Caroline Chung, MD, reflects on quantitative imaging, multimodal artificial intelligence, leadership, and the work required to turn data into better cancer care.
Cancer care is entering a period in which data is no longer simply collected, stored, and reviewed after the fact. It is increasingly becoming part of how clinicians anticipate risk, understand tumour biology, guide treatment decisions, and improve care delivery across health systems.
For Caroline Chung, MD, the future of oncology will not be determined by algorithms alone. It will depend on whether institutions can build the culture, standards, trust, and collaboration needed to use data responsibly.
Dr. Caroline Chung is Co-Director of the Institute for Data Science in Oncology at The University of Texas MD Anderson Cancer Center and a tenured professor in Radiation Oncology. A clinician specializing in central nervous system malignancies, she has spent her career connecting clinical challenges with quantitative imaging, data science, precision medicine, and health system strategy.
Her work has also included serving as MD Anderson’s inaugural Vice President and Chief Data & Analytics Officer, helping shape how a major cancer center approaches data access, analytics, technology, and organizational learning.
Seeing More Than an Image
For Dr. Caroline Chung, the connection between imaging and patient care developed over time through clinical practice, research, and the growing potential of computational methods.
In radiation oncology, defining the treatment target is one of the most important steps in delivering effective therapy. Early in her career, she became interested in whether medical images could reveal more than anatomy alone.
Could imaging provide information about the biology of a tumour? Could it help identify areas more likely to respond to treatment, recur after therapy, or require intensification?
These questions led her toward quantitative medical imaging, a field based on the idea that a pixel should not only be seen as an image signal but also understood as a measurable source of clinical information.
Medical imaging generates enormous volumes of data. Yet much of that data is still used primarily as a visual aid for clinical decision-making. Dr. Caroline Chung believes that the ability to extract quantitative measurements from imaging could help advance precision medicine, accelerate therapeutic development, and make clinical trials more efficient.
“Precision medicine relies on precision measurements,” she noted.
Why Culture Matters as Much as Technology
Dr. Caroline Chung’s approach to artificial intelligence and data strategy begins with people.
As MD Anderson’s inaugural Chief Data Officer and later Chief Data & Analytics Officer, she had the opportunity to consider a fundamental question: what should a data strategy look like in order to serve the mission of ending cancer?
Her answer was not focused only on which model or platform to adopt.
Instead, it centered on whether people across the organization could find, appropriately access, and trust data. It also required clinicians, researchers, administrators, and patients to see themselves as participants in a learning health system that can improve over time.
A future-ready health system, she argues, is not built through dashboards or models alone. It is built through shared trust, meaningful collaboration, and a common belief that data should improve care every day.
This is especially important because cancer care often extends beyond a single institution. Patients may receive imaging, systemic therapy, surgery, radiation, supportive care, and follow-up in different locations. Realizing the potential of data-driven oncology will require collaboration across systems, specialties, and stakeholders.
Multimodal Data May Change the Direction of Cancer Care
Among the developments shaping oncology in 2026, Dr. Chung identifies multimodal artificial intelligence as one of the most promising.
The ability to integrate pathology, radiology, genomics, and clinical information into a single analytical framework could offer a more complete understanding of a patient’s disease. Rather than examining each source of information in isolation, multimodal approaches may reveal patterns that are difficult to detect through one data type alone.
This has implications for diagnosis, risk assessment, treatment selection, and research.
However, Dr. Caroline Chung also stresses that progress should not be measured only by model accuracy. Much more research is needed to understand how these tools are implemented in clinical settings, whether they improve outcomes, and how they affect the relationship between clinicians and technology.
The human interface remains one of the most important unanswered questions. How will artificial intelligence affect learning, critical thinking, judgment, and clinical decision-making? How can health systems reduce risk while ensuring that these tools support rather than weaken professional expertise?
These questions will shape whether promising technologies translate into meaningful clinical impact.
From Reactive Imaging to Anticipatory Care
Quantitative imaging and predictive modeling are changing the role of imaging in oncology.
Traditionally, imaging workflows have been designed to allow radiologists to visually interpret scans and report whether disease appears to be progressing, responding, or remaining stable. While this remains essential, the broader opportunity lies in moving from reactive interpretation toward anticipatory clinical decision-making.
