|
a |
|
b/README.md |
|
|
1 |
# MI inverse inference |
|
|
2 |
Code for the inverse inference of infarcted area from ECG and MRI for myocardial infarction (MI) patients. This is achieved within a cardiac digital twin (CDT) framework, where the anatomical twinning personalizes the geometrical model, while functional twinning personalizes the electrophysiological model. |
|
|
3 |
|
|
|
4 |
 |
|
|
5 |
|
|
|
6 |
|
|
|
7 |
## Notes:<a id="Summary"/> |
|
|
8 |
1. *Cardiac_Personalisation-SenAnalysis* fold only contain partial code for the ECG simulation and the sensitivity analysis. For the full Eikonal-based ECG simulation code, please contact [Dr Julia Camps](mailto:julia.camps@cs.ox.ac.uk). |
|
|
9 |
2. *Cobiveco* fold only contrain partial code for converting biventicle mesh into cobiveco mesh. For the complete Cobiveco mesh reconstruction code, please visit [KIT-IBE Cobiveco Github repository](https://github.com/KIT-IBT/Cobiveco). |
|
|
10 |
|
|
|
11 |
## Package dependencies:<a id="Package dependencies"/> |
|
|
12 |
This repository is based on PyTorch, running on a computer with 3.50~GHz Intel(R) Xeon(R) E-2146G CPU and an NVIDIA GeForce RTX 3060. |
|
|
13 |
|
|
|
14 |
## Dataset:<a id="Dataset"/> |
|
|
15 |
Our network is trained based on UKB dataset, which contains paired multi-view MRIs and ECG data. We have reconstruced the cardiac geometry from the multi-view MRIs and converted it into cobiveco mesh. The scar/border zone area were assigned on the mesh and subsequently used for ECG simulation of MI patients. |
|
|
16 |
|
|
|
17 |
The data folder should be like: |
|
|
18 |
``` |
|
|
19 |
tree |
|
|
20 |
`-- UKB_clinical_data |
|
|
21 |
`-- patientID |
|
|
22 |
| |-- patientID_cobiveco_AHA17.vtu |
|
|
23 |
| |-- patientID_heart_cobiveco.vtu |
|
|
24 |
| |-- patientID_simulated_ECG_xxx_subendo.csv |
|
|
25 |
| |-- patientID_simulated_ECG_xxx_transmural.csv |
|
|
26 |
| |-- ... |
|
|
27 |
| |-- patientID_electrodePosition.csv |
|
|
28 |
|
|
|
29 |
``` |
|
|
30 |
|
|
|
31 |
## Citation:<a id="Citation"/> |
|
|
32 |
If you find this code useful in your research, please consider citing: |
|
|
33 |
``` |
|
|
34 |
@article{jounral/TMI/li2024, |
|
|
35 |
title={Towards Enabling Cardiac Digital Twins of Myocardial Infarction Using Deep Computational Models for Inverse Inference}, |
|
|
36 |
author={Li, Lei and Camps, Julia and Wang, Zhinuo and Banerjee, Abhirup and Rodriguez, Blanca and Grau, Vicente}, |
|
|
37 |
journal={IEEE Transactions on Medical Imaging}, |
|
|
38 |
year={2024} |
|
|
39 |
} |
|
|
40 |
|
|
|
41 |
``` |