abstract would benefit from quantitative support to better substantiate its claims. Although abstracts are concise summaries, we will revise the abstract in the next version to include key metrics (e.g., PSNR/SSIM for fidelity, AUC for MCI stratification) along with brief mentions of baseline comparisons and ablation results. The full results section already reports these with error bars, multi-center evaluations, and baseline comparisons; we will add explicit cross-references from the abstract to those sections. revision: yes

  2. Referee: Evaluation/Results section: The claim that synthetic PET accurately recapitulates subject-specific disease patterns (regional AV-45/FDG deviations) and drives improved MCI stratification requires explicit evidence that gains survive ablation of the PET synthesis branch and hold on fully held-out multi-center cohorts with independent clinical labels. Without such controls, improvements could arise from joint-training artifacts or BiomedCLIP conditioning leakage rather than genuine multi-tracer fidelity.

    Authors: We appreciate the emphasis on rigorous controls to isolate the contribution of the synthesized PET. Our multi-center evaluations already include quantitative comparisons to baselines and visual/metric-based recapitulation of disease-specific patterns. To directly address the concern, we will add new ablation experiments in the revised manuscript that remove the PET synthesis branch (showing that stratification gains depend on the synthetic PET) and report performance on fully held-out multi-center cohorts using independent clinical labels. These additions will confirm that improvements stem from multi-tracer fidelity rather than artifacts or leakage. revision: yes

Circularity Check

0 steps flagged

No significant circularity in the derivation chain

full rationale

The paper presents DIReCT++ as an empirical generative model (3D rectified flow modulated by BiomedCLIP) trained on multi-center datasets to synthesize PET from MRI plus clinical scores. All performance claims—fidelity, pattern recapitulation, and MCI stratification gains—are framed as outcomes of external evaluations rather than any closed mathematical derivation. No equations, uniqueness theorems, or self-referential definitions appear in the abstract or described content that would reduce outputs to fitted inputs by construction. The argument remains self-contained on held-out data without load-bearing self-citations or ansatzes imported from prior author work.

Axiom & Free-Parameter Ledger

1 free parameters · 2 axioms · 1 invented entities

The central claim rests on standard assumptions of deep generative models plus the unverified premise that the BiomedCLIP-guided flow produces clinically faithful PET images. No new physical entities are postulated.

free parameters (1)
  • model hyperparameters and training schedule
    Typical neural network weights and flow parameters are fitted to the training data; exact values not stated in abstract.
axioms (2)
  • domain assumption Rectified flow can capture complex cross-modal and cross-tracer relationships in 3D neuroimaging data
    Invoked by the choice of 3D rectified flow architecture.
  • domain assumption BiomedCLIP embeddings provide useful text-guided conditioning for personalized PET synthesis
    Central to the VLM-modulation component.
invented entities (1)
  • DIReCT++ architecture no independent evidence
    purpose: Domain-informed rectified flow for multi-tracer PET synthesis
    New proposed model combining existing components; no independent evidence outside this work.

pith-pipeline@v0.9.0 · 5564 in / 1558 out tokens · 48230 ms · 2026-05-10T16:06:43.692684+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

37 extracted references · 37 canonical work pages

  1. [1]

    Alzheimer’s disease.The Lancet, 397(10284):1577–1590, 2021

    Philip Scheltens, Bart De Strooper, Miia Kivipelto, Henne Holstege, Gael Chételat, Charlotte E Teunissen, Jeffrey Cummings, and Wiesje M van der Flier. Alzheimer’s disease.The Lancet, 397(10284):1577–1590, 2021

    work page 2021

  2. [2]

    Neuroimaging in dementia: more than typical Alzheimer disease

    Sven Haller, Hans Rolf Jäger, Meike W Vernooij, and Frederik Barkhof. Neuroimaging in dementia: more than typical Alzheimer disease. Radiology, 308(3):e230173, 2023

    work page 2023

  3. [3]

    Amyloid and tau PET-positive cognitively unimpaired individuals are at high risk for future cognitive decline.Nat

