Using machine learning to predict structural factors using electrogram characteristics in intact hearts
Lead Research Organisation:
Imperial College London
Department Name: National Heart and Lung Institute
Abstract
The study aims to achieve this by using contact electrodes on ex vivo whole hearts to form an extracellular electrogram. Abnormalities simulating atrial and ventricular fibrillation will be generated in the hearts to assess differences and changes in morphology. Machine learning algorithms will be developed for the analysis and characterization of the electrograms generated.
People |
ORCID iD |
Nicholas Peters (Primary Supervisor) | |
Joseph Brook (Student) |
Publications
Brook J
(2020)
Development of a pro-arrhythmic ex vivo intact human and porcine model: cardiac electrophysiological changes associated with cellular uncoupling.
in Pflugers Archiv : European journal of physiology
Kim MY
(2023)
Immunohistochemical characteristics of local sites that trigger atrial arrhythmias in response to high-frequency stimulation.
in Europace : European pacing, arrhythmias, and cardiac electrophysiology : journal of the working groups on cardiac pacing, arrhythmias, and cardiac cellular electrophysiology of the European Society of Cardiology
Scott AD
(2022)
Development of a cardiovascular magnetic resonance-compatible large animal isolated heart model for direct comparison of beating and arrested hearts.
in NMR in biomedicine
Description | We have shown that intact Human and Porcine hearts can be successfully restarted ex-vivo using a Langendorff apparatus. The experimental apparatus can be used for live electrophysiological studies, bridging a translational gap between smaller simpler laboratory models and clinical research. Using supervised machine learning, electrograms recorded from human and porcine hearts can be accurately classified between normal, or featuring abnormalities found to cause arrhythmia. When trying to differentiate between normal and 3 different arrhythmogenic abnormalities, supervised machine learning can also be used to correctly classify the specific abnormality responsible. |
Exploitation Route | More arrhythmogenic abnormalities could be appended to the model, showing that more abnormalities an be can be accurately predicted. The resultant machine learning model could also be tested against clinical data, to test the translational ability of the model to predict the presence of arrhythmogenic abnormalities in clinical electrograms. |
Sectors | Healthcare |
URL | https://link.springer.com/article/10.1007/s00424-020-02446-6 |
Description | BHF/ICTEM 4 year Phd Studentship |
Amount | £194,000 (GBP) |
Funding ID | FS/4yPhD/F/21/34165 |
Organisation | British Heart Foundation (BHF) |
Sector | Charity/Non Profit |
Country | United Kingdom |
Start | 08/2022 |
End | 09/2026 |
Description | Inferring structural properties of the myocardial substrate from multipolar electrogram recordings using deep learning |
Amount | £101,220 (GBP) |
Funding ID | 222845/Z/21/Z |
Organisation | Wellcome Trust |
Sector | Charity/Non Profit |
Country | United Kingdom |
Start | 08/2020 |
End | 09/2023 |
Title | Human Langendorff Perfusion System |
Description | The Langendorff ev vivo heart perfusion system has existed previously for rodent hearts. We have built a Langendorff perfusion system cabable of sustaining explanted human hearts. The success of translating rodent findings to the clinic is hindered by species differences. This ex vivo human Langendorff system allows for systematic, controlled investigation of human hearts. |
Type Of Material | Model of mechanisms or symptoms - human |
Year Produced | 2018 |
Provided To Others? | No |
Impact | Having built the initial system we are currently adding on-line monitoring systems. After this the system will be a research tool for investigating electrophysiological parameters in human hearts. |
Description | NHSBT donor hearts |
Organisation | NHS Blood and Transplant (NHSBT) |
Country | United Kingdom |
Sector | Public |
PI Contribution | We have acquired ethics and are using control donor heart samples (when not suitable for transplant) provided by NHS Blood and Transplant (NHSBT) for electrophysiological characterisation. |
Collaborator Contribution | NHSBT are managing provision of donor hearts to this research when not suitable for transplantation. |
Impact | The partnership has allowed the research to incorporate comparision against control samples. |
Start Year | 2016 |
Description | Porcine hearts and the Royal Veterinary College |
Organisation | Royal Veterinary College (RVC) |
Country | United Kingdom |
Sector | Academic/University |
PI Contribution | Implementation of 3Rs by using end of study pigs from the RVC. |
Collaborator Contribution | Provision of pigs at end of study and expertise in heart removal. |
Impact | Implementation of 3Rs |
Start Year | 2017 |