Pediatric Device Clinical Interview Transcripts and AI-Extracted Requirements

Data Sheet 1_A framework for generative AI-driven extraction of clinical user needs in pediatric device development.pdf is a 271.1 KB PDF file authored by Abdelrahman Abdou and last updated on 2026-04-14. The dataset contains transcripts from 29 healthcare practitioner interviews, totaling 26 hours of data, used to extract user requirements for an infant heart rate detector. The study compares manual annotations with GPT-4o extractions to evaluate GenAI's role in summarizing clinical needs.

Use Cases

• Benchmarking GenAI extraction accuracy based on comparison with manual annotations mentioned in the description
• Training NLP models for clinical requirement extraction based on interview themes like ease of use and accuracy
• Analyzing practitioner demographics and experience based on the quantitative analyses of years of experience and specialties described
• Developing frameworks for medical device design specification based on extracted themes such as reusability, display, and cost

Strengths

• Includes 26 hours of transcribed interview data from 29 healthcare practitioners
• Provides a direct comparison between GPT-4o extractions and manual ground-truth annotations
• User requirements are curated under specific themes like ease of use, accuracy, and cost

Limitations

• Row count and column-level documentation are absent; field semantics must be inferred after download
• The dataset is a 271.1 KB PDF, indicating limited scope and likely containing summarized results rather than raw data
• Description metadata is limited; actual data quality requires manual inspection after download

Provenance

Source

figshare, author Abdelrahman Abdou

Collection Method

29 semistructured interviews with healthcare practitioners, recorded and transcribed

Time Range

null

Freshness

Last updated 2026-04-14 05:32:11; freshness should be verified

Geography

null