Dataset Viewer
Auto-converted to Parquet Duplicate
report_id
stringclasses
8 values
patient_id
stringclasses
8 values
age_band
stringclasses
4 values
sex
stringclasses
2 values
collected_month
stringdate
2026-04-01 00:00:00
2026-09-01 00:00:00
test_name
stringlengths
3
22
value
float64
0.81
246
unit
stringclasses
8 values
ref_range
stringlengths
2
28
flag
stringclasses
3 values
method
stringclasses
8 values
source_format
stringclasses
2 values
consent_ref
stringclasses
5 values
LR-5001
pt_063dc1dc
45-49
M
2026-08
Vitamin D (25-OH)
17.9
ng/mL
30-100
low
CLIA
printed report (scanned)
src-0414 / v1
LR-5001
pt_063dc1dc
45-49
M
2026-08
TSH
3.11
mIU/L
0.4-4.0
normal
CLIA
printed report (scanned)
src-0414 / v1
LR-5001
pt_063dc1dc
45-49
M
2026-08
LDL cholesterol
178
mg/dL
<100
high
Direct enzymatic
printed report (scanned)
src-0414 / v1
LR-5001
pt_063dc1dc
45-49
M
2026-08
Platelet count
178
x10^9/L
150-450
normal
Automated analyser
printed report (scanned)
src-0414 / v1
LR-5002
pt_2c0aff4b
30-34
M
2026-09
HbA1c
8.5
%
4.0-5.6
high
HPLC
PDF report
src-0414 / v1
LR-5002
pt_2c0aff4b
30-34
M
2026-09
Fasting plasma glucose
85
mg/dL
70-99
normal
Hexokinase
PDF report
src-0414 / v1
LR-5002
pt_2c0aff4b
30-34
M
2026-09
TSH
10.96
mIU/L
0.4-4.0
high
CLIA
PDF report
src-0414 / v1
LR-5002
pt_2c0aff4b
30-34
M
2026-09
WBC count
14.9
x10^9/L
4.0-11.0
high
Automated analyser
PDF report
src-0414 / v1
LR-5003
pt_ce516d40
60-64
M
2026-08
Serum creatinine
0.81
mg/dL
0.6-1.2
normal
Jaffe
PDF report
src-0412 / v1
LR-5003
pt_ce516d40
60-64
M
2026-08
Fasting plasma glucose
187
mg/dL
70-99
high
Hexokinase
PDF report
src-0412 / v1
LR-5003
pt_ce516d40
60-64
M
2026-08
LDL cholesterol
180
mg/dL
<100
high
Direct enzymatic
PDF report
src-0412 / v1
LR-5003
pt_ce516d40
60-64
M
2026-08
ALT (SGPT)
94
U/L
7-56
high
IFCC
PDF report
src-0412 / v1
LR-5004
pt_7777460e
35-39
F
2026-06
TSH
6.71
mIU/L
0.4-4.0
high
CLIA
PDF report
src-0410 / v1
LR-5004
pt_7777460e
35-39
F
2026-06
WBC count
4.7
x10^9/L
4.0-11.0
normal
Automated analyser
PDF report
src-0410 / v1
LR-5004
pt_7777460e
35-39
F
2026-06
Platelet count
184
x10^9/L
150-450
normal
Automated analyser
PDF report
src-0410 / v1
LR-5004
pt_7777460e
35-39
F
2026-06
HbA1c
9.3
%
4.0-5.6
high
HPLC
PDF report
src-0410 / v1
LR-5005
pt_588d835d
60-64
F
2026-07
TSH
3.14
mIU/L
0.4-4.0
normal
CLIA
printed report (scanned)
src-0413 / v1
LR-5005
pt_588d835d
60-64
F
2026-07
C-reactive protein
40.5
mg/L
<5
high
Immunoturbidimetry
printed report (scanned)
src-0413 / v1
LR-5005
pt_588d835d
60-64
F
2026-07
Vitamin D (25-OH)
36.5
ng/mL
30-100
normal
CLIA
printed report (scanned)
src-0413 / v1
LR-5005
pt_588d835d
60-64
F
2026-07
Platelet count
246
x10^9/L
150-450
normal
Automated analyser
printed report (scanned)
src-0413 / v1
LR-5006
pt_715ba1f4
35-39
F
2026-04
Fasting plasma glucose
144
mg/dL
70-99
high
Hexokinase
PDF report
src-0410 / v1
LR-5006
pt_715ba1f4
35-39
F
2026-04
Serum creatinine
1.52
mg/dL
0.6-1.2
high
Jaffe
PDF report
src-0410 / v1
LR-5006
pt_715ba1f4
35-39
F
2026-04
HbA1c
5.1
%
4.0-5.6
normal
HPLC
PDF report
src-0410 / v1
LR-5006
pt_715ba1f4
35-39
F
2026-04
Haemoglobin
9.2
g/dL
12.0-15.5 (F), 13.5-17.5 (M)
low
Automated analyser
PDF report
src-0410 / v1
LR-5007
pt_f1be21f7
35-39
M
2026-07
TSH
8.7
mIU/L
0.4-4.0
high
CLIA
PDF report
src-0411 / v1
LR-5007
pt_f1be21f7
35-39
M
2026-07
Platelet count
139
x10^9/L
150-450
low
Automated analyser
PDF report
src-0411 / v1
LR-5007
pt_f1be21f7
35-39
M
2026-07
Vitamin D (25-OH)
22.2
ng/mL
30-100
low
CLIA
PDF report
src-0411 / v1
LR-5007
pt_f1be21f7
35-39
M
2026-07
Haemoglobin
15.7
g/dL
12.0-15.5 (F), 13.5-17.5 (M)
normal
Automated analyser
PDF report
src-0411 / v1
LR-5008
pt_42a6a985
45-49
F
2026-09
Serum creatinine
0.85
mg/dL
0.6-1.2
normal
Jaffe
PDF report
src-0411 / v1
LR-5008
pt_42a6a985
45-49
F
2026-09
Fasting plasma glucose
125
mg/dL
70-99
high
Hexokinase
PDF report
src-0411 / v1
LR-5008
pt_42a6a985
45-49
F
2026-09
Haemoglobin
10.6
g/dL
12.0-15.5 (F), 13.5-17.5 (M)
low
Automated analyser
PDF report
src-0411 / v1
LR-5008
pt_42a6a985
45-49
F
2026-09
WBC count
12.9
x10^9/L
4.0-11.0
high
Automated analyser
PDF report
src-0411 / v1

