SimICD : A Closed-Loop Simulation Framework For ICD Therapy
Hannah Lydon, Milad Kazemi, Martin Bishop, Nicola Paoletti · 2025
Read SimICD : A Closed-Loop Simulation Framework For ICD Therapy by Hannah Lydon, Milad Kazemi, Martin Bishop, Nicola Paoletti (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-01371.txt.
Introduction
Hannah Lydon and co-authors build a virtual test bench for implanted defibrillators, the devices that shock a racing heart back into rhythm. Their 2025 paper introduces SimICD, which couples the decision logic of a real device to detailed simulations of heart electricity. The loop lets a full therapy episode play out inside the computer.
Earlier in-silico tools could test diagnosis but not how treatment unfolds step by step during an episode. SimICD fills that gap across a cohort of virtual patients. Its signals and device responses track the logic of real products closely. Clinicians can then retune device settings for particular episodes in simulation first, long before any setting touches a patient.
Catalog record
| CREATOR | Hannah Lydon, Milad Kazemi, Martin Bishop, Nicola Paoletti |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 29 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-01371
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
SimICD : A Closed-Loop Simulation Framework For ICD Therapy Hannah Lydon, Milad Kazemi, Martin Bishop, Nicola Paoletti Abstract Virtual studies of ICD behaviour are crucial for testing device functionality in a controlled environment prior to clinical application. Although previous works have shown the viability of using in silico testing for diagnosis, there is a notable gap in available models that can simulate therapy progression decisions during arrhythmic episodes. This work introduces SimICD , a simulation tool which combines virtual ICD logic algorithms with cardiac electrophysiology simulations in a feedback loop, allowing the progression of ICD therapy protocols to be simulated for a range of tachy-arrhythmia episodes. Using a cohort of virtual patients, we demonstrate the ability of SimICD to simulate realistic cardiac signals and ICD responses that align with the logic of real-world devices, facilitating the reprogramming of ICD parameters to adapt to specific episodes. I Introduction Implantable Cardioverter Defibrillators (ICDs) are essential tools for terminating potentially fatal rapid arrhythmias (tachy-arrhythmias) in high-risk patients using a range of pacing and electric shock therapy interventions. Whilst ICDs have proven effective in improving patient outcomes [ 1 ] , they are susceptible to misdiagnoses and ineffective therapy delivery [ 2 , 3 ] , highlighting the need for robust testing of ICD decision algorithms prior to clinical application. Such vital testing can be conducted through in silico experiments, which have been playing an increasingly significant role in clinical research by enabling safer, more comprehensive testing without invasive procedures whilst also allowing researchers to evaluate novel algorithms prior to use in real patient devices. Whilst previous work has shown the viability of using virtual devices for diagnoses of arrhythmia episodes [ 4 , 5 ] , these attempts only consider open-loop signals and thus cannot reproduce the effects of the device’s therapy on the patient’s cardiac electrophysiology (EP). Hence, a significant research gap remains in modelling the interactions between realistic virtual devices and advanced cardiac EP simulations in a closed-loop fashion, capable of both prescribing and simulating therapy interventions [ 6 , 7 ] . Fig. 1 : Visualization of SimICD simulation pipeline. Cardiac arrhythmia episodes are modelled in OpenCARP [ 8 ] , the signals are extracted for analysis in the virtual ICD, and therapy decisions are relayed to the simulator. We address this gap by proposing SimICD , an automated closed-loop simulation environment (represented in Fig. 1 ) that utilises the abilities of open-source cardiac modelling software and the algorithms of one of the leading device manufacturers to model therapy interventions during simulated cardiac episodes. Related work We build on previous works [ 4 , 9 , 10 ] for the encoding of the Boston Scientific ICD algorithm but significantly expand the algorithm to be in line with the latest available device manuals and support the ability for the virtual ICD to perform live signal monitoring, prescribe therapy and perform re-detection. Our work also leverages some of the methods for modelling idealized infarcts and re-entrant circuits [ 11 , 12 ] but