JMIR Publications

ABSTRACT

Background:

The integration of large language models (LLMs) into medicine has reshaped healthcare delivery, education, and research. Although proprietary models face challenges such as data privacy, regulation, and adaptability, DeepSeek, an open-source LLM, has emerged as a customizable and cost effective alternative with significant potential for clinical and operational applications. However, the rapid expansion of research in this area necessitates a systematic mapping of its landscape, applications, and challenges.

Objective:

This study combines bibliometric analysis with a scoping review to systematically map and characterize the literature on DeepSeek's medical applications. The aims were to: (1) analyze publication trends, leading contributors, and research themes and (2) identify primary application domains, strengths, limitations, and future directions.

Methods:

Following Arksey and O’Malley’s framework and the PRISMA-ScR guidelines, a systematic search was conducted using PubMed, Web of Science, and Scopus from January 20 to November 30, 2025. Bibliometric analysis was then used to quantify publication trends, productivity, and research themes across 371 papers. The scoping review thematically synthesized the applications, strengths, and limitations of 353 original articles.

Results:

The publication output showed strong polynomial growth (R² = 0.978), with China, Türkiye, and the United States as leading contributors. Keyword co-occurrence analysis identified clinical decision support, medical education, and patient communication as core keywords. DeepSeek has demonstrated competitive performance across multiple domains: patient education, clinical decision support, medical education, workflow optimization, and medical research. The challenges include variable performance across specialties, ethical risks (autonomy, justice, beneficence, and non-maleficence), data privacy, regulatory gaps, and the need for human oversight.

Conclusions:

DeepSeek has emerged as a high-performance, open-weight LLM with significant potential for enhancing clinical practice, education, and operational efficiency in medicine. Its on premises deployability and cost-effectiveness are notable advantages. However, challenges remain, including variable performance across specialties, model hallucinations, ethical concerns, and regulatory uncertainty. Future integration requires robust validation, multimodal capabilities, bias mitigation, human-in-the-loop governance, and equitable access strategies to ensure the safe, effective, and ethically aligned adoption in global health systems.