Contributed Commentary by Rakesh Nagarajan, MD, Chief Medical Officer at Velsera 

February 14, 2025 | The healthcare industry has experienced tremendous growth in the use of artificial intelligence (AI) and large language models (LLMs) over the past decade, both of which are already impacting how precision medicine is applied. AI plays a key role in identifying patterns within complex datasets, improving accuracy, and facilitating the exploration of vast volumes of clinical and ‘omics-based research data. Meanwhile, LLMs, with their ability to analyze and generate human-like text, hold significant promise for compiling medical knowledge, interpreting research papers, and making predictions based on clinical data. This allows researchers to engage with the system conversationally and begin the analytical process.  

However, these technologies reach their full potential only when integrated into a cohesive, standardized platform. This platform should process raw data or signals to identify discrete analytes, such as genetic variants, expression profiles, and protein levels, then extract insights using advanced AI algorithms and annotations, including those derived by LLMs.  

By integrating these technologies, researchers and clinicians can efficiently analyze large datasets, uncover disease mechanisms, and recognize trends that might otherwise go unnoticed. This approach not only drives research forward but also accelerates the development of new treatments, significantly improving patient care. 

Protecting Assets: Data Security and Privacy 

The integration of AI and LLMs in clinical settings brings data security and patient privacy to the forefront of discussion. A report in 2024 revealed over 300 cyberattacks on healthcare systems, highlighting the inherent risks when handling sensitive medical data. To mitigate these challenges, healthcare organizations must implement robust data security protocols, including: 

• Establishing Private Networks: Researchers and precision medicine companies should prioritize the establishment of dedicated, private cloud networks, ensuring sensitive data is processed, stored, and accessed securely with complete control over the environment. 

• Anonymizing Patient Records: Ensuring private information is removed from medical data before its use in AI systems prevents reidentification, allowing AI to analyze data, identify patterns, and produce predictions without compromising patient privacy or safety. 

• Establishing Private Environments for AI/LLMs: Precision medicine companies should ensure proprietary models are able to be version-controlled, with local copies preventing unauthorized data use. These models must operate within isolated environments with strict validation to ensure accuracy and reliability. 

Overcoming Challenges in AI and LLM Integration 

Although the promise of AI and LLMs is undeniable, their integration into healthcare presents several challenges that must be addressed: 

• Data Quality and Integration: Curated datasets, such as annotation databases and processed multimodal data, in conjunction with unstructured data like clinical notes or research papers, are crucial for ensuring the accuracy of LLM-driven insights. To get accurate results, a wide range of analytical tools, including AI algorithms, LLMs, clustering/classification methods, biostatistics, and advanced computational techniques, must be carefully used in the right research or clinical settings to guarantee high specificity and sensitivity. 

• Digital Hallucinations: Hallucinations where LLMs produce incorrect information can be a serious issue in healthcare. "Fit-for-purpose" models, fine-tuned with curated data or guided by prior knowledge, improve accuracy and reduce errors by focusing on specific tasks. 

• Model Transparency: Tailored models that offer traceable evidence behind generated insights allow professionals to correlate results back to their original references, ensuring more reliable verification. This streamlined fact-checking process is more efficient and reliable than traditional methods. 

Human Touch and its Importance 

Despite the impressive capabilities of AI and LLMs, the human touch remains a key element in the integration of these technologies into precision medicine. Researchers and healthcare professionals play an irreplaceable part when it comes to understanding AI findings, applying insights to patients, and adhering to established standards. 

By fostering collaboration between human expertise and technological advances, a new synergy can be achieved—one that accelerates the translation of discoveries into clinical practice while maintaining a focus on high-quality patient care. This collaboration ensures that, while AI and LLMs may accelerate research, human oversight guarantees that the end results are tailored to meet the patients’ unique needs. 

Accelerating to Discovery 

The pace of medical discovery today is way too slow for the rapidly growing healthcare landscape. A comprehensive approach—incorporating AI, LLMs, and the expertise of human touch—within a secure data-sharing and analytic platform will allow the precision medicine industry to accelerate scientific breakthroughs and bring novel treatments to patients faster, ultimately improving patient outcomes sooner than ever before.  

Dr. Rakesh Nagarajanis the Chief Medical Officer at Velsera, focusing on democratizing clinical genomics and advancing precision research through 'omics technologies. With nearly thirty years of experience at the intersection of computer science, informatics, and medicine, he is a trained physician scientist committed to clinical and translational research. Velsera is a Health Tech company providing advanced software enabled by expert services that accelerates the discovery, development and delivery of Precision Medicine globally. We are on a mission to make breakthroughs happen faster, so that lives are improved sooner.