Files

PDF

HOW TO CITE

SAFEGUARDING AI-MEDIATED DIAGNOSES: A COMPREHENSIVE REVIEW OF CYBERSECURITY CHALLENGES AND SOLUTIONS IN LARGE LANGUAGE MODEL-ASSISTED MEDICAL APPLICATIONS
Authors (Affiliation): Param Ahir (NFSU), Mehul Parikh (Information Technology Department L. D. College of Engineering)
Abstract:

This extensive review discusses the connection between large language models (LLMs) and medical imaging, exploring into the complex realm of cybersecurity challenges and solutions in the context of AI-assisted diagnoses. The paper provides a critical review of key studies that address the vulnerabilities and opportunities associated with the integration of LLMs in the analysis and interpretation of textual data related to medical images. Utilising a variety of research methods, such as qualitative analyses, quantitative studies, and ethical evaluations, this review combines findings from each study to provide a comprehensive overview of the present state of cybersecurity in LLM-assisted medical imaging. The comprehensive literature review establishes a basis for examining various obstacles, including potential risks to data privacy, malicious attacks, ensuring data integrity, and safeguarding network security. The paper explores various aspects related to cybersecurity measures, such as encryption protocols, strategies for defending against adversaries, and ethical considerations about the use of AI in healthcare. In this review, we seek to make a valuable contribution to the ongoing discussion surrounding the future of medical imaging at a time when large language models are becoming more prevalent. 

Keywords: large language model, medical imaging, artificial intelligence, cyber security
Vol & Issue: VOL.2, ISSUE No.2, December 2023