    Rik Ossenkoppele, Alexa Pichet Binette, Colin Groot, Ruben Smith, Olof Strandberg, Sebastian Palmqvist, Erik Stomrud, Pontus Tideman, Tomas Ohlsson, Jonas Jögi, et al. Amyloid and tau PET-positive cognitively unimpaired individuals are at high risk for future cognitive decline.Nat. Med., 28(11):2381–2387, 2022

    work page 2022

  4. [4]

    Sex modifies APOE𝜀4 dose effect on brain tau deposition in cognitively impaired individuals.Brain, 144(10):3201–3211, 2021

    Shaozhen Yan, Chaojie Zheng, Manish D Paranjpe, Yanxiao Li, Weihua Li, Xiuying Wang, Tammie LS Benzinger, Jie Lu, and Yun Zhou. Sex modifies APOE𝜀4 dose effect on brain tau deposition in cognitively impaired individuals.Brain, 144(10):3201–3211, 2021

    work page 2021

  5. [5]

    TheuseofPETinAlzheimerdisease.Nat.Rev.Neurol.,6(2):78–87, 2010

    AgnetaNordberg,JuhaORinne,AhmadulKadir,andBengtLångström. TheuseofPETinAlzheimerdisease.Nat.Rev.Neurol.,6(2):78–87, 2010

    work page 2010

  6. [6]

    Deep learning based synthesis of MRI, CT and PET: Review and analysis.Med

    Sanuwani Dayarathna, Kh Tohidul Islam, Sergio Uribe, Guang Yang, Munawar Hayat, and Zhaolin Chen. Deep learning based synthesis of MRI, CT and PET: Review and analysis.Med. Image Anal., 92:103046, 2024

    work page 2024

  7. [7]

    Synthesizing images of tau pathology from cross-modal neuroimaging using deep learning.Brain, 147(3):980–995, 2024

    Jeyeon Lee, Brian J Burkett, Hoon-Ki Min, Matthew L Senjem, Ellen Dicks, Nick Corriveau-Lecavalier, Carly T Mester, Heather J Wiste, Emily S Lundt, Melissa E Murray, et al. Synthesizing images of tau pathology from cross-modal neuroimaging using deep learning.Brain, 147(3):980–995, 2024

    work page 2024

  8. [8]

    Synthesizing missing PET from MRI with cycle- consistent generative adversarial networks for Alzheimer’s disease diagnosis.Int

    Yongsheng Pan, Mingxia Liu, Chunfeng Lian, Tao Zhou, Yong Xia, and Dinggang Shen. Synthesizing missing PET from MRI with cycle- consistent generative adversarial networks for Alzheimer’s disease diagnosis.Int. Conf. Med. Image Comput. Comput.-Assist. Interv., pages 455–463, 2018

    work page 2018

  9. [9]

    Spatially-constrained fisher representation for brain disease identification with incomplete multi-modal neuroimages.IEEE Trans

    Yongsheng Pan, Mingxia Liu, Chunfeng Lian, Yong Xia, and Dinggang Shen. Spatially-constrained fisher representation for brain disease identification with incomplete multi-modal neuroimages.IEEE Trans. Med. Imaging, 39(9):2965–2975, 2020

    work page 2020

  10. [10]

    Applications of generative adversarialnetworks(GANs)inpositronemissiontomography(PET)imaging:Areview.Eur.J.Nucl.Med.Mol.Imaging,49(11):3717–3739, 2022

    Ioannis D Apostolopoulos, Nikolaos D Papathanasiou, Dimitris J Apostolopoulos, and George S Panayiotakis. Applications of generative adversarialnetworks(GANs)inpositronemissiontomography(PET)imaging:Areview.Eur.J.Nucl.Med.Mol.Imaging,49(11):3717–3739, 2022

    work page 2022

  11. [11]

