Kaphathy/Dataset

MM-OphBench: Multi-Center Multimodal Clinical Ophthalmic Benchmark Dataset

<div align="center">

A Large-Scale, Standardized Multi-Center Benchmark Covering 7 Imaging Modalities & 4.3M+ Clinical Records

  ![Modalities]() ![Images]()

</div>

1. Executive Summary & Repository Overview

The MM-OphBench repository hosts a petabyte-scale, clinically harmonized ophthalmic image archive compiled from leading ophthalmic hospitals and benchmark cohorts. It spans 4,307,415 high-resolution diagnostic images and multimodal pairs across 7 distinct imaging modalities, with comprehensive clinical metadata, expert lesion segmentations, and hierarchical diagnostic taxonomies.

All archives are compressed into high-speed Zstandard chunks (.tar.zst) with pre-indexed SHA-256 manifests. A centralized, fully de-identified master index is provided in metadata/adapter_manifest_reviewed.parquet (39.9 MB), allowing instant cross-cohort querying across imaging modalities, device models, scan protocols, and clinical diagnoses without requiring massive raw downloads.

2. Multi-Center Clinical Provenance & Ethical Governance

All patient identifiable information (PII) including patient names, hospital record numbers (MRNs), national identity codes, and date-of-birth timestamps have been removed or replaced with deterministic SHA-256 cryptographic hashes (patient_hash, study_id), strictly complying with HIPAA Safe Harbor and GDPR pseudonymization principles.

Key Contributing Clinical Centers & Cohorts

1. 1.The Second Affiliated Hospital of Zhejiang University School of Medicine (浙二眼科)
2. 2.Contributions: Longitudinal FFA/ICGA dynamic angiography sequences (Final_eryuan_FFA), 17-class retinal disease benchmark (eryuanFundusT17RDCls), anterior segment photography.
3. 3.Clinical Leads: Certified ophthalmologists (Prof. Wang Guoping and retinal angiography specialist team).
4. 4.Beijing Tongren Hospital, Capital Medical University (北京同仁医院)
5. 5.Contributions: Longitudinal multi-condition clinical cohorts (tongren_archives), 66-class fine-grained disease screening benchmark (trhd3_FRD_Cls66_Train), and the standard 95-disease clinical taxonomy (tongren95_taxonomy.csv).
6. 6.The Second Xiangya Hospital of Central South University (中南大学湘雅二医院)
7. 7.Contributions: High-frequency ocular B-scan ultrasonography with detailed diagnostic text reports (BUltrasound).
8. 8.InEye Hospital of Chengdu University of TCM (成都中医药大学附属银海眼科医院)
9. 9.Contributions: Comprehensive Ultra-Widefield (UWF) Optos color imaging (yinhai_uwf, 274.6 GB), FFA dynamic sequences, and ocular B-scans.
10. 10.Chengdu First People's Hospital (成都市第一人民医院)
11. 11.Contributions: 50MHz Ultrasound Biomicroscopy (UBM) ciliary body and anterior chamber angle quantitative morphometry (ChengduShiyiUBM).
12. 12.Handan Eye Hospital (河北邯郸眼科医院)
13. 13.Contributions: Fluorescein angiography paired cohorts (handan_ffa, handan_ffa2) and epidemiological diabetic retinopathy screenings.
14. 14.Zhongshan Hospital, Fudan University (复旦大学附属中山医院)
15. 15.Contributions: Multimodal cohort comprising 419,014 clinical visits with cross-system validation.

3. Optical Modalities & Hardware Instrumentation

The dataset captures full optical, acoustic, and angiography representations across the anterior and posterior segments of the human eye:

4. Archive Directory & Shard Inventory

Archives are organized into self-contained sub-directories chunked at ~1024 MiB boundaries. Each shard is compressed using zstd -3 for fast decompressibility and high compression efficiency:

5. Master Metadata (metadata/) & Clinical Crosswalks

Instead of downloading multi-hundred-gigabyte archives to inspect sample attributes, download the lightweight metadata and clinical crosswalk files in metadata/:

1. metadata/adapter_manifest_reviewed.parquet (39.9 MB)

Contains 4,307,415 rows mapping every asset with standard schema:

import duckdb

con = duckdb.connect()
# Query device distribution across modalities
con.execute("""
    SELECT modality_subtype, device_vendor, device_model, count(*) AS count
    FROM 'metadata/adapter_manifest_reviewed.parquet'
    GROUP BY 1, 2, 3
    ORDER BY count DESC
    LIMIT 10
""").df()

2. metadata/trhd66_clinical_abbreviations.csv (25 KB)

Full 66-class fine-grained disease taxonomy codebook from Beijing Tongren & Handan Eye Hospital (trhd3_FRD_Cls66_Train). Includes abbreviation code (class_code), standardized Chinese name (official_zh), canonical English term (canonical_en), anatomical category (category), clinical manifestation, and vital disambiguation rules.

