{"paper":{"title":"JI-ADF: Joint-Individual Learning with Adaptive Decision Fusion for Multimodal Skin Lesion Classification","license":"http://creativecommons.org/licenses/by/4.0/","headline":"JI-ADF integrates joint-individual learning and adaptive decision fusion for improved multimodal skin lesion classification.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Dat Cao, Hien Chu, Hien Kha, Minh Le, Nguyen Quoc Khanh Le, Phan Nguyen, Trang Pham","submitted_at":"2026-04-30T02:48:15Z","abstract_excerpt":"Skin lesion classification is essential for early dermatological diagnosis, yet many existing computer-aided systems rely primarily on dermoscopic images and underutilize the multimodal evidence routinely available in clinical practice. To address this gap, we propose \\textbf{JI-ADF}, a trimodal deep learning framework that integrates dermoscopic images, clinical photographs, and structured patient metadata for clinically grounded skin lesion classification. The proposed architecture combines joint multimodal representation learning with modality-specific auxiliary supervision and an adaptive "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"The proposed JI-ADF method demonstrates strong and well-balanced performance across lesion categories on the MILK10k benchmark, improving sensitivity and Dice score while maintaining high specificity and good calibration.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the MILK10k benchmark faithfully represents real-world clinical acquisition conditions and severe class imbalance, and that the observed improvements arise from the joint-individual learning and adaptive fusion rather than dataset-specific tuning or implementation details.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"JI-ADF fuses three modalities with adaptive decision fusion and a multimodal attention module to achieve balanced, well-calibrated performance on the imbalanced MILK10k skin lesion benchmark.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"JI-ADF integrates joint-individual learning and adaptive decision fusion for improved multimodal skin lesion classification.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"779a8483ed76c2dec6a0dea677f2a7a11df86439bf2dbe9881e8b80415df628e"},"source":{"id":"2604.27343","kind":"arxiv","version":2},"verdict":{"id":"7cf67d6a-9250-44b5-afaa-29f9a1f58a76","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-07T09:31:38.136986Z","strongest_claim":"The proposed JI-ADF method demonstrates strong and well-balanced performance across lesion categories on the MILK10k benchmark, improving sensitivity and Dice score while maintaining high specificity and good calibration.","one_line_summary":"JI-ADF fuses three modalities with adaptive decision fusion and a multimodal attention module to achieve balanced, well-calibrated performance on the imbalanced MILK10k skin lesion benchmark.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the MILK10k benchmark faithfully represents real-world clinical acquisition conditions and severe class imbalance, and that the observed improvements arise from the joint-individual learning and adaptive fusion rather than dataset-specific tuning or implementation details.","pith_extraction_headline":"JI-ADF integrates joint-individual learning and adaptive decision fusion for improved multimodal skin lesion classification."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.27343/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-20T22:39:37.051068Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T19:22:50.496392Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"26b5f34cb5e0b9e33b9cd8d84e6372fa4a6e70b9dfb809335149090736bb64ee"},"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"}