NCI Workshop on Emerging Technologies in Cancer Epidemiology (Speaker Bios)

Abstract: In this talk, Dr. Platz will showcase examples of molecular epidemiology research on cancer incidence and progression that leverage large scale -omics platforms and complex tissue-based measurements. Across these examples, Dr. Platz will illustrate challenges related to samples, measurement, classification based on measurements, external cohort confirmation, and inferences when implementing emerging technologies in large observational studies. Dr. Platz will also highlight the critical need for collaboration with basic scientists to achieve meaningful “back translation” to strengthen causal inference and thus advance the development of strategies for cancer prevention and treatment.

Biography: Elizabeth A. Platz, ScD, MPH, is a Professor and the Martin D. Abeloff, MD Scholar in Cancer Prevention in the Department of Epidemiology at the Johns Hopkins Bloomberg School of Public Health. Dr. Platz holds joint appointments in the Departments of Oncology and Urology at the Johns Hopkins University School of Medicine and is the Associate Director for Population Sciences at the Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins. Her research, conducted in large cohort studies, focuses on the molecular epidemiology of cancer incidence and progression with an eye toward translation of findings into strategies for prevention and treatment. Dr. Platz is the editor-in-chief of the American Association for Cancer Research’s journal Cancer Epidemiology, Biomarkers and Prevention and is an elected Fellow of the American Association for the Advancement of Science.

Abstract: Rapid advances in microbiome and multi-omics profiling have created new opportunities to improve cancer risk prediction, yet translation to large-scale population studies remains limited by data heterogeneity, batch effects, missing modalities, and a lack of uncertainty-aware modeling. In this talk, Dr. Mallick will discuss emerging multimodal AI frameworks for integrating and harmonizing microbiome multi-omics data across studies, assays, and populations, with a focus on feasibility for epidemiological research.

Biography: Himel Mallick, PhD, is a tenure-track Assistant Professor of Population Health Sciences and an adjunct faculty member of Statistics and Data Science and Gastroenterology and Hepatology at Cornell University. Prior to joining Cornell, Dr. Mallick was an Associate Director at Merck Research Laboratories and a postdoctoral fellow at Harvard University and the Broad Institute. He previously served as the primary statistician for the Integrated Human Microbiome Project (iHMP), published in Nature in 2019. He is the developer of several bioBakery tools, including MelonnPan and MaAsLin2, which are widely used in microbiome data analysis. Dr. Mallick has authored multiple book chapters on microbiome multi-omics integration, including Metagenomics for Microbiology, and co-authored Orchestrating Microbiome Analysis, a widely adopted community resource outlining best practices for microbiome research. He is the recipient of multiple early-career awards in statistics and data science and is a Fellow of the American Statistical Association and an Elected Member of the International Statistical Institute.

Abstract: Emerging artificial intelligence (AI) technologies offer unprecedented opportunities to enhance cancer epidemiology by enabling dynamic, scalable risk prediction and personalized survivorship care across large populations. Dr. Lin will present an integrated AI framework, spanning pre-treatment risk stratification to post-diagnosis survivorship management.

In head and neck cancer (HNC) patients undergoing radiotherapy, Dr. Lin’s research team developed and validated an AI system to predict post-radiotherapy epigenetic aging acceleration (EAA) using only pre-treatment sociodemographic, clinical, symptom, and immune biomarker data. Deep learning approaches, particularly TabNet, outperformed conventional algorithms (average RMSE 4.08 ± 0.32), with strongest performance immediately post-treatment (Time 2 RMSE 4.87). Baseline immune markers, including absolute eosinophil count and hemoglobin levels, consistently predicted longitudinal EAA trajectories. Importantly, this approach enables identification of high-risk patients without costly epigenetic assays, demonstrating feasibility for large-scale population implementation.

Building on this predictive infrastructure, Dr. Lin’s research team is developing AI-enabled digital survivorship platforms (e.g., CancerCompanion, DietAI24) that leverage large language models, reinforcement learning, and just-in-time adaptive interventions (JITAI) to dynamically monitor diet, physical activity, and symptoms in real time. These platforms function not only as interventions but also as scalable digital cohort infrastructures, generating high-frequency longitudinal data that continuously refine risk prediction models and enable adaptive, personalized care.

Collectively, this integrated AI paradigm supports a shift from static risk estimation to dynamic, population-scale cancer trajectory modeling. By combining pre-treatment biological risk prediction with post-treatment digital phenotyping and intervention, AI has the potential to enhance precision cancer epidemiology, improve early identification of vulnerable populations, and facilitate cost-effective, equitable implementation across diverse clinical settings.

Biography: Yufen Lin, PhD, is an Assistant Professor at Emory University's Nell Hodgson Woodruff School of Nursing and a member of the Cancer Prevention and Control Research Program at the Winship Cancer Institute. She completed her postdoctoral fellowship at Emory School of Nursing and earned her PhD in Nursing from Duke University. Her research interests focus on cancer symptom science and management, innovative interventions, -omics science, and health equity. Her current research program centers on two main themes: (i) developing and evaluating technology- and AI-based interventions to improve health outcomes for cancer patients and their family caregivers through a lens of health equity; and (ii) investigating the biological mechanisms underlying cancer-related aging outcomes using AI-powered multi-omics approaches. Dr. Lin’s research has been supported by multiple organizations, including the National Institutes of Health, the American Cancer Society, the Oncology Nursing Foundation, and intramural funding sources. Her honors include the Distinguished Dissertation Award from Duke University, the Early-Stage Investigator Fellowship from NRG Oncology, and the Rising Star Award from Sigma Theta Tau International Honor Society of Nursing.

Abstract: Advances in multi-omics have provided new sources of data to characterize tumor biology, but translating these data into accurate, population-scale outcome predictions is a major challenge. Generative artificial intelligence (AI) offers a promising paradigm shift. By learning latent representations across heterogeneous data modalities (including genomic, transcriptomic, and clinical data), emerging models may better capture complex disease patterns to enhance prognostic modeling beyond traditional approaches. This presentation will provide an overview of the role of generative AI in predicting cancer outcomes, including reviewing key considerations for scaling and implementation, such as harmonizing multi-omic data, computational infrastructure requirements, interpretability, and governance needed for responsible implementation. Drawing on real-world efforts to integrate molecular and clinical data at scale, this talk will highlight practical pathways and current limitations.

Biography: Anai N. Kothari, MD, MS, is an Assistant Professor in the Division of Surgical Oncology, Department of Surgery, at the Medical College of Wisconsin. He holds additional faculty appointments in the Clinical & Translational Science Institute of Southeast WI (CTSI) and the MCW Data Science Institute. Dr. Kothari grew up in Manitowoc, Wisc., and received his BS and MD at the University of Wisconsin-Madison. Dr. Kothari completed his General Surgery training at Loyola University Medical Center, where he also earned a master’s degree in Epidemiology, followed by a fellowship in Complex General Surgical Oncology at the University of Texas MD Anderson Cancer Center. Dr. Kothari’s clinical expertise is the surgical treatment of upper gastrointestinal tract and advanced cancers of the GI tract, including the use of advanced techniques such as minimally invasive and robot-assisted surgery.