3. metadata/ophthalmic_abbreviations_crosswalk.csv (27 KB)

Unified multi-cohort crosswalk indexing 135+ ophthalmic abbreviations and disease codes across trhd3_FRD_Cls66_Train, eryuanFundusT17RDCls, RFMiD1, ODIR-5K, and multimodal diagnostic biomarkers (CFP, UWF, OCT, OCTA, FFA, UBM, B-Scan, RNFL, CDR).

4. metadata/tongren95_taxonomy.csv (12 KB)

Comprehensive 95-class disease taxonomy mapping local clinical diagnoses (raw_zh) to standardized English terms (canonical_en), abbreviations (AMD, PDR, RVO, CSC), disease families (family_en), and specificity levels.

5. metadata/trhd59_taxonomy.csv (5.5 KB)

59-class baseline model taxonomy with strict semantic relations mapped to Tongren-95 hierarchical leaves and clinical notes.

6. metadata/labeled_assets_summary.json (2.0 KB)

Complete breakdown of 1,817,159 instruction-tuned samples and 1,459,879 downstream high-confidence diagnostic splits.

6. Clinical Abbreviations & Diagnostic Codebook (临床病种缩写与全称速查手册)

Many ophthalmic datasets adopt hospital-specific, short-hand abbreviations that can cause severe diagnostic ambiguity if misinterpreted. Below is the authoritative clinical dictionary aligned with the codebase (class_labels_zh.py, trhd59_taxonomy.csv, and trhd66_clinical_abbreviations.csv).

A. Critical Clinical Disambiguation & Avoidance of Misnomers (核心消歧与避坑指南)

[!IMPORTANT] Key Disambiguation Rules: 1. `PMM_CNV` = 特发性息肉状脉络膜血管病变 (Polypoidal Choroidal Vasculopathy / PCV) Misconception: Frequently mistaken for "Pathologic Myopia with CNV". Ground Truth: In this cohort, PMM_CNV strictly represents PCV. True myopic CNV is separately encoded as `PM_CNV` (高度近视性脉络膜新生血管). 2. `COC` = 先天性大视杯 (Congenital Large Optic Cup) Misconception: Mistaken for Choroidal Osteoma (脉络膜骨瘤) or glaucomatous cupping. Ground Truth: Codebase explicitly documents: "COC must not be interpreted as choroidal osteoma". It represents physiological large optic cup with normal, pink neuroretinal rim obeying the ISNT rule. 3. `TC` vs `TR`: 外伤性脉络膜视网膜病变 vs 外伤性视网膜病变 Misconception: TC mistaken for Toxocariasis (眼弓蛔虫病). Ground Truth: In this project, TC = Traumatic Chorioretinopathy (外伤性脉络膜视网膜病变, deeper choroidal involvement with pigmentary scarring); TR = Traumatic Retinopathy (外伤性视网膜病变, retinal contusion/Purtscher-like cotton wool spots). 4. `EMM` vs `ERM`: 黄斑膜前膜 vs 视网膜前膜 Ground Truth: EMM is strictly restricted to the foveal/macular area (Epiretinal Macular Membrane), whereas ERM represents generalized Epiretinal Membrane across the posterior pole. 5. `SOD` = 小视盘 / 视隔-视神经发育不良 (Small Optic Disc / Septo-Optic Dysplasia) Ground Truth: Represents severe optic nerve hypoplasia, linked with de Morsier syndrome (midline septum pellucidum absence). 6. `PR` = 视网膜静脉周围炎 (Periphlebitis Retinae / Eales Disease) Misconception: Mistaken for Retinal Detachment or prolapse. Ground Truth: Represents vascular periphlebitis with characteristic perivenous sheathing and recurrent vitreous hemorrhages in young adults. 7. `SO eye` = 硅油填充眼 (Silicone Oil Filled Eye) Misconception: Mistaken for Sympathetic Ophthalmia. Ground Truth: Iatrogenic post-vitrectomy status showing hyper-reflective silicone oil reflex and meniscus droplets. 8. `PLS` = 激光光凝斑 (Retinal Laser Spots), `PM` = 视盘前膜 (Prepapillary Membrane), and `VO` = 玻璃体混浊 (Vitreous Opacity): None of these represent Pathological Myopia or Vein Occlusion.

