{
  "abstract": "Introduction: To develop and evaluate a multimodal deep learning (DL) model for subtype classification of viral keratitis.\n\nMethods: This retrospective multicenter study included 3,501 slit-lamp images (807 eyes) as the internal development set and two external validation cohorts (2,957 images total). DeepFusion-HSK integrates three complementary illumination modalities—diffuse light imaging (DLI), slit-beam imaging (SBI), and fluorescein staining imaging (FSI)—through a novel fusion architecture. The framework uses a tri-way higher-order interaction layer to capture cross-modal pathological signatures, dual-task learning with auxiliary heads to prevent modality-specific feature degradation, and a dynamic gating mechanism that uses high-confidence coarse priors to calibrate fine-grained predictions and mitigate softmax probability spillover. To simulate real-world data deficiency, we evaluated model resilience under complete FSI dropout.\n\nResults: On the internal test set, DeepFusion-HSK achieved an accuracy of 85.1% (95% CI: 78.5%–90.0%) and AUCs of 0.947–0.949 for discriminating epithelial herpes simplex keratitis (E-HSK) and non-epithelial herpes simplex keratitis (NE-HSK), significantly outperforming Swin-Transformer, EfficientNetV2, and ResNet50. External validation confirmed robust generalization, with accuracies of 81.7% on Dataset I and 78.2% on Dataset II, demonstrating clinical resilience. Ablation studies confirmed the hierarchical synergy of multimodal inputs: the best single modality (FSI) yielded 78.5% accuracy, the optimal dual-modality combination (DLI + FSI) reached 81.7%, and full tri-modal fusion improved performance to 85.1%, validating that integrated global morphology, depth contour, and superficial ulcer signatures are essential for robust HSK subtyping.\n\nDiscussion: DeepFusion-HSK offers a robust, generalizable, and clinically deployable solution for therapeutic HSK subtyping. Its resilience to missing imaging modalities makes it well suited for heterogeneous real-world ophthalmic practice, providing a decision-support tool to reduce vision-threatening misdiagnoses and guide therapy.",
  "abstract_shown": true,
  "affiliations": [
    "National Clinical Research Center for Ocular Diseases, Eye Hospital, Cixi Biomedical Research Institute, Wenzhou Medical University , ,",
    "School of Electronic Engineering, Xi’an University of Posts and Telecommunications , ,"
  ],
  "article_type": "research article",
  "arxiv": null,
  "author_total": 12,
  "authors": [
    {
      "affs": [
        0
      ],
      "name": "Zuhui Zhang"
    },
    {
      "affs": [
        0
      ],
      "name": "Tao Lv"
    },
    {
      "affs": [
        0
      ],
      "name": "Jiahao Tang"
    },
    {
      "affs": [
        0
      ],
      "name": "Kunhui Xu"
    },
    {
      "affs": [
        0
      ],
      "name": "Fen-Fen Li"
    },
    {
      "affs": [
        0
      ],
      "name": "Xinxin Yu"
    },
    {
      "affs": [
        0
      ],
      "name": "Chenchen Wang"
    },
    {
      "affs": [
        0
      ],
      "name": "Shuangqing Wu"
    },
    {
      "affs": [
        0
      ],
      "name": "Ya-na Fu"
    },
    {
      "affs": [
        1
      ],
      "name": "Jie-Wei Jiang"
    },
    {
      "affs": [
        0
      ],
      "name": "Fan Lu"
    },
    {
      "affs": [
        0
      ],
      "name": "Qi Dai"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/8d5dvbdrq28k9.bib",
  "cited_by": 0,
  "doi": "10.3389/fcell.2026.1978214",
  "fulltext_url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13724611/?tool=EBI",
  "id": "8d5dvbdrq28k9",
  "issns": [
    "2296-634X"
  ],
  "landing_url": "https://europepmc.org/article/PMC/PMC13724611",
  "language": "eng",
  "licence": "CC-BY",
  "licence_label": "CC BY",
  "licence_source": "article",
  "licence_url": "https://creativecommons.org/licenses/by/4.0/",
  "metadata_licence": null,
  "native_subjects": [
    [
      "heading",
      "Original Research"
    ],
    [
      "openalex",
      "2731|Ocular Infections and Treatments"
    ]
  ],
  "openalex": "W7222982406",
  "pdf_url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13724611/pdf/?tool=EBI",
  "pmcid": "PMC13724611",
  "pmid": "42973666",
  "published": "2026-10-19",
  "publisher": "Frontiers Media SA",
  "ris": "https://research.draconitic.com/paper/8d5dvbdrq28k9.ris",
  "sources": [
    {
      "first_seen": "2026-11-03T15:20:51Z",
      "native_id": "PMC:PMC13724611",
      "source": "europepmc",
      "url": "https://europepmc.org/article/PMC/PMC13724611"
    },
    {
      "first_seen": "2026-11-04T12:23:09Z",
      "native_id": "PMC13724611",
      "source": "pmc",
      "url": "https://doi.org/10.3389/fcell.2026.1978214"
    },
    {
      "first_seen": "2026-11-04T15:20:48Z",
      "native_id": "MED:42973666",
      "source": "europepmc",
      "url": "https://europepmc.org/article/MED/42973666"
    }
  ],
  "subjects": [
    {
      "field": "medicine",
      "origin": "openalex:2731|Ocular Infections and Treatments",
      "subfield": "general"
    }
  ],
  "title": "DeepFusion-HSK: a multimodal deep learning for subtype classification of herpes simplex keratitis with multicenter validation",
  "title_text": "DeepFusion-HSK: a multimodal deep learning for subtype classification of herpes simplex keratitis with multicenter validation",
  "type": "peer-reviewed",
  "updated_at": "2026-11-05T03:21:01Z",
  "url": "https://research.draconitic.com/paper/8d5dvbdrq28k9",
  "venue": "Frontiers in cell and developmental biology",
  "versions": []
}