    Phys., 51(8):5250–5269, 2024

    TaofengXie,ChentaoCao,Zhuo-xuCui,YuGuo,CaiyingWu,XuemeiWang,QingnengLi,ZhanliHu,TaoSun,ZiruSang,etal.Synthesizing PET images from high-field and ultra-high-field MR images using joint diffusion attention model.Med. Phys., 51(8):5250–5269, 2024

    work page 2024

  12. [12]

    Amulti-viewlearningapproach withdiffusionmodeltosynthesizeFDGPETfromMRIT1WIfordiagnosisofAlzheimer’sdisease.AlzheimersDement.,21(2):e14421,2025

    KeChen,YingWeng,YueqinHuang,YimingZhang,TomDening,AkramAHosseini,andWeizhongXiao. Amulti-viewlearningapproach withdiffusionmodeltosynthesizeFDGPETfromMRIT1WIfordiagnosisofAlzheimer’sdisease.AlzheimersDement.,21(2):e14421,2025

    work page 2025

  13. [13]

    MRI2PET: Realistic PET image synthesis from MRI for automated inference of brain atrophy and Alzheimer’s.medRxiv, pages 2025–04, 2025

    Brandon Theodorou, Anant Dadu, Brian Avants, Mike Nalls, Jimeng Sun, and Faraz Faghri. MRI2PET: Realistic PET image synthesis from MRI for automated inference of brain atrophy and Alzheimer’s.medRxiv, pages 2025–04, 2025

    work page 2025

  14. [14]

    Unisyn: a generativefoundationmodelforuniversalmedicalimagesynthesisacrossMRI,CTandPET.Int.Conf.Med.ImageComput.Comput.-Assist

    Yulin Wang, Honglin Xiong, Kaicong Sun, Jiameng Liu, Xin Lin, Ziyi Chen, Yuanzhe He, Qian Wang, and Dinggang Shen. Unisyn: a generativefoundationmodelforuniversalmedicalimagesynthesisacrossMRI,CTandPET.Int.Conf.Med.ImageComput.Comput.-Assist. Interv., pages 673–682, 2025

    work page 2025

  15. [15]

    DIReCT:Domain-informedrectifiedflowforcontrollable brain MRI to PET translation.Int

    TuoLiu,HaifengWang,HengChang,FanWang,ChunfengLian,andJianhuaMa. DIReCT:Domain-informedrectifiedflowforcontrollable brain MRI to PET translation.Int. Conf. Inf. Process. Med. Imaging, pages 218–231, 2025

    work page 2025

  16. [16]

    Imaging the evolution and pathophysiology of Alzheimer disease.Nat

    William Jagust. Imaging the evolution and pathophysiology of Alzheimer disease.Nat. Rev. Neurosci., 19(11):687–700, 2018

    work page 2018

  17. [17]

    A multimodal biomedical foundation model trained from fifteen million image–text pairs.NEJM AI, 2(1):AIoa2400640, 2025

    ShengZhang,YanboXu,NaotoUsuyama,HanwenXu,JaspreetBagga,RobertTinn,SamPreston,RajeshRao,MuWei,NaveenValluri,etal. A multimodal biomedical foundation model trained from fifteen million image–text pairs.NEJM AI, 2(1):AIoa2400640, 2025

    work page 2025

  18. [18]

    TheAlzheimer’sdiseaseneuroimaginginitiative:progressreportandfutureplans.AlzheimersDement., 6(3):202–211, 2010

    Michael W Weiner, Paul S Aisen, Clifford R Jack Jr, William J Jagust, John Q Trojanowski, Leslie Shaw, Andrew J Saykin, John C Morris, NigelCairns,LaurelABeckett,etal. TheAlzheimer’sdiseaseneuroimaginginitiative:progressreportandfutureplans.AlzheimersDement., 6(3):202–211, 2010

    work page 2010

  19. [19]

    OASIS-3: longitudinal neuroimaging, clinical, and cognitive dataset for normal aging and Alzheimer disease.medrxiv, pages 2019–12, 2019

