Large language models in radiology: fundamentals, applications, ethical considerations, risks, and future directions

被引：88

作者：

D'Antonoli, Tugba Akinci ^{[1
]}

Stanzione, Arnaldo ^{[2
]}

Bluethgen, Christian ^{[3
]}

Vernuccio, Federica ^{[4
]}

Ugga, Lorenzo ^{[2
]}

Klontzas, Michail E. ^{[5
,6
,7
]}

Cuocolo, Renato ^{[8
]}

Cannella, Roberto ^{[9
]}

Kocak, Burak ^{[10
]}

机构：

[1] Cantonal Hosp Baselland, Inst Radiol & Nucl Med, Liestal, Switzerland

[2] Univ Naples Federico II, Dept Adv Biomed Sci, Naples, Italy

[3] Univ Zurich, Univ Hosp Zurich, Inst Diagnost & Intervent Radiol, Zurich, Switzerland

[4] Univ Hosp Padova, Dept Radiol, Padua, Italy

[5] Univ Hosp Heraklion, Dept Med Imaging, Iraklion, Greece

[6] Univ Crete, Dept Radiol, Iraklion, Greece

[7] FORTH, Inst Comp Sci, Computat Biomed Lab, Iraklion, Greece

[8] Univ Salerno, Dept Med Surg & Dent, Baronissi, Italy

[9] Univ Palermo, Sect Neurosurg, Dept Biomed Neurosci & Adv Diagnost, Palermo, Italy

[10] Univ Hlth Sci, Basaksehir Cam & Sakura City Hosp, Clin Radiol, Istanbul, Turkiye

来源：

DIAGNOSTIC AND INTERVENTIONAL RADIOLOGY | 2024年 / 30卷 / 02期

关键词：

Large language models; natural language processing; artificial intelligence; ChatGPT; deep learning;

D O I：

10.4274/dir.2023.232417

中图分类号：

R8 [特种医学]; R445 [影像诊断学];

学科分类号：

1002 ; 100207 ; 1009 ;

摘要：

With the advent of large language models (LLMs), the artificial intelligence revolution in medicine and radiology is now more tangible than ever. Every day, an increasingly large number of articles are published that utilize LLMs in radiology. To adopt and safely implement this new technology in the field, radiologists should be familiar with its key concepts, understand at least the technical basics, and be aware of the potential risks and ethical considerations that come with it. In this review article, the authors provide an overview of the LLMs that might be relevant to the radiology community and include a brief discussion of their short history, technical basics, ChatGPT, prompt engineering, potential applications in medicine and radiology, advantages, disadvantages and risks, ethical and regulatory considerations, and future directions.

引用

页码：80 / 90

页数：11

共 96 条

[51]

Lu Y., 2021, CoRR

[52] Assessing the usefulness of a large language model to query and summarize unstructured medical notes in intensive care [J].

Madden, Michael G. ;

McNicholas, Bairbre A. ;

Laffey, John G. .

INTENSIVE CARE MEDICINE, 2023, 49 (08) :1018-1020

[53] Unstructured clinical notes within the 24 hours since admission predict short, mid & long-term mortality in adult ICU patients [J].

Mahbub, Maria ;

Srinivasan, Sudarshan ;

Danciu, Ioana ;

Peluso, Alina ;

Begoli, Edmon ;

Tamang, Suzanne ;

Peterson, Gregory D. .

PLOS ONE, 2022, 17 (01)

[54] Large language models for structured reporting in radiology: performance of GPT-4, ChatGPT-3.5, Perplexity and Bing [J].

Mallio, Carlo A. ;

Sertorio, Andrea C. ;

Bernetti, Caterina ;

Beomonte Zobel, Bruno .

RADIOLOGIA MEDICA, 2023, 128 (07) :808-812

[55] The imperative for regulatory oversight of large language models (or generative AI) in healthcare [J].

Mesko, Bertalan ;

Topol, Eric J. J. .

NPJ DIGITAL MEDICINE, 2023, 6 (01)

[56]

Mialon Gregoire, 2023, ARXIV

[57]

Mikolov T, 2010, 11TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION 2010 (INTERSPEECH 2010), VOLS 1-2, P1045

[58] Recent Advances in Natural Language Processing via Large Pre-trained Language Models: A Survey [J].

Min, Bonan ;

Ross, Hayley ;

Sulem, Elior ;

Ben Veyseh, Amir Pouran ;

Nguyen, Thien Huu ;

Sainz, Oscar ;

Agirre, Eneko ;

Heintz, Ilana ;

Roth, Dan .

ACM COMPUTING SURVEYS, 2024, 56 (02)

[59]

Mishra Gaurav, 2022, arXiv

[60]

Nicholson C., on Artificial Intelligence and Machine Learning

← 1 2 3 4 5 6 7 8 9 10 →