B. 66-Class Fine-Grained Retinal Disease Cohort (trhd3_FRD_Cls66_Train)

C. Multi-Cohort Abbreviation Crosswalk Summary

1. 17-Class Retinal Disease Cohort (archives/eryuanFundusT17RDCls/)

• —`CNV`: 脉络膜新生血管 (Choroidal Neovascularization)
• —`ION`: 缺血性视神经病变 (Ischemic Optic Neuropathy)
• —`ME`: 黄斑水肿 (Macular Edema)
• —`OA`: 视神经萎缩 (Optic Atrophy)
• —`OpticNP`: 其他视神经病变 (Optic Neuropathy)
• —`OpticNR`: 视神经炎 (Optic Neuritis)
• —`PapEdema`: 视盘水肿 (Papilledema)
• —`VH`: 玻璃体积血 (Vitreous Hemorrhage)
• —`VKH`: 小柳原田病 (Vogt-Koyanagi-Harada Disease)
• —`AMD` / `DR` / `RVO` / `RAO` / `RD` / `RP` / `CSC` / `ERM` / `MH`: Standard canonical terms matching the 66-class definitions above.

2. RFMiD Multi-Label Dataset (archives/RFMiD1/)

• —`DN`: 玻璃膜疣 (Drusen)
• —`TSLN`: 豹纹状眼底 (Tessellated Fundus)
• —`LS`: 激光斑 (Laser Scars)
• —`MS`: 黄斑区瘢痕 (Macular Scar)
• —`CSR`: 中心性浆液性脉络膜视网膜病变 (Central Serous Chorioretinopathy / CSC)
• —`ODP`: 视盘小凹 (Optic Disc Pit)
• —`ODE`: 视盘水肿 (Optic Disc Edema)
• —`ST`: 后巩膜葡萄肿 (Posterior Staphyloma)
• —`AION`: 前部缺血性视神经病变 (Anterior Ischemic Optic Neuropathy)
• —`PT`: 中心凹旁毛细血管扩张症 (Parafoveal Telangiectasia)
• —`CRS`: 脉络膜视网膜炎 (Chorioretinitis)
• —`EDN`: 视网膜渗出病灶 (Exudation)
• —`CM` / `MC`: 缺损 (Coloboma) / 黄斑缺损 (Macular Coloboma)
• —`PRH`: 视网膜前出血 (Preretinal Hemorrhage)
• —`CWS`: 棉絮斑 (Cotton Wool Spots)
• —`CME`: 黄斑囊样水肿 (Cystoid Macular Edema)
• —`CF`: 脉络膜皱褶 (Choroidal Folds)
• —`VS`: 视网膜血管炎 (Retinal Vasculitis)

3. ODIR-5K Single-Letter Diagnostic Codes (archives/ODIR/)

• —`N`: Normal Fundus (正常眼底)
• —`D`: Diabetic Retinopathy (糖尿病视网膜病变)
• —`G`: Glaucoma (青光眼)
• —`C`: Cataract (白内障)
• —`A`: Age-Related Macular Degeneration (年龄相关性黄斑变性)
• —`H`: Hypertensive Retinopathy (高血压性视网膜病变)
• —`M`: Pathological Myopia (病理性近视)
• —`O`: Other Abnormalities / Diseases (其他异常及眼底疾病)

4. Multimodal Instrumentation & Clinical Parameters

• —`CFP`: 彩色眼底照相 (Color Fundus Photography, 30°~50°)
• —`UWF`: 超广角眼底照相 (Ultra-Widefield Fundus Photography, 200° Optos)
• —`OCT`: 光学相干断层扫描 (Optical Coherence Tomography B-Scans)
• —`OCTA`: 光学相干断层扫描血管成像 (OCT Angiography En-face Slabs)
• —`FFA`: 眼底荧光素血管造影 (Fundus Fluorescein Angiography Sequences)
• —`UBM`: 超声生物显微镜 (50MHz High-Frequency Anterior Segment Ultrasound)
• —`B-Scan`: 眼科B型超声 (10MHz Ocular Ultrasound Sweeps)
• —`C/D (CDR)`: 杯盘比 / 垂直杯盘比 (Cup-to-Disc Ratio / vCDR)
• —`RNFL`: 视网膜神经纤维层 (Retinal Nerve Fiber Layer)
• —`PED`: 色素上皮脱离 (Pigment Epithelial Detachment)
• —`FAZ`: 中心凹无血管区 (Foveal Avascular Zone)

