2025 IARS Mentored Research Award Recipient Interview: Lichy Han, MD, PhD

A Hybrid Deep Learning Framework for Causal Inference and Prediction of Postoperative Outcomes

Abstract Presentation:

Mini Oral Abstract Session C on Saturday, May 2, 2026, 10:00 am – 11:00 am, in Square Dorchester + Place du Canada, Convention Level at the Fairmont The Queen Elizabeth, during the 2026 Annual Meeting, Presented by IARS and SOCCA

As an anesthesiologist and bioinformatician, Lichy Han, MD, PhD, has been fascinated by the intersection of clinical care and data science throughout her research career. Initially, Dr. Han sought out to investigate the potential for advanced computational methods to solve the complex perioperative challenges. In 2025, as Instructor of Anesthesiology, Perioperative and Pain Medicine at Stanford University, Dr. Han’s studies into “A Foundation Model for Perioperative Prediction and Causal Inference” drew the attention of the IARS and she was recognized with an IARS Mentored Research Award. Since receiving this award, she has shifted her investigative focus from predictive modeling to include integrating causal inference frameworks that allow their research team to identify actionable interventions in addition to identifying high-risk patients. Although she is only four months into her IMRA-funded study, Dr. Han is making significant headway, facilitating collaborations within the Stanford Department of Anesthesiology and gaining the resources necessary to begin expanding her models. She is confident that this instrumental funding and these collaborations will prove vital in launching her trajectory as a physician-scientist. Already, she has found that leaning into interdisciplinary expertise is vital for innovation and she continues to do just that. On Saturday, May 2, during the Mini Oral Abstract Session C at the 2026 Annual Meeting, presented by IARS and SOCCA, she will share the results of her investigations so far. Below, she discusses how her research is developing, how the Annual Meeting provides an opportunity to gain vital feedback and her hopes for the future of her study.

1. For this research, you are…

Principal Investigator on my IMRA, which this project is directly related to.

2. What inspired you to pursue this area of research, and how has your focus evolved since your IMRA award in 2025?

My background as an anesthesiologist and bioinformatician has always driven me toward the intersection of clinical care and data science. I am inspired by the potential for advanced computational methods to solve complex perioperative challenges. Since receiving the IMRA award, my focus has shifted from predictive modeling to include integrating causal inference frameworks that allow us to identify actionable interventions in addition to identifying high-risk patients.

3. What key question or challenge does your current research aim to address?

My research addresses the limitations of traditional risk prediction tools, which often fail to capture the temporal evolution and granular complexity of a patient’s physiologic state. Specifically, we aim to overcome these challenges by incorporating high-resolution, dynamic intraoperative data and a framework for evaluating causal effects in retrospective data using machine learning.

4. How do you hope that your research, whether through direct clinical application or by advancing foundational knowledge, will impact the field of anesthesiology and ultimately improve patient care?

By successfully developing a framework that utilizes both the preoperative profile and dynamic intraoperative data, I hope to provide clinicians with highly granular risk assessments to reduce postoperative complications. Beyond that, I hope to provide new methodologies to assess the causal impact of intraoperative interventions retrospectively, addressing the limitations of traditional observational studies and randomized controlled trials.

5. What does presenting your abstract at the IARS Annual Meeting mean to you, and how does it support your professional growth?

Presenting at the IARS Annual Meeting allows me to engage with a community dedicated to academic excellence in anesthesia. As an early-career instructor, sharing my work with peers provides vital feedback that helps refine my methodology for future large-scale studies.

6. Acknowledging that you are only four months into your IMRA funded research, has receiving the 2025 IARS Mentored Research Award impacted your career path, research opportunities, or collaborations yet? How do you anticipate its impact moving forward?

Receiving the award has been instrumental in launching my trajectory as a physician-scientist. It has already facilitated collaborations within the Stanford Department of Anesthesiology and provided the resources necessary to begin expanding my models. Moving forward, I anticipate this support will be a cornerstone in my development toward research independence.

7. What advice would you offer to early-career researchers or future IARS award applicants?

While challenging, I believe that leaning into interdisciplinary expertise is important for innovation. Combining clinical medicine and technical fields enabled me to approach old problems with new, more powerful tools, fostering creativity at the intersection and enabling translation between worlds. Additionally, I find that when it comes to computational modeling, focusing on the “why” helps ensure that even the most complex machine learning models remain grounded in the clinical context.

8. Is there anyone you would like to acknowledge as part of this work?

I would like to acknowledge my co-authors Marc Ghanem, Chloe Stanwyck, Philip Chung, and my primary mentor Nima Aghaeepour for their invaluable contributions to this work.

9. Outside of your research, what is something colleagues might be surprised to learn about you?

Despite being an anesthesiologist, I don’t like coffee!