Diagnostic lab reports: sample

Structured lab results: test name, value, unit, reference range, high/low flag and method, one row per test. This file shows the schema and labels of the lab data that Datadhara licenses to AI teams.

What this is

  • 32 rows: 8 reports with 4 tests each, one row per test.
  • 11 tests: HbA1c, fasting plasma glucose, LDL cholesterol, TSH, vitamin D (25-OH), haemoglobin, WBC count, platelet count, serum creatinine, ALT (SGPT), C-reactive protein.
  • Each result has a flag (high, low or normal) against its reference range, and the assay method (for example HPLC, CLIA, Jaffe, IFCC).
  • Each row records whether it was taken from a PDF report or a scanned printed report.
  • Patients are identified only by a pseudonymous ID. Ages are five-year bands and dates are months.

The rows in this file were created to show the format, labels and de-identification. They do not come from real patients or labs, and the consent_ref values are format examples. Licensed deliveries contain real records collected under written agreements.

Fields

Column Type Description
report_id string Report identifier, e.g. LR-5001
patient_id string Pseudonymous patient ID, e.g. pt_063dc1dc
age_band string Five-year age band, e.g. 45-49
sex string F or M
collected_month string Sample collection month, YYYY-MM (exact date coarsened)
test_name string Test name as reported
value number Result value
unit string Unit, e.g. mg/dL, mIU/L, x10^9/L
ref_range string Reference range as printed, e.g. 4.0-5.6, <100, or sex-specific ranges
flag string high, low or normal
method string Assay or measurement method
source_format string PDF report or printed report (scanned)
consent_ref string Reference to the source agreement and its version, e.g. src-0414 / v1

Example row

{
  "report_id": "LR-5002",
  "patient_id": "pt_2c0aff4b",
  "age_band": "30-34",
  "sex": "M",
  "collected_month": "2026-09",
  "test_name": "HbA1c",
  "value": 8.5,
  "unit": "%",
  "ref_range": "4.0-5.6",
  "flag": "high",
  "method": "HPLC",
  "source_format": "PDF report",
  "consent_ref": "src-0414 / v1"
}

Load it

from datasets import load_dataset

ds = load_dataset("datadhara/lab-reports-sample", split="train")

Intended uses

Use this sample to check the schema and label set against your pipeline before requesting real data. The licensed data is used for:

  • Lab-result extraction and structuring from PDFs and scans
  • Clinical NLP over structured and semi-structured results
  • Evaluation sets for medical assistants

Limitations of this sample

  • Structured values only. The full data can include the source PDFs and scans; none are in this repository.
  • 8 reports is enough to see the format, not to train or benchmark anything.
  • Reference ranges are stored as printed strings, not parsed into numbers.

Full dataset and licensing

This sample file is released under CC BY 4.0, as on the samples page.

The full data is not on Hugging Face. It is available under a commercial licence. Datadhara sources lab reports, outpatient notes, prescriptions and discharge summaries directly from diagnostic labs, clinics and hospitals, as PDFs, scans and structured values. Test panels, document types and volume are agreed per project. Licences cover training and evaluating commercial models, and can be exclusive or non-exclusive.

To request data, describe what you need at datadhara.in/healthcare-data-for-ai or email hello@datadhara.in.

Consent and de-identification

As described in Datadhara's methodology:

  • Every source organisation signs a written agreement that explicitly covers use for AI training.
  • Patient names, record numbers, phone numbers, addresses and ID numbers are removed before data leaves the source. Exact dates are shifted or coarsened, and rare conditions in small clinics are reviewed so they can't single a patient out.
  • The identifier categories in the U.S. HIPAA Safe Harbor method are used as a baseline checklist, alongside the requirements of the law that applies to the source.
  • Every record carries a consent_ref linking it to its source agreement. In a real delivery, that reference resolves to the signed agreement.

Citation

The same sample files are archived on Zenodo: https://doi.org/10.5281/zenodo.24349905

Downloads last month
-