7. Benchmark Evaluation Tasks

1. 1.17-Class Retinal Disease Classification (`eryuanFundusT17RDCls`): Standardized 17-class classification for common and sight-threatening retinal conditions (Normal, DR, AMD, RVO, Pathological Myopia, Macular Hole, Epiretinal Membrane, Retinal Detachment, Retinitis Pigmentosa, etc.).
2. 2.66-Class Multi-Center Fine-Grained Diagnosis (`trhd3_FRD_Cls66_Train`): Challenging 66-class classification covering rare and subtle ophthalmic pathologies validated across Beijing Tongren and Handan populations.
3. 3.95-Class Clinical Semantic Classification (`tongren16G95DFundus`, `tongren95_taxonomy`): Full-spectrum hospital diagnostic classification aligning free-text clinical reports and fundus photographs across 16 major disease families and 95 clinical subcategories.
4. 4.Dynamic FFA Vessel & Microvascular Leakage Segmentation (`Final_eryuan_FFA`, `handan_ffa`): Temporal sequence segmentation across early, arteriovenous, and late phases.
5. 5.Quantitative Anterior Chamber Morphometry & Angle-Closure Risk (UBM): Automated measurement of Anterior Chamber Depth (ACD in mm), Trabecular-Iris Angle (TIA), and angle-closure risk assessment across ChengduShiyiUBM, YinhaiUBM, and UBM_SW3200L.
6. 6.Ophthalmic Image Quality & Pre-Filtering Assessment (`FQ_Datasets`): Automated detection of non-diagnostic fundus images, lens opacity, uneven illumination, and motion artifacts for reliable clinical pipeline pre-filtering.
7. 7.Dual-View Diabetic Retinopathy Grading (`deepdrid`): Multi-field consistency and lesion grading combining macular and optic disc perspectives.
8. 8.Multi-Center Glaucoma Screening & Cup-to-Disc Ratio Estimation (`kaggle_glaucoma_v4`, `DGOCF`): Optic nerve head segmentation, vertical cup-to-disc ratio (vCDR) calculation, and multi-device glaucoma detection.

8. Quick Start: Extraction & Usage Guide

A. Download & Extract Archives (Linux / macOS)

To extract multi-part .tar.zst archives with maximum parallelism:

# Install zstd if not present
sudo apt-get install -y zstd  # Ubuntu/Debian
brew install zstd             # macOS

# Extract all shards for a specific dataset (e.g., BUltrasound)
cat archives/BUltrasound/BUltrasound-*.tar.zst | tar -I zstd -xvf - -C /path/to/destination/

B. Python Fast Metadata Query

import pandas as pd

# Read metadata directly from Hugging Face or local path
df = pd.read_parquet("metadata/adapter_manifest_reviewed.parquet", 
                     columns=["image_id", "modality_subtype", "device_model", "hospital_domain"])

# Filter for SS-OCT B-scans from Topcon Triton
topcon_oct = df[(df["modality_subtype"] == "oct_bscan") & (df["device_model"] == "DRI OCT Triton")]
print(f"Loaded {len(topcon_oct)} Topcon SS-OCT B-scans")

9. Citation & Acknowledgments

If you use this benchmark or any subsets in your research, please cite:

@dataset{mm_ophbench2026,
  title={MM-OphBench: A Multi-Center Multimodal Clinical Ophthalmic Benchmark Dataset},
  author={Zhejiang University, Beijing Tongren Hospital, InEye Hospital, Xiangya Hospital, and Contributors},
  year={2026},
  publisher={Hugging Face},
  howpublished={\url{https://huggingface.co/datasets/Kaphathy/Dataset}}
}

10. License & Terms of Use

• —The aggregated metadata and clinical annotations are made available under Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0).
• —Research use only. Not for direct diagnostic or medical intervention without regulatory clearance.