HEARTMAP: Hybrid Engineered Arrhythmia-Resistant Regenerative Tissue with Machine learning-Augmented Placement

🧬imagine this:

you’re one of thousands waiting. your name’s on a transplant list, but the call never comes. your heart is failing, and the clock is ticking. no backup. no match. no options. meanwhile, the science to rebuild organs exists — just not where it’s needed. not yet.

over 4,000 canadians are currently waiting for an organ. many will never get one. stem cell-based patches exist, but they’re not personalized — and not ready for clinical use.

🧬the problem: engineered patches fail to integrate electrically

• Whole heart transplants are limited by organ availability, matching criteria, and logistics — thousands die each year waiting. Patches don’t rely on organ donors, which are in critically short supply.
• Placement is imprecise, based on guesswork. Inaccurate placement of electrodes can significantly affect the morphology of ECG waveforms, potentially leading to misinterpretation of cardiac rhythms. This underscores the importance of precise placement to ensure accurate diagnosis and effective treatment.
• Electrical mismatch leads to arrhythmias. Studies have shown that spatial profiles of electrical mismatch can determine vulnerability to reentrant arrhythmias. This highlights the critical need for electrical compatibility between the patch and the host tissue to prevent adverse outcomes.
• Post-MI hearts vary widely — but patches don’t. After a myocardial infarction (MI), the heart undergoes complex remodeling, resulting in heterogeneous tissue characteristics that are not adequately addressed by standardized patches (Proceedings of the National Academy of Sciences).
• iPSC-derived cardiomyocytes (iPSC-CMs) often exhibit immature electrical properties, leading to poor integration with native cardiac tissue.

🧬 wait, what are hiPSC-CMs?

human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) offer a promising cell source for cardiac regeneration due to their ability to differentiate into functional cardiomyocytes. Preclinical studies have demonstrated improved cardiac function following transplantation of hiPSC-CM patches in animal models.

🧬 proposed solution

the HEARTMAP team proposes the development of a novel cardiac patch scaffold embedded with conductive polymers (e.g., polypyrrole or reduced graphene oxide) seeded with iPSC-CMs. To enhance therapeutic precision, we will integrate machine learning (ML) models trained on patient-specific electrogram and imaging data to guide patch placement and predict arrhythmic zones.

regenerating heart tissue starts long before a patch ever reaches the heart. It begins at the cellular level — where a single stem cell is guided to become a beating cardiomyocyte.

🧬 the journey:

• Activin A triggers mesoderm formation, the germ layer that gives rise to the heart.
• BMP4 nudges cells toward cardiac progenitor identity.
• FGF2 supports proliferation and survival of early cardiac cells.

• Matrix stiffness must mimic natural heart tissue.
• Electrical cues guide alignment and contractility of cardiomyocytes.

these factors work together to ensure the resulting cells are ready to sync with the patient’s heart.

🧬 biocompatibility: beyond just “safe”

• Non-cytotoxicity: materials must not harm cells or trigger rejection.
• Mechanical compliance: patches must “beat” with the heart.
• Cellular integration: scaffolds must support gap junctions (e.g., connexin 43).

materials like rGO (reduced graphene oxide nanosheets) and PPy (an organic polymer) are promising, since they offer conductivity and support alignment of cardiomyocytes.

Referenced: Guo et al., 2023, Bioactive Materials.

🧬 machine learning model

for many bioengineered solutions, scalability and biocompatibility remain barriers. however, an ML approach could be a promising change.

• Input: Electrogram data, cardiac MRI, LGE-MRI scar segmentation
• Output: Predictive placement heatmap, arrhythmic vulnerability index
• Algorithm: 3D U-Net with attention layers, trained on >100 anonymized ECG/imaging datasets

🧬 how do we build it?

Proposed Clinical Trial Structure: Trial #1

• Animal Model: Yorkshire pigs, MI induced via LAD ligation (n=10 with patch, n=10 control)
• Implantation Site Determination: AI-based predictive modeling using real-time electrogram mapping
• Procedure: Patch applied epicardially under general anesthesia
• Endpoints:
  • Electromechanical coupling (optical mapping + ECG)
  • Fibrosis volume (histology)
  • Arrhythmia incidence (telemetry)
  • Histological evidence of iPSC-CM retention and alignment
• Reference Study: Wong RC et al., UHN (2020), “Transplantation of iPSC-derived cardiac tissue in swine post-MI improves functional recovery” – Canadian Journal of Cardiology

Proposed Clinical Trial Structure: Trial #2

• Eligibility Criteria: Adult patients (18–70) with ischemic cardiomyopathy, NYHA Class II–III symptoms, and documented infarction within 1 year
• Surgical Procedure:
  • Patch generated in GMP facility from autologous or allogeneic iPSCs
  • Conductive scaffold fabricated and seeded ex vivo, validated for sterility
  • AI-generated placement map used to guide epicardial placement
• Safety Monitoring:
  • ECG and Holter monitoring for 30 days post-implant
  • Echocardiogram at baseline, 30, 90, and 180 days
  • Cardiac MRI at 90 days to assess remodeling and patch integration

🧬 how do we scale?

proposed partners: Ted Rogers Centre for Heart Research (UHN + SickKids + U of T) – Comprehensive cardiology and iPSC expertise

🧬 behind the scenes:

This project combines regenerative engineering, AI, and material science. Illustrations were done through the use of ChatGPT, with supporting research from recent publications in cardiac bioengineering. Website + prototype by Elena Osipyan :)

🧬 references:

• Zhang, Donghui, et al. “3D Bioprinting of Cell-Laden Cardiac Patch for Heart Regeneration.” NPJ Regenerative Medicine, vol. 6, 2021. Link
• Ntagiantas, Evagelos, and Marco Pignatelli. “Estimating Cardiac Tissue Conductivity from Electrograms with Fully Convolutional Networks.” Semantic Scholar, 2022. Link
• Shin, Seung Rim, et al. “Electroconductive Biomaterials for Cardiac Tissue Engineering.” Frontiers in Bioengineering and Biotechnology, vol. 10, 2022. Link
• Khan, Tareq, et al. “Advances in Conductive Biomaterials for Cardiac Tissue Engineering: Design, Fabrication, and Functional Integration.” Polymers, vol. 17, no. 5, 2025. Link
• Liu, Yuxuan, et al. “Large Cardiac Muscle Patches Engineered From Human Induced-Pluripotent Stem Cell–Derived Cardiac Cells Improve Recovery from Myocardial Infarction in Swine.” Circulation, vol. 137, no. 16, 2018. Link
• Miyagawa, Shigeru, et al. “Safety Confirmation of Induced Pluripotent Stem Cell–Derived Cardiomyocyte Patches for Clinical Application.” NPJ Regenerative Medicine, vol. 8, no. 1, 2023. Link
• Menasché, Philippe. “Cell Therapy Trials for Heart Regeneration—Lessons Learned and Future Directions.” Nature Reviews Cardiology, vol. 15, 2018. Link
• Guo, Rui, et al. “Engineering a Conduction-Consistent Cardiac Patch with rGO/PLCL Electrospun Nanofibrous Membrane and Human iPSC-Derived Cardiomyocytes.” Bioactive Materials, vol. 20, 2023. Link
• Ting, Daniel S.W., et al. “Artificial Intelligence and Deep Learning in Cardiovascular Imaging.” Nature Reviews Cardiology, vol. 18, 2021. Link