About

Hi, my name is Teya Bergamaschi, and I am a Ph.D. student at MIT in the EECS department working in the Computational Cardiovascular Research Group. My research has focused on developing machine learning models for cardiovascular health monitoring, leveraging multimodal data, including waveform, image/video, text, and EHR data. Primarily, I have worked in self-supervised and representation learning approaches to enable robust phenotyping, disease prediction, and patient stratification using large-scale medical datasets.

Before coming to MIT, I spent a year working in the Trayanova Lab at Johns Hopkins University, where I also earned my M.S.E. in Biomedical Engineering and my B.S. in Biomedical Engineering and Applied Mathematics.

News

• December 2024: I attended ML4H as an author and organizer.
• December 2024: I attended NeurIPS and presented work at the Time Series in the Age of Large Models workshop.
• October 2024: I attended and gave a talk at IEEE BSN.

Publications

• MEDS-torch: An ML Pipeline for Inductive Experiments for EHR Medical Foundation Models

  Nassim Oufattole, Teya Bergamaschi, Pawel Renc, Aleksia Kolo, Matthew McDermott, Collin Stultz, "MEDS-torch: An ML Pipeline for Inductive Experiments for EHR Medical Foundation Models." NeurIPS Workshop on Time Series in the Age of Large Models, 2024.

• MEDS-Tab: Automated tabularization and baseline methods for MEDS datasets

  Nassim Oufattole*, Teya Bergamaschi*, Aleksia Kolo, Hyewon Jeong, Hanna Gaggin, Collin Stultz, Matthew McDermott, "MEDS-Tab: Automated tabularization and baseline methods for MEDS datasets." arXiv preprint arXiv:2411.00200, 2024.

• MEDS Decentralized, Extensible Validation (MEDS-DEV) Benchmark: Establishing Reproducibility and Comparability in ML for Health

  MEDS Team, "MEDS Decentralized, Extensible Validation (MEDS-DEV) Benchmark: Establishing Reproducibility and Comparability in ML for Health." ML4H 2024 Demonstration Track, 2024.

• Heart Block Identification from 12-Lead ECG: Exploring the Generalizability of Self-Supervised AI

  Teya Bergamaschi, Collin Stultz, Ridwan Alam, "Heart Block Identification from 12-Lead ECG: Exploring the Generalizability of Self-Supervised AI." IEEE 20th International Conference on Body Sensor Networks (BSN), 2024.

• Predicting intensive care delirium with machine learning: Model development and external validation

  Kirby Gong, Ryan Lu, Teya Bergamaschi, Akaash Sanyal, Joanna Guo, Han Kim, Hieu Nguyen, Joseph Greenstein, Raimond Winslow, Robert Stevens, "Predicting intensive care delirium with machine learning: Model development and external validation." Anesthesiology, 2023.

• Customized gaming system engages young children in reaching and balance training

  Sundari Parise, Katharine Lee, Joshua Park, Cari Sullivan, Rebecca Schlesinger, Maggie Li, Samiksha Ramesh, Nicholas Maritato, Teya Bergamaschi, "Customized gaming system engages young children in reaching and balance training." Journal of Rehabilitation and Assistive Technologies Engineering, 2023.

• Rationale: Patients with ischemic cardiomyopathy (ICMP) are at high risk for malignant arrhythmias

  Konstantinos Aronis, Adityo Prakosa, Teya Bergamaschi, Ronald Berger, Patrick Boyle, Jonathan Chrispin, Suyeon Ju, Joseph Marine, Sunil Sinha, Harikrishna Tandri, "Rationale: Patients with ischemic cardiomyopathy (ICMP) are at high risk for malignant arrhythmias." Artificial Intelligence in Heart Modelling, 2022.

• Machine Learning for Intensive Care Delirium Prediction

  Kirby Gong, Ryan Lu, Teya Bergamaschi, Akaash Sanyal, Joanna Guo, Han Kim, Robert Stevens, "Machine Learning for Intensive Care Delirium Prediction." ANESTHESIA AND ANALGESIA, 2021.

• Characterization of the electrophysiologic remodeling of patients with ischemic cardiomyopathy by clinical measurements and computer simulations coupled with machine learning

  Konstantinos Aronis, Adityo Prakosa, Teya Bergamaschi, Ronald Berger, Patrick Boyle, Jonathan Chrispin, Suyeon Ju, Joseph Marine, Sunil Sinha, Harikrishna Tandri, "Characterization of the electrophysiologic remodeling of patients with ischemic cardiomyopathy by clinical measurements and computer simulations coupled with machine learning." Frontiers in Physiology, 2021.

• 743: Computational Endotypes of ICU Delirium

  Kirby Gong, Ryan Lu, Joanna Guo, Teya Bergamaschi, Akaash Sanyal, Hanbiehn Kim, Robert Stevens, "743: Computational Endotypes of ICU Delirium." Critical Care Medicine, 2021.

• 27: Machine Learning Prediction of Intensive Care Unit Delirium

  Kirby Gong, Ryan Lu, Teya Bergamaschi, Akaash Sanyal, Joanna Guo, Hanbiehn Kim, Robert Stevens, "27: Machine Learning Prediction of Intensive Care Unit Delirium." Critical Care Medicine, 2021.

(Some) Ongoing Projects

• MEDS-Torch: A PyTorch-based ML pipeline for inductive experiments for EHR medical foundation models.
• MEDS-DEV: The MEDS Decentralized Extensible Validation benchmarking project. Establishing reproducibility and comparability in ML for health.
• MEDS-Evaluation: A repository for evaluating MEDS models, offering tools to assess the performance of machine learning models in the context of electronic health record binary classification tasks.

To see how some of these projects fit together, check out our MEDS: a Health AI Ecosystem tutorial and start-up guide.

Installable Tools

• MEDS-Tab: An easy-to-use Python package available on PyPI for efficient tabularization and featurization of MEDS format datasets for ML baseline generation.