Digital twin for presurgical assessment of cardiac ablation

Cardiac arrhythmias are irregularities of the heartbeat that can seriously affect the life of the patient, and can be fatal in severe cases. A typical treatment for serious cases of arrhythmias is the radiofrequency ablation based on catheterization. This is a minimally invasive therapy that has been practised for many years and proven to be effective. Ablation therapy is typically efficient, however insufficient treatment might occur resulting in a repetition of the procedure. Additionally, catheterization is not exempt from complications that can prolong the procedure and on rare occasions can even lead to the death of the patient. To eliminate complications and to provide a more efficient treatment, different ablation strategies and catheter technologies are being researched, the testing of which heavily relies on animal experimentation. Typically these experiments require large animals, such as pigs, sheep and cows, and are conducted either by using the extracted heart of the animal or by performing the experiment on an alive sedated animal. This research aims to provide accurate complex mathematical models to foster the use of computer simulations to optimize the efficiency and safety of catheterization, thus reducing animal experimentation. Using medical imaging data from patients which have been acquired and transferred following the ethical protocols and with the patients consent, a database of virtual human upper bodies and their internal organs will be created. These virtual bodies will be the digital twins of the patients, on which virtual therapies will be performed. In particular, computer designed catheters will be used and placed within the heart of the digital twin. Virtual catheterization will then be performed using the models developed in the project to accurately describe the biophysics and mechanics of the procedure. Virtual catheterizations will simulate the ablation of the different cardiac areas that are typically targeted by the therapy in the hospitals, providing a personalization framework for the assessment of the procedure.

Computer models of individual patients can show, before an operation, how a heart treatment will affect that specific person. In this project we built such "digital twins" for catheter ablation, a common treatment for atrial fibrillation, the most frequent heart rhythm disorder. We used them to study questions that are difficult, costly or impossible to answer with experiments on patients or animals. During ablation, a physician guides a thin tube (a catheter) through a blood vessel into the heart and destroys small areas of tissue that trigger the irregular rhythm. This is done either by heating the tissue with radiofrequency current or, in the newer pulsed field technique, with very short, strong electric pulses. The treatment works well, but the right settings depend on the patient and the procedure, and today they are chosen mostly from general experience. Together with cardiologists at the Hospital de la Santa Creu i Sant Pau in Barcelona, we turned medical scans of real patients into detailed three dimensional computer models of the heart and the surrounding body. On these virtual patients we simulated the electric current, the heating of the tissue and the resulting tissue damage. The results show that factors outside the heart matter. The position of the grounding pad on the skin changes how much of the energy actually reaches the heart. A patient's sex and body mass index also influence the delivered power, which suggests that uniform settings may not suit everyone equally well. We also showed how the angle and pressure of the catheter shape the treated area, and we developed a model that predicts not only the size of the treated area but also the risk of damage to the esophagus, which lies directly behind the heart and whose injury is a rare but serious complication. For the newer pulsed field ablation, we found that the simple rule used in most computer studies, a single electric field strength above which tissue dies, cannot explain the lesion shapes observed in experiments. This points to the need for more realistic models, which we are now developing. All methods were combined in AblaFEMx, a software package that will be released free of charge so that other researchers can use, check and extend it, and simulation data from our studies are openly shared. Computer studies of this kind can replace part of the animal experiments currently used to test new catheters and treatment strategies. The project also trained young researchers, including one completed doctoral thesis, and led to a new project with an Austrian company that develops heart simulation software, bringing these models closer to clinical practice.