update dataset README

README.md

@@ -20,47 +20,105 @@ configs:
 
 # ClinSeek-Bench
 
-ClinSeek-Bench provides the metadata files used to reconstruct the ClinSeekAgent
-evaluation benchmark from the original source datasets. This repository is not a
-fully materialized benchmark package: MIMIC-derived patient databases, source
-tables, chest X-ray JPG files, radiology reports, and rendered EHR contexts are
-protected assets and must be rebuilt locally under each user's own credentialed
-access and data-use agreements.
-
-| Task set | # data points | Modality | Released metadata file |
-| --- | ---: | --- | --- |
-| ClinSeek-Bench text tasks | 1,800 | text-only EHR | `ClinSeek-Bench_text.json` |
-| ClinSeek-Bench multimodal tasks | 989 | EHR + chest X-ray | `ClinSeek-Bench_multimodal.jsonl` |
-
-The ClinSeek-Bench multimodal tasks are restricted to image-grounded examples.
-They contain 497 EHRXQA-derived CXR question-answering rows and 492
-MedMod-derived ICU/CXR prediction rows.
+ClinSeek-Bench is the evaluation suite introduced in
+[ClinSeekAgent: Automating Multimodal Evidence Seeking for Agentic Clinical
+Reasoning](https://arxiv.org/abs/2605.20176). It evaluates clinical reasoning
+under two paired settings with the same task definitions and answer labels:
+
+- **Curated Input**: the model answers from the evidence package provided by
+  the source benchmark.
+- **Automated Evidence-Seeking**: the curated context is removed, and the model
+  must retrieve evidence from raw clinical data using ClinSeekAgent tools.
+
+This Hugging Face dataset release provides the metadata needed to reconstruct
+ClinSeek-Bench. It is not a fully materialized benchmark package.
+
+## Why This Release Contains Metadata Only
+
+ClinSeek-Bench is built from credentialed clinical datasets, including MIMIC-IV,
+MIMIC-IV-Note, MIMIC-IV-ED, MIMIC-CXR, MIMIC-CXR-JPG, EHRXQA, and MedMod. These
+sources contain protected clinical information and must be obtained under each
+user's own credentialed access and data-use agreements.
+
+For privacy and licensing reasons, this repository does **not** redistribute:
+
+- raw MIMIC tables;
+- patient-level SQLite databases;
+- rendered EHR contexts or `input_text`;
+- chest X-ray image files;
+- radiology report text;
+- experiment logs or model trajectories.
+
+Instead, this repository releases the metadata needed to rebuild the complete
+runtime benchmark locally from the official source datasets.
+
+## Released Metadata
+
+| Split | Released metadata | Rows | Modality | What it contains | What it excludes |
+| --- | --- | ---: | --- | --- | --- |
+| Text-only | `ClinSeek-Bench_text.json` | 1,800 | EHR | `qid`, `subject_id`, `hadm_id`, prediction-time metadata, task names, questions, labels, candidates | patient SQLite DBs and raw MIMIC rows |
+| Multimodal | `ClinSeek-Bench_multimodal.jsonl` | 989 | EHR + CXR | released `qid`s, questions, normalized labels, patient/time metadata, CXR image references, report references | CXR JPG files, report text, rendered EHR contexts, patient SQLite DBs |
+
+The metadata files are the source of truth for reconstruction and evaluation.
+During local reconstruction, the ClinSeekAgent scripts materialize protected
+runtime assets such as patient databases, rendered EHR contexts, linked CXR
+files, and radiology reports from the official source datasets.
+
+## Benchmark Composition
+
+### Text-only EHR tasks
+
+The text-only split is derived from EHR-Bench in
+[EHR-R1](https://github.com/MAGIC-AI4Med/EHR-R1). It contains 45 EHR subtasks
+covering risk-prediction and decision-making scenarios. We sample 40 examples
+per subtask, resulting in 1,800 text-only examples. The released metadata
+contains 1,563 unique patients.
+
+### Multimodal tasks
+
+The multimodal split adapts EHRXQA and MedMod, both grounded in MIMIC-IV EHRs
+and MIMIC-CXR chest radiographs. It contains 989 image-grounded examples:
+
+- 497 EHRXQA-derived CXR question-answering rows;
+- 492 MedMod-derived ICU/CXR prediction rows.
+
+The six multimodal task groups are CXR finding presence, CXR finding
+enumeration, CXR temporal change comparison, 24-hour decompensation prediction,
+in-hospital mortality prediction, and phenotype prediction.
 
