{"paper":{"title":"Multimodal Graph-based Classification of Esophageal Motility Disorders","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Multimodal graph neural networks that fuse esophageal pressure graphs with patient data improve classification of motility disorders over single-modality baselines.","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Alexander Geiger, Alissa Jell, Alois Knoll, Daniel Rueckert, Dirk Wilhelm, Lars Wagner","submitted_at":"2026-05-13T14:52:12Z","abstract_excerpt":"Diagnosing esophageal motility disorders pose significant challenges due to the complexity of high-resolution impedance manometry (HRIM) data and variability in clinical interpretation. This work explores the feasibility of a multimodal Machine Learning (ML)-based classification approach that combines HRIM recordings with patient-specific information and incorporates a graph-based modeling of esophageal physiology. We analyze HRIM recordings with corresponding patient information from 104 patients with esophageal motility disorders. Patient data includes demographic, clinical, and symptom info"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"The proposed multimodal approach indicates improvements over models that rely solely on HRIM-derived features across all classification categories. Additionally, the graph-based modeling provides gains compared to vision-based baselines.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The assumption that the spatio-temporal graph representation of HRIM recordings encodes physiologically meaningful features that, when fused with patient embeddings, lead to better multi-class classification, based on ablation studies whose details are not provided.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Graph-based multimodal ML model shows improved classification of esophageal motility disorders by fusing HRIM spatio-temporal graphs with patient embeddings.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Multimodal graph neural networks that fuse esophageal pressure graphs with patient data improve classification of motility disorders over single-modality baselines.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"399177d391b08d1f2c0862db93ee7dd26ff101d91dcfeb85fa4dd28089dc2c72"},"source":{"id":"2605.13623","kind":"arxiv","version":1},"verdict":{"id":"ea4834b4-5589-4eeb-ae3e-2e44812710f3","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T19:30:29.684280Z","strongest_claim":"The proposed multimodal approach indicates improvements over models that rely solely on HRIM-derived features across all classification categories. Additionally, the graph-based modeling provides gains compared to vision-based baselines.","one_line_summary":"Graph-based multimodal ML model shows improved classification of esophageal motility disorders by fusing HRIM spatio-temporal graphs with patient embeddings.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The assumption that the spatio-temporal graph representation of HRIM recordings encodes physiologically meaningful features that, when fused with patient embeddings, lead to better multi-class classification, based on ablation studies whose details are not provided.","pith_extraction_headline":"Multimodal graph neural networks that fuse esophageal pressure graphs with patient data improve classification of motility disorders over single-modality baselines."},"references":{"count":43,"sample":[{"doi":"","year":2024,"title":"Otolaryn- gologic Clinics of North America57(4) (2024)","work_id":"f4971d87-8708-4726-b1f4-4214de92416f","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2023,"title":"International Journal of Language & Communication Disorders58(2) (2023)","work_id":"d9da8df0-3398-4265-8e1a-9180add14d42","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2020,"title":"The Laryngoscope130(4) (2020)","work_id":"dcd8d89c-cf1d-42a8-b43f-a55f65b60b1b","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2008,"title":"Rheumatology47(6) (2008)","work_id":"34f8b606-a97a-4270-b13f-ca6fe0bd725d","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2020,"title":"American Family Physician102(5) (2020)","work_id":"2480e63d-21cb-4d6f-a3c9-9316f2306e7d","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":43,"snapshot_sha256":"edd42b3a8d8bcabdb7af188dd63a00cdff20434b134e209a445d02e80eebd08a","internal_anchors":1},"formal_canon":{"evidence_count":2,"snapshot_sha256":"231004d5c89acbe7ff50a05f35046747f3ef555ef448ef9799e577c5a73b38ee"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}