The future landscape of large language models in medicine

Abstract

Large language models (LLMs) are artificial intelligence (AI) tools specifically trained to process and generate text. LLMs attracted substantial public attention after OpenAI’s ChatGPT was made publicly available in November 2022. LLMs can often answer questions, summarize, paraphrase and translate text on a level that is nearly indistinguishable from human capabilities. The possibility to actively interact with models like ChatGPT makes LLMs attractive tools in various fields, including medicine. While these models have the potential to democratize medical knowledge and facilitate access to healthcare, they could equally distribute misinformation and exacerbate scientific misconduct due to a lack of accountability and transparency. In this article, we provide a systematic and comprehensive overview of the potentials and limitations of LLMs in clinical practice, medical research and medical education.

Evaluation of Vision LLMs GTP-4V and LLaVA for the Recognition of Features Characteristic of Melanoma

“…1 The classification of tumour stages depends on the diameter of the primary tumour (stage I: <2 cm, stage II: >2 cm), metastatic spread into regional lymph nodes (stage III), and distant metastasis (stage IV). 1 The 5-year survival is reported to be 50.6% for local disease, 35.4% for nodal disease, and 13.5% for distant disease. 1 Surgical therapy remains the basis of management for MCC including primary excision and locoregional evaluation by sentinel node biopsy (SNB).…”

“…To the Editor, Merkel cell carcinoma (MCC) is a rare neuroendocrine malignancy with a low incidence between 0.3/100000/year and 1.6/100000/year but with high mortality. 1 The classification of tumour stages depends on the diameter of the primary tumour (stage I: <2 cm, stage II: >2 cm), metastatic spread into regional lymph nodes (stage III), and distant metastasis (stage IV). 1 The 5-year survival is reported to be 50.6% for local disease, 35.4% for nodal disease, and 13.5% for distant disease.…”

“…The ability of LLMs to respond to diverse requests and questions, and process complex concepts combined with that of realtime user interaction translates to their potential incorporation within medicine. 1 GTP-4V (Generative Pre-Training Transformer-4th in series with Vision) and LLaVA (Large Language-and-Vison Assistant) are 2 broadly available, state-of-the-art multimodal LLMs with the additional ability to analyze image inputs. This enables them to solve new and increasingly complex tasks compared to language-only systems as these models possess novel abilities resulting from the intersection of language and vision modalities in addition to the intelligence and reasoning abilities of these large scale models.…”

“…The ability of LLMs to respond to diverse requests and questions, and process complex concepts combined with that of real-time user interaction translates to their potential incorporation within medicine. 1…”

“…To the Editor, Large language models (LLMs) are artificial intelligence (AI) tools trained to process and generate text at a standard that resembles and is nearly indistinguishable from human capabilities. 1 Competencies of LLMs include the capacity to answer questions, summarize, paraphrase, and translate text following the input of several keywords or queries. The ability of LLMs to respond to diverse requests and questions, and process complex concepts combined with that of realtime user interaction translates to their potential incorporation within medicine.…”