Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,92 @@
|
|
| 1 |
-
---
|
| 2 |
-
|
| 3 |
-
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language:
|
| 3 |
+
- en
|
| 4 |
+
license: mit
|
| 5 |
+
pretty_name: Clinical Narrative–Biomarker Alignment Mapping v0.1
|
| 6 |
+
dataset_name: clinical-narrative-biomarker-alignment-mapping-v0.1
|
| 7 |
+
tags:
|
| 8 |
+
- clarusc64
|
| 9 |
+
- clinical
|
| 10 |
+
- patient-safety
|
| 11 |
+
- narrative
|
| 12 |
+
- biomarkers
|
| 13 |
+
task_categories:
|
| 14 |
+
- tabular-classification
|
| 15 |
+
- tabular-regression
|
| 16 |
+
size_categories:
|
| 17 |
+
- n<1K
|
| 18 |
+
configs:
|
| 19 |
+
- config_name: default
|
| 20 |
+
data_files:
|
| 21 |
+
- split: train
|
| 22 |
+
path: data/train/clinical-narrative-biomarker-alignment-mapping-train.csv
|
| 23 |
+
- split: test
|
| 24 |
+
path: data/test/clinical-narrative-biomarker-alignment-mapping-test.csv
|
| 25 |
+
---
|
| 26 |
+
|
| 27 |
+
Clinical Narrative–Biomarker Alignment Mapping v0.1
|
| 28 |
+
|
| 29 |
+
Goal
|
| 30 |
+
Given a patient narrative timeline and a biomarker timeline
|
| 31 |
+
judge whether they tell the same story.
|
| 32 |
+
|
| 33 |
+
This is not diagnosis prediction.
|
| 34 |
+
It is integrity checking across modalities.
|
| 35 |
+
|
| 36 |
+
What this dataset tests
|
| 37 |
+
- models that over-trust biomarkers when narratives lead
|
| 38 |
+
- models that over-trust narratives when objective deterioration exists
|
| 39 |
+
- models that fail to name divergence points
|
| 40 |
+
|
| 41 |
+
Required model outputs
|
| 42 |
+
- alignment_label
|
| 43 |
+
- aligned
|
| 44 |
+
- partial
|
| 45 |
+
- divergent
|
| 46 |
+
- alignment_score
|
| 47 |
+
- 0.00 to 1.00
|
| 48 |
+
- divergence_points
|
| 49 |
+
- short phrase naming where the split occurs
|
| 50 |
+
|
| 51 |
+
CSV schema
|
| 52 |
+
|
| 53 |
+
train columns
|
| 54 |
+
- id
|
| 55 |
+
- setting
|
| 56 |
+
- narrative_timeline
|
| 57 |
+
- biomarker_timeline
|
| 58 |
+
- clinical_context
|
| 59 |
+
- alignment_label
|
| 60 |
+
- alignment_score
|
| 61 |
+
- divergence_points
|
| 62 |
+
- why_it_matters
|
| 63 |
+
- constraints
|
| 64 |
+
|
| 65 |
+
test columns
|
| 66 |
+
- id
|
| 67 |
+
- setting
|
| 68 |
+
- narrative_timeline
|
| 69 |
+
- biomarker_timeline
|
| 70 |
+
- clinical_context
|
| 71 |
+
- notes
|
| 72 |
+
- constraints
|
| 73 |
+
|
| 74 |
+
Evaluation
|
| 75 |
+
- label accuracy
|
| 76 |
+
- RMSE on alignment_score
|
| 77 |
+
- divergence_points quality check
|
| 78 |
+
|
| 79 |
+
Use cases
|
| 80 |
+
- safety audits for clinical assistants
|
| 81 |
+
- triage support quality checks
|
| 82 |
+
- training clinicians on mismatch patterns
|
| 83 |
+
- bench tests for multimodal clinical models
|
| 84 |
+
|
| 85 |
+
Files
|
| 86 |
+
- data/train/clinical-narrative-biomarker-alignment-mapping-train.csv
|
| 87 |
+
- data/test/clinical-narrative-biomarker-alignment-mapping-test.csv
|
| 88 |
+
- tester.csv
|
| 89 |
+
- scorer.py
|
| 90 |
+
|
| 91 |
+
Path
|
| 92 |
+
ClarusC64/clinical-narrative-biomarker-alignment-mapping-v0.1
|