with a different research goal, namely studying closed-loop ICD therapy. The modelling of ATP therapy using the resources available in openCARP was primarily inspired by [ 13 ] and the heart meshes, ICD electrode placement and the stimulus sites used in the patient simulations were leveraged from previous cardiac modelling investigations [ 14 , 15 ] . In summary, our key contributions are as follows: • We present SimICD , a novel closed-loop simulation environment that enables realistic in-silico evaluation of multiple ICD therapies and several relevant cardiac episodes. Within this framework, we achieve: • Implementation of a realistic virtual ICD model using the latest device manuals from Boston Scientific [ 16 ] , one of the leading device manufacturers, • Generation of a broad cohort of virtual patients for ICD evaluations using state-of-the-art cardiac EP modelling software [ 8 ] . The codebase for the SimICD tool is publicly available at: https://github.com/janet-9/SimICD II Background II-A ICD Functionality ICDs monitor the electric signals in the heart via leads embedded within the cardiac tissue and a pulse generator, also acting as a far-field sensor, placed in the torso. In a dual chamber device, these signals are traces from the atria, ventricles and the far-field signal and in a single chamber device, only the ventricular and far-field signals are measured. We focus on a single chamber device in this work as we are concerned with modelling arrhythmias of a ventricular origin. The recorded signals are processed to form intracardiac electrograms (EGMs) that are analysed using internal device logic to diagnose potentially fatal arrhythmic episodes. These are typically classified as either Ventricular Tachycardia (VT), heart rates above 100 bpm, or Ventricular Fibrillation (VF), usually heart rates above 150 bpm and characterised by a lower amplitude, irregular rhythm [ 17 ] . VT, whilst considered the less severe of the two types, has the potential to develop into the (higher-risk) VF condition, and both episodes can result in sudden cardiac death if left untreated. A common pacing therapy protocol used by ICDs to treat these arrhythmias is Anti-Tachycardia Pacing (ATP): rapid low-voltage pulses delivered from an electrode embedded within cardiac tissue designed to interrupt the mechanisms of the arrhythmia to restore a healthy heart rate (known as normal sinus rhythm or NSR, which is typically defined as a rhythm with a bpm between 60 - 100 [ 18 ] ). If this intervention is unsuccessful, the device can proceed to deliver high-energy shocks, delivered via a current exchange between an electrode within the tissue and the pulse generator of the ICD, which increases the likelihood of termination but can cause discomfort to the patient and, if used excessively, have shown a correlation with worsened long-term patient outcomes [ 19 ] . As such, the majority of devices are designed only to treat episodes that are potentially fatal and use the lower energy therapy options first before escalating the intensity. The synthetic EGMs used for analysis by the virtual ICD in SimICD were generated using the state-of-the-art source cardiac modelling software OpenCARP [ 8 ] , which allows for flexible implementation of a variety of cardiac meshes and electrophysiology models. II-B Electrophysiology Model For the cardiac EP simulations, we used various models of a human bi-ventricular (BiV) system to simulate both healthy and arrhythmic episodes. During all simulations, conduction in the myocardium was modelled using the monodomain model. This model, defined by equations 1 , 2 , describes propagation via a single parabolic reaction diffusion equation under the assumption that the intra- and extracellular spaces have proportional conductivities. It was chosen compared to the more complex bidomain model [ 20 ] , which models the intra- and extracellular spaces separately because, due to its lower computational cost. The monodomain model, despite the reduced physiological accuracy, was found to have sufficient accuracy in simulating EGMs as it allowed the recovery of the extracellular potential, \phi_{e} , (the signal measured by ICD electrodes) by performing numerical integration at specific locations [ 21 ] . \nabla\cdot\bm{\sigma_{m}}\nabla V_{m}=\beta(I_{m}+I_{stim}). (1) \bm{\sigma_{m}}=\frac{\bm{\sigma_{i}}\bm{\sigma_{e}}}{\bm{\sigma_{i}}+\bm{\sigma_{e}}}. (2) In these equations, the tensor matrices \sigma_{i} , \sigma_{e} and values of I_{i} , I_{e} describe the conductivities and current densities of the intra and extracellular regions