    Pamela J LaMontagne, Tammie LS Benzinger, John C Morris, Sarah Keefe, Russ Hornbeck, Chengjie Xiong, Elizabeth Grant, Jason Hassenstab, Krista Moulder, Andrei G Vlassenko, et al. OASIS-3: longitudinal neuroimaging, clinical, and cognitive dataset for normal aging and Alzheimer disease.medrxiv, pages 2019–12, 2019

    work page 2019

  20. [20]

    Hierarchical fully convolutional network for joint atrophy localization and Alzheimer’s disease diagnosis using structural MRI.IEEE Trans

    Chunfeng Lian, Mingxia Liu, Jun Zhang, and Dinggang Shen. Hierarchical fully convolutional network for joint atrophy localization and Alzheimer’s disease diagnosis using structural MRI.IEEE Trans. Pattern Anal. Mach. Intell., 42(4):880–893, 2018

    work page 2018

  21. [21]

    Densely connected convolutional networks.Proc

    Gao Huang, Zhuang Liu, Laurens Van Der Maaten, and Kilian Q Weinberger. Densely connected convolutional networks.Proc. IEEE Conf. Comput. Vis. Pattern Recognit., pages 4700–4708, 2017

    work page 2017

  22. [22]

    Freesurfer.NeuroImage, 62(2):774–781, 2012

    Bruce Fischl. Freesurfer.NeuroImage, 62(2):774–781, 2012

    work page 2012

  23. [23]

    A 12-step user guide for analyzing voxel-wise gray matter asymmetries in statistical parametric mapping (SPM).Nat

    Florian Kurth, Christian Gaser, and Eileen Luders. A 12-step user guide for analyzing voxel-wise gray matter asymmetries in statistical parametric mapping (SPM).Nat. Protoc., 10:293–304, 2015

    work page 2015

  24. [24]

    Greve, Oula Puonti, Axel Thielscher, Koen Van Leemput, Bruce Fischl, Adrian V

    Colin Billot, Douglas N. Greve, Oula Puonti, Axel Thielscher, Koen Van Leemput, Bruce Fischl, Adrian V. Dalca, and Juan E. Iglesias. SynthSeg: Segmentation of brain MRI scans of any contrast and resolution without retraining.Med. Image Anal., 86:102789, 2023

    work page 2023

  25. [25]

    Jun-Yan Zhu, Taesung Park, Phillip Isola, and Alexei A. Efros. Unpaired image-to-image translation using cycle-consistent adversarial networks.2017 IEEE Int. Conf. Comput. Vis. ICCV, pages 2242–2251, 2017

    work page 2017

  26. [26]

    Swin-unet:Unet-likepuretransformer for medical image segmentation.ArXiv Prepr

    HaoyuCao,YueyueWang,JienengChen,DongshengJiang,XiaohuiZhang,QiTian,andMengWang. Swin-unet:Unet-likepuretransformer for medical image segmentation.ArXiv Prepr. ArXiv210505537 CsCV, 2021

    work page 2021

  27. [27]

    Mazurowski

    Nicholas Konz, Yuwen Chen, Haoyu Dong, and Maciej A. Mazurowski. Anatomically-controllable medical image generation with segmentation-guided diffusion models.ArXiv Prepr. ArXiv240205210 EessIV, 2024. Accepted at Medical Image Computing and Computer- Assisted Intervention (MICCAI)

    work page 2024

  28. [28]

    Denoising diffusion-based MRI to CT image translation T

    Robert Graf, Joachim Schmitt, Sarah Schlaeger, Hendrik Kristian Möller, Vasiliki Sideri-Lampretsa, Anjany Sekuboyina, Sandro Manuel Krieg, Benedikt Wiestler, Bjoern Menze, Daniel Rueckert, and Jan Stefan Kirschke. Denoising diffusion-based MRI to CT image translation T. Liu et al.:Preprint Page 14 of 15 enables automated spinal segmentation.Eur. Radiol. E...

    work page 2023

  29. [29]

    Flow straight and fast: Learning to generate and transfer data with rectified flow.ArXiv Prepr

