City of Hope creates 1st oncology-specific large language model

City of Hope has created one of the first oncology-specific large language models, Nasim Eftekhari, executive director of applied AI and data science at the Duarte, Calif.-based organization, told Becker's.

City of Hope was able to develop, refine and validate an oncology-specific large language model. This process involved leveraging more than 8 billion words extracted from around 6 million clinical, radiology and pathology notes at City of Hope. 

According to Ms. Eftekhari, the tool holds promise for diverse applications, including document summarization, answering questions, data structuring, clinical trials matching and more.

AI's future at City of Hope 

When it comes to large language models and AI, Ms. Eftekhari said the organization sees a lot of promise, especially in easing the administrative load on healthcare providers.  

"By automating routine tasks such as documentation, coding and report generation, generative AI can free up healthcare professionals to focus more on patient care," she said. "This approach not only improves efficiency but also reduces the risk of burnout among healthcare staff, enhancing overall healthcare delivery."

This means in 2024, Ms. Eftekhari's role will be heavily focused on AI and generative AI. 

For example, Ms. Eftekhari said she is working on continuing to establish and enforce comprehensive governance structures for AI, with a focus on ethical guidelines, algorithmic transparency and accountability. 

Additionally, she will also be homing in on the strategic implementation of generative AI.

"I will be working to identify and evaluate viable applications of generative AI and large-language models at City of Hope, focusing on operational efficiency, drug discovery and precision medicine," she said. "Artificial intelligence is playing a huge role in streamlining the drug discovery process, from identifying potential drug candidates to assisting in clinical trial design. This can significantly reduce the time and cost of bringing new drugs to market."

Ms. Eftekhari also detailed plans to invest in education initiatives and collaboration so the organization can enlighten all stakeholders about the capabilities and limitations of generative AI, and foster partnerships between AI technologists both within and outside City of Hope. 

AI's impact on healthcare roles

Ms. Eftekhari said the emergence and rapid evolution of generative AI, particularly LLMs, marked a significant shift in her role in 2023.

"This [the emergence] required strategizing the integration of these advanced AI technologies into City of Hope's health system to enhance various aspects, such as patient care, administrative efficiency and clinical research," she said. "The fast-paced changes in the AI landscape demanded agility and flexibility in our approach, ensuring that we not only kept pace with these technological advancements but also continued to lead in the application of AI in healthcare."

Adapting to and leveraging the capabilities of generative AI, according to Ms. Eftekhari, is becoming essential for healthcare organizations like City of Hope so they can maintain a "cutting-edge position and deliver innovative healthcare solutions."

Editor's note: Some of the responses have been lightly edited for clarity. 

Copyright © 2024 Becker's Healthcare. All Rights Reserved. Privacy Policy. Cookie Policy. Linking and Reprinting Policy.


Featured Whitepapers

Featured Webinars