Hello, thank you for visiting my website! 👋
I am currently a Ph.D. student at the School of Information Sciences at the University of Illinois Urbana-Champaign (UIUC), where I am a member of the Causal Science Lab and am honoured to be advised by Prof. Yonghan Jung.
Prior to joining the lab, I studied Electronic and Electrical Engineering at University College London (UCL) on a four-year integrated master's programme (MEng), graduating with a First Class Honours Degree in 2021. There, I was advised by Prof. Yiannis Andreopoulos (now at Sony Interactive Entertainment UK).
After graduating, I completed a three-year alternative military service, during which I worked as a researcher at Severance Hospital, Yonsei University Health System, Korea, under the guidance of Prof. Hyuk-Jae Chang. I also worked as a research assistant at the Max Planck Institute for Security and Privacy (MPI-SP), under the mentorship of Prof. Cheng Yaw Low (now at Changwon National University, Korea) and Prof. Meeyoung Cha. In addition, I have collaborated with Prof. June-Woo Kim (Wonkwang University, Korea) on respiratory sound classification.
I am deeply grateful to all of my mentors and collaborators, and to everyone who has supported me on this journey, including many not mentioned here.
For more publications, please visit my Google Scholar page.
Next-visit diagnosis prediction forecasts the diagnoses documented at a patient's next encounter, turning retrospective electronic health records into prospective clinical prediction. However, conventional aggregate evaluation obscures its multifaceted nature: distributionally, models must handle variable-size diagnosis sets and long-tailed labels; temporally, they must distinguish persistent, newly recorded, and recurrent diagnoses; and both vary with observed history depth. We introduce an evaluation framework along these dimensions and apply it to rule-based heuristics, LLM-based prompting, retrieval, memory, and multi-agent approaches. This reveals that repeating the last visit outperforms most LLM-based approaches overall despite predicting no new or recurrent diagnoses, and that retrieval and memory exhibit distinct profiles whose relative strengths vary across temporal transitions and backbones. Motivated by this, we propose LONGINUS, a retrieval-memory multi-agent adjudication framework combining evidence retrieved from patient records with structured memory of the patient trajectory. Using cross-view agreement as a reliability heuristic, it preserves shared predictions and adjudicates disagreements without revealing their source. Across MIMIC-III and MIMIC-IV with four LLM backbones, LONGINUS achieves the best transition-specific performance among LLM-based methods in nearly all settings while remaining competitive with them on aggregate metrics.
@article{koo2026rethinking,
title={Rethinking Next-Visit Diagnosis Prediction Beyond Aggregate Metrics: Retrieval--Memory Multi-Agent Adjudication},
author={Koo, Heejoon},
journal={arXiv preprint arXiv:2611.03826},
year={2026}
}
Multimodal respiratory sound classification offers promise for early pulmonary disease detection by integrating bioacoustic signals with patient metadata. Nevertheless, current approaches remain vulnerable to spurious correlations from attributes such as age, sex, or acquisition device, which hinder their generalization, especially under distribution shifts across clinical sites. To this end, we propose a counterfactual adversarial debiasing framework. First, we employ a causal graph-based counterfactual debiasing methodology to suppress non-causal dependencies from patient metadata. Second, we introduce adversarial debiasing to learn metadata-insensitive representations and reduce metadata-specific biases. Third, we design counterfactual metadata augmentation to mitigate spurious correlations further and strengthen metadata-invariant representations. By doing so, our method consistently outperforms strong baselines in evaluations under both in-distribution and distribution shifts. Code is available at https://github.com/RSC-Toolkit/BTS-CARD.
@inproceedings{koo2026empowering,
title={Empowering multimodal respiratory sound classification with counterfactual adversarial debiasing for out-of-distribution robustness},
author={Koo, Heejoon and Toikkanen, Miika and Kim, Yoon Tae and Kim, Soo Yong and Kim, June-Woo},
booktitle={ICASSP 2026-2026 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
pages={14967--14971},
year={2026},
organization={IEEE}
}
Predicting next visit diagnosis using Electronic Health Records (EHR) is an essential task in healthcare, critical for devising proactive future plans for both healthcare providers and patients. Nonetheless, many preceding studies have not sufficiently addressed the heterogeneous and hierarchical characteristics inherent in EHR data, inevitably leading to sub-optimal performance. To this end, we propose NECHO, a novel medical code-centric multimodal contrastive EHR learning framework with hierarchical regularisation. First, we integrate multifaceted information encompassing medical codes, demographics, and clinical notes using a tailored network design and a pair of bimodal contrastive losses, all of which pivot around a medical codes representation. We also regularise modality-specific encoders using a parental level information in medical ontology to learn hierarchical structure of EHR data. A series of experiments on MIMIC-III data demonstrates effectiveness of our approach.
@inproceedings{koo2024next,
title={Next visit diagnosis prediction via medical code-centric multimodal contrastive ehr modelling with hierarchical regularisation},
author={Koo, Heejoon},
booktitle={Findings of the Association for Computational Linguistics: EACL 2024},
pages={41--55},
year={2024}
}
Last updated: November 17, 2026 · Based on the Minimal Light theme.