The age of AI is here. How can we develop the AI tools of tomorrow to positively benefit patient and survivor outcomes for all who battle cancer? This is the question fueling the work of several researchers at the Cancer Center at Illinois (CCIL). Within the complex landscape of AI and cancer research, according to CCIL Director Rohit Bhargava, the CCIL holds a “unique position” to engage this critical task.
Like other paradigm-shifting discoveries in history, AI’s influence is already far-reaching and likely to become more apparent in the coming years. We’ve already grown accustomed to AI’s presence in our phones, cameras, social media platforms, email accounts, transportation, finances, health care, artistic creation, and more. While AI as a technology has been in use for decades, the pace, scale, and reach of recent AI innovations have inaugurated a global reckoning with the momentum of this potent technology. Leaders in government, business, academia, and private institutions are congregating not only around how this technology can benefit human health but also questions to promote and safeguard AI’s ethical implementation.
“How can we develop the AI tools of tomorrow to positively benefit patient and survivor outcomes for all who battle cancer?” The CCIL is developing AI solutions in response to this question.
While AI might be presently inert, left to its own—dystopian fears of an age of AI dominion aside— its potential power for cancer research will be missed without our careful attention in developing and implementing this technology.
“AI is like fire. It’s going to transform human existence and it is here to stay,” said Bhargava. “The challenge today is to harness this technology to provide useful, cost-effective, and life-changing tools. That’s why the CCIL is focused on building equitable and effective AI tools of tomorrow that will ultimately help us enhance personalized treatments for all patients and survivors who seek to overcome cancer.”
As a basic cancer center existing within the diverse, multidisciplinary ecosystem of engineers and scientists at the University of Illinois, the CCIL is on the leading edge of AI’s future. “We are uniquely positioned to develop and deploy beneficial AI tools,” said Bhargava. “The key idea is that AI is not a tool to be used in isolation. The CCIL has leading expertise to integrate AI in all that we do to advance drug discovery, new imaging technologies, understanding cancer and make treatments more effective. We have a once in a lifetime opportunity to accelerate fundamental discoveries to use by harnessing the emerging capabilities of AI.”
AI is pervasive in all six core themes of the CCIL’s two cancer research programs. Be it imaging to diagnose cancers, molecular measurement for early detection, computational biology to design drugs, understanding the molecular pathways that drive tumors, discovering life-saving drug, or using cancer models to enhance all research, CCIL members are working within interdisciplinary collaborations to use AI tools to increase positive outcomes for patients and survivors.
One of the ways the CCIL is doing this is by addressing the gaps in data. Like a pet trained on treats, AI is trained by data. What you feed it matters. If you provide AI systems with data that are biased in representation, it has real-world consequences for underrepresented patients. “We are developing solutions that help change cancer for everyone, but we are especially cognizant that this emerging area can transform how we address cancer in our local communities,” said Bhargava.
To address this, the CCIL recently launched the BEAT Cancer Initiative in partnership with OSF Healthcare in Peoria, IL. OSF’s timely installation of proton beam therapy technology at their cancer center positions them as a regional leader in cancer healthcare. This increasing reach in clinical care provides a consequential opportunity for the CCIL and OSF to address gaps in patient and science data that ultimately fuel the development of equitable AI solutions.
“You can only build a solution based on the data of the community you are drawing from. We want to ensure we build a solution that works for all OSF patients,” said Bhargava. “One disparity we are addressing is the rural-urban disparity. Our next goal will be to expand the inclusion criteria to other underserved groups.”
Another example of the CCIL’s frontier work in AI solutions for cancer includes the new Cancer Digital Insights working group. Leveraging a collaboration with the National Center for Supercomputing Applications (NCSA), the CCIL’s newest working group seeks to ensure that the data and algorithms available will maximize and organize their insights into cancer progression, ideally down to the single-patient level. The Cancer Digital Insights group aims to gain understanding into cancer biology to design cancer progression models better. This partnership builds on decades of exceptional computing research and expertise at NCSA, which continues to be a national leader in computing and AI.
“Collaborating with the CCIL provides us with a unique opportunity to apply our world-class expertise and cutting-edge computing and data infrastructure to tackle complex problems using emerging AI-based methods,” said William Gropp, NCSA Director and professor of computer science at Illinois. “Our partnership with the CCIL’s Cancer Digital Insights group is a prime example of NCSA’s commitment to push the boundaries of what is possible while addressing pressing societal needs. By combining the CCIL’s vision and drive with NCSA’s computing technologies and expertise in AI, we aim to amplify the impact of the Cancer Digital Insights working group.”
In addition to the advantageous AI partnership with NCSA, CCIL members also leverage another research benefit that Illinois affords: a collaboration with the College of Veterinary Medicine, where comparative oncology provides helpful data from companion cats and dogs. Through the diagnosis and treatment of cancer in pet dogs and cats, large databases populated with digital information including magnetic resonance and computed tomographic imaging, tissue histology, and correlative clinical outcome data serve as a treasure trove for AI interrogation and development of unrealized algorithms that can aid in the most accurate diagnosis, AI-guided treatment stratification, and disease prognostication for improving veterinary healthcare. “Through the vision and leadership of CCIL researchers, the time is ripe to accelerate the development of new, more precise methods for treating cancer that can be piloted through a comparative oncology approach,” said Timothy Fan, a professor of veterinary medical oncology and Associate Director for Translational Research and Development at the CCIL.
“We have our eyes fixed on growing as a leader in AI and cancer research. We have ten faculty lines open right now for researchers who want to join us at this critical nexus of fundamental science and AI development,” notes Bhargava. “The CCIL is bringing AI-powered tools to the foundational level of cancer research, and we have the potential to improve the quality of life for survivors and patients from the ground up. Investing the time and energy in this effort will lead to transformative advances that we can only imagine today.”
Editor’s notes:
This story was written by Jonathan King, CCIL Communications Specialist