Dr. Caroline Chung describes this as a repositioning of imaging from a qualitative adjunct to a quantitative measurement platform.
In early detection, for example, artificial intelligence models using digital breast tomosynthesis have shown the potential to estimate five-year breast cancer risk from routine screening images. Other models are being used to anticipate the need for cardiac catheterization.
The ability to identify risk earlier could change how clinicians approach prevention, screening, surveillance, and intervention.
Yet major barriers remain.
The Standardization Problem Behind Predictive Medicine
Health systems are still underprepared for one of the most practical requirements of predictive medicine: standardization.
Imaging data can vary across institutions, scanners, acquisition protocols, and technical settings. Without reproducible measurements, predictive models may struggle to perform consistently across different environments.
“You can’t build a predictive model on top of data that isn’t reproducible,” Dr. Caroline Chung explained.
She argues that image acquisition should be treated with the same rigor applied to laboratory assays. Until this happens, valuable biological signals may remain hidden beneath technical variability.
The opportunity is substantial. Quantitative imaging may help clinicians identify features related to the tumour microenvironment, molecular phenotype, and treatment response without requiring an additional biopsy.
The challenge is ensuring that these signals remain meaningful across diverse imaging environments.
Caroline Chung Received the AI in Oncology Yvonne Award Named by ZS 2026
Mentorship Gives Advice. Sponsorship Opens Doors.
Dr. Caroline Chung’s leadership extends beyond data science and radiation oncology. As Chair of Women in Cancer – All in Cancer, she has worked to promote leadership development across clinicians, researchers, educators, administrators, STEM professionals, industry representatives, regulatory experts, and patient advocates.
A central part of this work has been the “Strengthening Through Perspectives” event series, designed to bring together people with different professional backgrounds, experiences, and expertise.
The goal is not simply representation. It is meaningful dialogue.
When people with different perspectives engage around a shared purpose, they can develop deeper understanding, challenge assumptions, and create momentum around common problems. For individuals facing structural barriers, these conversations can also help shift the perception that systemic challenges are personal failures.
Dr. Caroline Chung emphasizes the importance of both mentorship and sponsorship.
“Mentorship gives advice. Sponsorship opens doors.”
For women and underrepresented leaders, sponsorship can mean advocacy, nomination, visibility, and access to rooms where critical decisions are made. It requires leaders to move beyond guidance and actively create opportunities for others.
A Career at the Intersection of Medicine, Data, and Technology
For young professionals interested in oncology, data science, and technology, Dr. Caroline Chung does not offer a conventional checklist.
Instead, she highlights several deeper capabilities.
The first is translational fluency: the ability to move between disciplines while helping people understand one another. This may mean speaking about oncology in one setting and data architecture in another, while ensuring that everyone is working toward the same clinical goal.
The second is communication. Clear, authentic communication is essential for building trust, sharing a vision, and helping teams act together. It also requires listening.
The third is the ability to ask critical questions and distinguish meaningful signal from noise. In a field moving as quickly as data science and artificial intelligence, excitement can easily outpace evidence. The ability to evaluate data, technology, and clinical relevance is increasingly important.
Dr. Caroline Chung also challenges the belief that professionals must choose between depth and breadth.
The strongest leaders in this space, she argues, are those who develop enough expertise to be taken seriously while remaining broad enough to connect disciplines and lead across them.
A Goal Measured in One Billion Lives
Dr. Caroline Chung describes her long-term ambition as a “Big Hairy Audacious Goal”: to positively impact the lives of at least one billion people.
For her, the number is not about visibility. It is a compass.
It changes the questions she asks about where to invest time, what to build, what to publish, and how to use her expertise and network. It also places equity at the center of scale.
The future of health data, computational medicine, and artificial intelligence will be shaped by decisions made over the next few years. Those decisions could influence cancer care for decades.
Dr. Caroline Chung hopes to contribute to the systems, policies, and platforms that can bring precision medicine beyond major academic centers and into communities that have historically had less access to advanced cancer care.
Her professional journey, she says, has been guided by a belief that the greatest impact is not achieved alone or in a single step. It is created by connecting knowledge, resources, and people through shared purpose and collective learning.
Written by Nare Hovhannisyan, MD