 ## Files
 
 ```text
 ClinSeek-Bench/
-├── ClinSeek-Bench_text.json
-└── ClinSeek-Bench_multimodal.jsonl
++-- README.md
++-- ClinSeek-Bench_text.json
++-- ClinSeek-Bench_multimodal.jsonl
 ```
 
-`ClinSeek-Bench_text.json` stores the text benchmark records and the patient,
-admission, task, timing, question, label, and candidate metadata needed by the
-ClinSeekAgent preparation and evaluation scripts. It does not include the
-patient-level SQLite databases or raw MIMIC source tables.
+The paths above are the released metadata paths in this dataset repository. If
+you clone this dataset next to the ClinSeekAgent repository, reference the
+metadata as:
+
+```text
+ClinSeek-Bench/ClinSeek-Bench_text.json
+ClinSeek-Bench/ClinSeek-Bench_multimodal.jsonl
+```
 
-`ClinSeek-Bench_multimodal.jsonl` stores released `qid`s, questions, normalized
-labels, patient identifiers, prediction-time cutoffs, and relative CXR/report
-references for the multimodal split. It intentionally does not store rendered
-`input_text`, patient SQLite databases, CXR JPG files, or report text.
+If you use ClinSeekAgent's default local data layout, you may copy or symlink
+the text manifest to `data/text/ClinSeek-Bench_text.json`. That location is a
+local workspace convenience, not a released metadata path.
 
-## Reconstructing Benchmark Assets
+## Reconstructing The Full Benchmark
 
-Reconstruction scripts are maintained in the ClinSeekAgent GitHub repository:
+Reconstruction code and detailed instructions are maintained in the
+ClinSeekAgent GitHub repository:
 
-- text benchmark documentation: https://github.com/UCSC-VLAA/ClinSeekAgent/blob/main/docs/ClinSeek-Bench_text_data_prepare.md
-- documentation: https://github.com/UCSC-VLAA/ClinSeekAgent/blob/main/docs/ClinSeek-Bench_multimodal_data_prepare.md
-- scripts: https://github.com/UCSC-VLAA/ClinSeekAgent/tree/main/scripts/data_build
+- Code: https://github.com/UCSC-VLAA/ClinSeekAgent
+- Text data preparation: https://github.com/UCSC-VLAA/ClinSeekAgent/blob/main/docs/ClinSeek-Bench_text_data_prepare.md
+- Automated Evidence-Seeking text evaluation: https://github.com/UCSC-VLAA/ClinSeekAgent/blob/main/docs/ClinSeek-Bench_text_evaluation.md
+- Curated Input text evaluation: https://github.com/UCSC-VLAA/ClinSeekAgent/blob/main/docs/ClinSeek-Bench_text_curated_input_evaluation.md
+- Multimodal data preparation: https://github.com/UCSC-VLAA/ClinSeekAgent/blob/main/docs/ClinSeek-Bench_multimodal_data_prepare.md
+- Data build scripts: https://github.com/UCSC-VLAA/ClinSeekAgent/tree/main/scripts/data_build
 
 At a high level:
 
