The Outlook and Responsibility of AI in the Next Few Years

Artificial intelligence (AI) is playing a significant role in how cancer is treated, both in terms of clinical outcomes and patient experience.

At City of Hope, a cancer treatment and research center, Simon Nazarian, the system executive vice president and chief digital and technology officer, and Nasim Eftekhari, the executive director of applied AI and data science, are at the forefront of integrating AI into cancer care.

Their leadership is helping shape how technology not only improves treatment but also supports the emotional and mental recovery of patients.

Nazarian explained that City of Hope’s focus is on the full patient journey, beyond just physical healing.

“AI is really helping us create digital tools that will lead to really equitable access to very multidisciplinary, personalized, supportive care services,” he said.

These tools are being used to extend access to mental and behavioral health services, part of a broader goal to care for the whole person and not just the disease, he shared.

By expanding these services, AI is helping redefine treatment, providing crucial support to both patients and caregivers facing cancer.

The integration of AI into healthcare, however, brings with it serious ethical and practical challenges.

One major concern is algorithmic bias, according to Eftekhari.

She stressed that addressing bias requires ongoing attention at every step, from how algorithms are developed to how they are used.

“It’s really important, on a use-case-by-use-case basis, to look at how an algorithm is predicting on different patient populations,” Eftekhari said.

She explained that City of Hope has a rigorous AI governance process that includes ethical, legal and technical reviews to ensure fairness and safety.

With AI models constantly evolving, it’s critical to continually review their limitations and refine them to be as unbiased as possible.

One particular application of AI at City of Hope is the development of a large language model (LLM) made specifically for oncology.

This model has been trained on six million clinical notes, allowing it to understand the complexity and context of cancer care, Eftekhari shared.

This tool is already being used to summarize lengthy medical documents, some stretching over decades, into clean, two-page summaries that help doctors quickly understand a patient’s past treatments, side effects and disease progression.

“That's something that I believe is very impactful for providers and doctors, but as well for patients and their outcomes,” Eftekhari said. “It also removes the burden of patients having to inform their doctor without having the medical knowledge about what they've been through so far.”

In addition, she added that City of Hope’s LLM is becoming multimodal, meaning it will soon be able to process not just text, but also images and voice.

This could allow the AI to assist with interpreting radiology scans or even transcribing and analyzing spoken conversations between patients and providers.

Looking forward, Eftekhari sees robotics playing a larger role, with AI-driven machines potentially supporting routine healthcare tasks, a development she described as similar to ChatGPT for language processing.

“I think this year we will see a lot of breakthroughs in robots that can do everyday stuff very effectively,” she said.