    Qiang Liu Xingchao Liu, Chengyue Gong. Flow straight and fast: Learning to generate and transfer data with rectified flow.ArXiv Prepr. ArXiv220903003, 2022

    work page 2022

  30. [30]

    Mueller, Michael W

    Susanne G. Mueller, Michael W. Weiner, Leon J. Thal, Ronald C. Petersen, Clifford Jack, William Jagust, John Q. Trojanowski, Arthur W. Toga, and Laurel Beckett. The alzheimer’s disease neuroimaging initiative.Neuroimaging Clin. N. Am., 15(4):869–877, 2005

    work page 2005

  31. [31]

    Wong, Paul B

    Dean F. Wong, Paul B. Rosenberg, Yun Zhou, Anil Kumar, Victoria Raymont, Hayden T. Ravert, Robert F. Dannals, Abhijit Nandi, James R. Brasić, Weiguo Ye, John Hilton, Constantine Lyketsos, Hank F. Kung, Abhinay D. Joshi, Daniel M. Skovronsky, and Michael J. Pontecorvo. Invivoimagingofamyloiddepositioninalzheimerdiseaseusingtheradioligand 18F-AV-45(florbeta...

    work page 2010

  32. [32]

    Controlling the false discovery rate: a practical and powerful approach to multiple testing.J

    Yoav Benjamini and Yosef Hochberg. Controlling the false discovery rate: a practical and powerful approach to multiple testing.J. R. Stat. Soc. Ser. B Methodol., 57(1):289–300, 1995

    work page 1995

  33. [33]

    SynthSeg: Domain randomization for segmentation of brain scans of any contrast and resolution

    AdrianVDalca,ElodiePetit,BenjaminBillot,JuanEugenioIglesias,BruceFischl,MertRSabuncu,JasonTourville,WilliamMWells,Koen Van Leemput, François Rousseau, et al. SynthSeg: Domain randomization for segmentation of brain scans of any contrast and resolution. Med. Image Anal., 80:102480, 2022

    work page 2022

  34. [34]

    Blood biomarkers for Alzheimer’s disease in clinical practice and trials

    Oskar Hansson, Kaj Blennow, Henrik Zetterberg, and Jeffrey Dage. Blood biomarkers for Alzheimer’s disease in clinical practice and trials. Nat. Aging, 3(5):506–519, 2023

    work page 2023

  35. [35]

    Flexibly distilled 3D rectified flow with anatomicalconstraintsfordevelopmentalinfantbrainMRIprediction.Int.Conf.Med.ImageComput.Comput.-Assist.Interv.,pages228–237, 2025

    Haifeng Wang, Zehua Ren, Heng Chang, Xinmei Qiu, Fan Wang, Chunfeng Lian, and Jianhua Ma. Flexibly distilled 3D rectified flow with anatomicalconstraintsfordevelopmentalinfantbrainMRIprediction.Int.Conf.Med.ImageComput.Comput.-Assist.Interv.,pages228–237, 2025

    work page 2025

  36. [36]

    Controllableflowmatchingfor 3D contrast-enhanced brain MRI synthesis from non-contrast scans.Int

    HengChang,YuShang,HaifengWang,YuxiaLiang,HaoyuWang,FanWang,ChenNiu,andChunfengLian. Controllableflowmatchingfor 3D contrast-enhanced brain MRI synthesis from non-contrast scans.Int. Conf. Med. Image Comput. Comput.-Assist. Interv., pages 119–128, 2025

    work page 2025

  37. [37]

    Portable,low-fieldmagneticresonanceimagingforevaluationofAlzheimer’sdisease

    AnnabelJSorby-Adams,JenniferGuo,PabloLaso,JohnEKirsch,JuliaZabinska,Ana-LuciaGarciaGuarniz,PamelaWSchaefer,Seyedmehdi Payabvash,AdamdeHavenon,MatthewSRosen,etal. Portable,low-fieldmagneticresonanceimagingforevaluationofAlzheimer’sdisease. Nat. Commun., 15(1):10488, 2024. T. Liu et al.:Preprint Page 15 of 15

    work page 2024