@@ -69,29 +127,29 @@ git clone https://huggingface.co/datasets/UCSC-VLAA/ClinSeek-Bench
 git clone https://github.com/UCSC-VLAA/ClinSeekAgent.git
 ```
 
-Then run the scripts under `ClinSeekAgent/scripts/data_build/` with local paths
-to the official source datasets.
+Then obtain the required official source datasets, prepare local paths, and
+follow the text and multimodal reconstruction guides above.
 
-Required source data:
+## Required Source Data
 
-| Dataset | Source |
-| --- | --- |
-| MIMIC-IV | https://physionet.org/content/mimiciv/ |
-| MIMIC-CXR | https://physionet.org/content/mimic-cxr/ |
-| MIMIC-CXR-JPG | https://physionet.org/content/mimic-cxr-jpg/ |
-| MIMIC-IV-Note | https://physionet.org/content/mimic-iv-note/ |
-| EHRXQA | https://physionet.org/content/ehrxqa/ |
-| MedMod code | https://github.com/nyuad-cai/MedMod |
+Download clinical source data only from official sources under your own
+credentialed access and data-use agreements.
 
-The final local reconstruction materializes the runtime benchmark assets needed
-for evaluation, including patient SQLite databases, table-description metadata,
-rendered EHR contexts, linked CXR JPGs, and radiology reports when available.
-The downloaded `ClinSeek-Bench_text.json` and `ClinSeek-Bench_multimodal.jsonl`
-remain the metadata sources of truth used to drive that reconstruction.
+| Dataset or source | Official source | Used for |
+| --- | --- | --- |
+| MIMIC-IV | https://physionet.org/content/mimiciv/ | structured EHR tables |
+| MIMIC-IV-Note | https://physionet.org/content/mimic-iv-note/ | discharge and radiology notes |
+| MIMIC-IV-ED | https://physionet.org/content/mimic-iv-ed/ | emergency department tables |
+| MIMIC-CXR | https://physionet.org/content/mimic-cxr/ | CXR metadata and report text |
+| MIMIC-CXR-JPG | https://physionet.org/content/mimic-cxr-jpg/ | linked chest X-ray JPG files |
+| EHRXQA | https://physionet.org/content/ehrxqa/ | source-aligned CXR question-answering rows |
+| MedMod | https://github.com/nyuad-cai/MedMod | source-aligned ICU/CXR prediction rows |
+| EHR-R1 / EHR-Bench | https://github.com/MAGIC-AI4Med/EHR-R1 | source benchmark for text-only EHR tasks |
 
 ## Validation
 
-After cloning ClinSeekAgent, validate the released multimodal metadata file:
+After cloning ClinSeekAgent, validate the released multimodal metadata without
+requiring protected assets:
 
 ```bash
 python ClinSeekAgent/scripts/data_build/validate_multimodal_release.py \
@@ -100,16 +158,34 @@ python ClinSeekAgent/scripts/data_build/validate_multimodal_release.py \
   --manifest-only
 ```
 
-Expected multimodal metadata counts:
+Expected released metadata counts:
 
-- 989 total multimodal rows.
-- 497 EHRXQA-derived rows.
-- 492 MedMod-derived rows.
-- 350 unique EHRXQA linked CXR JPG paths.
-- 477 unique MedMod linked CXR JPG paths.
+- 1,800 text-only rows;
+- 45 text-only EHR subtasks;
+- 1,563 unique text-only patients;
+- 989 multimodal rows;
+- 497 EHRXQA-derived multimodal rows;
+- 492 MedMod-derived multimodal rows;
+- 350 unique EHRXQA linked CXR JPG paths;
+- 477 unique MedMod linked CXR JPG paths;
 - 356 unique EHRXQA linked CXR report paths.
 
-## Citation
+After local reconstruction, the same validation script can also check the
+materialized runtime assets, including patient DBs, image paths, and report
+paths. See the ClinSeekAgent reconstruction documentation for the full commands.
+
+## Evaluation
+
+Evaluation code and instructions are maintained in the ClinSeekAgent repository:
+https://github.com/UCSC-VLAA/ClinSeekAgent.
+
+After reconstructing the protected runtime assets locally, follow the
+ClinSeekAgent documentation for Automated Evidence-Seeking evaluation, Curated
+Input evaluation, and scoring.
+
+## Responsible Use
 
-Please cite the ClinSeek paper and the upstream source benchmarks used for any
-reconstructed multimodal assets.
+ClinSeek-Bench is for research on clinical evidence seeking. It is not a
+medical device and must not be used for clinical diagnosis, treatment, triage,
+or patient management without separate validation, governance, and regulatory
+review.