Med-Convo-Nig: Simulated Doctor-Patient Medical Conversations in Nigerian-Accented English

This work is licensed under a
Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

---

Overview

Med-Convo-Nig is a curated dataset of 25 simulated doctor-patient tele-consultations conducted in Nigerian-accented English. The dataset comprises approximately 4.2 hours of speech, with each conversation recorded as a single .wav file. The dialogues cover a range of medical scenarios commonly encountered in Nigeria, including cardiovascular conditions, reproductive health issues, and physical injuries.

All conversations were performed by Nigerian medical professionals and transcribed by the same. Each interaction includes a free-form conversation between a patient and a doctor, with no turn-by-turn speaker labeling.

Dataset Summary

	Summary
Number of conversations	25
Total duration	4.20 hours
Avg. conversation turns	99
Avg. word count per dialogue	1,220
Sample rate	16kHz mono
Gender distribution (F/M)	4 / 7
Speaker ages	25–35 years

Use Cases

• Evaluating ASR models on Nigerian medical speech
• Studying real-world phrasing in telemedicine consultations

Dataset Structure

Each row in the dataset includes:

• file_name: Path to the .wav audio file
• transcript: Full conversation transcript
• duration: Length of the audio in seconds
• medical_condition: Main medical issue discussed (e.g., hypertension, STI, physical injury)

Data Split

This dataset is a test-only set consisting of 25 complete conversations.

How to Load the Dataset

from datasets import load_dataset

dataset = load_dataset("intronhealth/med-convo-nig")