{"paper":{"title":"SFL-Net: Source-Factorized Latent Representation Learning for Multi-Contrast MRI to Tau-PET Synthesis","license":"http://creativecommons.org/licenses/by/4.0/","headline":"A disentangled quantized Half-UNet generates tau-PET images from T1-weighted and FLAIR MRI with top fidelity and Braak-stage accuracy.","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Agamdeep S. Chopra, Caitlin Neher, Hesamoddin Jahanian, Juampablo E. Heras Rivera, Mehmet Kurt, Tianyi Ren","submitted_at":"2026-02-26T02:37:38Z","abstract_excerpt":"Tau positron emission tomography supports Alzheimer's disease staging but is difficult to scale because of tracer, scanner, and radiation constraints. Synthesis from structural MRI is therefore attractive, but it is a particularly difficult setting. T1-weighted and FLAIR MRI provide anatomy and disease correlated morphology, but they do not directly measure Tau-PET relevant signal. We introduce SFL-Net, a multi-input synthesis framework that predicts Tau-PET from T1-weighted and FLAIR MRI. SFL-Net factorizes the latent representation into shared, T1-specific, FLAIR-specific, and complementary "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Across 17 evaluated models, the proposed DQ2H-MSE-Inf variant achieved the best raw PET fidelity and the strongest downstream Braak-stage performance, while remaining competitive on SUVR reconstruction and regional agreement.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the partial information decomposition components remain clinically meaningful and that the synthesized tau-PET does not introduce artifacts that would alter Braak staging or regional uptake interpretation in real patients.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"A Partial Information Decomposition-guided disentangled quantized Half-UNet synthesizes tau-PET from multimodal MRI and outperforms baselines on raw fidelity and Braak-stage tracking in ADNI-3 and OASIS-3 data.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A disentangled quantized Half-UNet generates tau-PET images from T1-weighted and FLAIR MRI with top fidelity and Braak-stage accuracy.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"f5de15659f7098aaf462555b3f510ca2a6b19df4c20f0c06b31a1aebbd7377c2"},"source":{"id":"2602.22545","kind":"arxiv","version":3},"verdict":{"id":"20001449-e406-486c-87a6-cc89f8c9ea03","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T19:33:53.718502Z","strongest_claim":"Across 17 evaluated models, the proposed DQ2H-MSE-Inf variant achieved the best raw PET fidelity and the strongest downstream Braak-stage performance, while remaining competitive on SUVR reconstruction and regional agreement.","one_line_summary":"A Partial Information Decomposition-guided disentangled quantized Half-UNet synthesizes tau-PET from multimodal MRI and outperforms baselines on raw fidelity and Braak-stage tracking in ADNI-3 and OASIS-3 data.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the partial information decomposition components remain clinically meaningful and that the synthesized tau-PET does not introduce artifacts that would alter Braak staging or regional uptake interpretation in real patients.","pith_extraction_headline":"A disentangled quantized Half-UNet generates tau-PET images from T1-weighted and FLAIR MRI with top fidelity and Braak-stage accuracy."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2602.22545/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}