CoolFace
Datasetpublic

Chinapa/endoscopy_dataset

-- license: unknown language: - en pretty_name: Endoscopy Bowel Prep Risk Dataset (Synthetic) task_categories: - text-classification tags: - synthetic - tabular - endoscopy - nursing - bowel-prep - risk-prediction size_categories: - <1K Endoscopy Bowel Prep Risk Dataset (Synthetic) 📌 Dataset Summary This is a synthetic dataset created for research and demonstration in endoscopy nursing care.It is designed to support AI-assisted… See the full description on the dataset page: https://huggingface.co/datasets/Chinapa/endoscopy_dataset.

sourceHugging Faceupdated 1y agoView on Hugging Face
0likes2downloads
Dataset Card

-- license: unknown language:

  • en prettyname: Endoscopy Bowel Prep Risk Dataset (Synthetic) taskcategories:
  • text-classification tags:
  • synthetic
  • tabular
  • endoscopy
  • nursing
  • bowel-prep
  • risk-prediction size_categories:
  • <1K ---

Endoscopy Bowel Prep Risk Dataset (Synthetic)

📌 Dataset Summary

This is a synthetic dataset created for research and demonstration in endoscopy nursing care. It is designed to support AI-assisted risk prediction and nurse-led coaching systems for bowel preparation before colonoscopy.

⚠️ Note: This dataset does not contain real patient data. It is for educational and research demonstration purposes only.


📊 Data Fields

  • age: Patient age (years, int)
  • sex: Sex (M or F)
  • bmi: Body Mass Index (float)
  • dm: Diabetes mellitus (0/1)
  • ckd: Chronic Kidney Disease (0/1)
  • osa: Obstructive Sleep Apnea (0/1)
  • constipated: Chronic constipation (0/1)
  • iron: Iron supplementation (0/1)
  • opioid: Opioid use (0/1)
  • prior_inadequate: History of inadequate prep (0/1)
  • understands_splitdose: Patient understanding of split-dose regimen (0/1)
  • fiber48h: Fiber intake in last 48 hours (low, normal, high)
  • bbps_total: Boston Bowel Preparation Scale score (0–9)
  • inadequate: Binary target label (1 = inadequate, 0 = adequate)

✅ Intended Uses

  • Building AI-assisted pre-visit risk calculators
  • Supporting nurse-led interventions (personalized coaching, reminders)
  • Demonstration in digital health education and nursing informatics research

⚠️ Limitations

  • This dataset is synthetic, not derived from real patient data.
  • Not suitable for clinical decision-making.
  • Any deployment in real-world healthcare must use validated datasets under ethical approval (IRB).

📊 Example: BBPS Score and Adequacy

The Boston Bowel Preparation Scale (BBPS) evaluates bowel cleanliness in three segments of the colon (right, transverse, left), each scored 0–3. The total score ranges from 0–9.

BBPS ScoreInterpretationAdequacy label
0–1Unprepared / solid stoolInadequate (1)
2–4PoorInadequate (1)
5–6Fair / SuboptimalBorderline (often marked Inadequate=1)
7–9Good / ExcellentAdequate (0)

Target column in this dataset:

  • inadequate = 1 → BBPS below adequacy threshold
  • inadequate = 0 → BBPS adequate (usually ≥ 7)

🧾 Example Records

agesexbmidmckdosaconstipatedironopioidprior_inadequateunderstands_splitdosefiber48hbbps_totalinadequate
63F2810011010high41
52M24.500000001normal80

📚 Citation

If you use this dataset, please cite:

Chinapa M., Endoscopy Nursing Research Team (2025).  
Synthetic Endoscopy Bowel Preparation Risk Dataset. Hugging Face Datasets.  

🇹🇭 (สรุปภาษาไทย)

ชุดข้อมูลนี้เป็น ข้อมูลจำลอง (Synthetic) เพื่อใช้ในการเรียนการสอนและสาธิตงานวิจัย เกี่ยวกับ การพยาบาลส่องกล้องทางเดินอาหาร และการประเมินความเสี่ยงการเตรียมลำไส้ก่อนส่องกล้อง

  • ไม่ใช่ข้อมูลผู้ป่วยจริง
  • ใช้เพื่อการศึกษาและวิจัยต้นแบบเท่านั้น
  • ไม่ควรใช้ในการตัดสินใจทางคลินิกโดยตรง