In the Czech Republic, abdominal aortic aneurysm affects 2 to 3 percent of people over 60 years of age. In case the aneurysm ruptures, half of the patients die. While smaller aneurysms are usually only closely watched by physicians, larger ones are usually operated on. Physicians must then weight the risk of aneurysm rupture against the risks of the surgery itself. In the future, models of biomechanics of the BUT Faculty of Mechanical Engineering could help them in their challenging decision making. They were first tested by clinicians during the Covid crisis.

While in the spring the planned surgeries were getting postponed due to the COVID restrictions, clinicians who care for aneurysm patients faced the dilemma of whether the delayed procedures would endanger their patients’ lives. Clinicians of the St. Anne’s University Hospital of Brno therefore welcomed the help of an algorithm created by scientists of the BUT Faculty of Mechanical Engineering and VŠB-TU Ostrava. Out of twelve patients, surgeons immediately operated on two who, according to calculations, were at serious risk of rupture of their aneurysm. "An aneurysm rupture is actually an internal arterial bleeding. Although many people are eventually saved, everyone who is affected is on the verge of death," says Jiří Burša, Head of the Department of Tissue Biomechanics, who has been researching abdominal aortic aneurysm with his team for over ten years.

BUT experts focused on the question of how to make the prediction of an aneurysm rupture more accurate. "Not every aneurysm is to burst. According to the medical standards, an aneurysm larger than five centimeters should be operated, but there also exist cases of patients who used to live with a ten-centimeter aortic aneurysm. These were usually people who refused surgery or could not be operated on due to their age or other circumstances. We are talking especially about elderly people who suffer from a number of other health issues and end up usually dying of a completely different reason than a ruptured aneurysm. In this case, the postponed operation may lead to improvement of the patient's quality of life," Burša explains. In other words, clinicians often weigh whether it is riskier for the patient to undergo surgery or to continue living with the aneurysm.

Mathematical models of biomechanics from Brno could help them. "Our goal is to find the weakest link in the calculation chain that causes the highest error rate and attempt to improve it. Thanks to this, we will move forward and look for another source of inaccuracies in the new model and so on and so forth. We have solved about five similar partial problems during our ten years of research, and the accuracy of our model has increased by tens of percent. A significant improvement in the prediction accuracy has been achieved especially in patients who were supposed to be operated on and were not, but still have not suffered an aneurysm rupture. In other words, they would have been operated unnecessarily," says Burša.

Welcome clot

For example, the researchers focused on the question of how to obtain the default geometry for the model from CT scans. This geometry must be obtained under pressureless conditions, as during imaging the aneurysm is, like all arteries, distorted by blood pressure, which stretches them by up to 15 mm and strongly modifies its shape. "We have developed an algorithm that is able to reconstruct the pressureless shape. We have also found a way how to reduce uncertainty with respect to the mechanical properties of the aorta in this situation, when prior surgery it is not possible to perform a mechanical test similar to that commonly done for industrial materials. This way, we are trying to minimize the inaccuracy of our calculations," Burša explains. The basic settings of the mechanical models were determined using pig aortas, which were supplied by the Steinhauser slaughterhouse. These samples were also used to develop methods to determine the structure of vascular tissue, primarily the arrangement of collagen fibers, which provide aortas with necessary rigidity and strength. Only then come samples from operated patients or from the autopsy room. This methodology is being further developed within the current project, which is already focused on another issue – the risk of rupture of the sclerotic plaque in the carotid artery and subsequent stroke.

Researchers also looked at the significance of the so-called intraluminal thrombus, i.e. the blood clot that fills the aneurysm. For a layman, clots in the bloodstream are synonymous with heart attack, stroke or embolism. In the case of an aneurysm, however, they can play a positive role. "In the treatment of cerebral artery aneurysms, the clot is induced in a targeted manner. The aneurysm, which can be imagined as a blind channel of a river, fills up due to a blood clot and the thrombus gradually turns into stiffer tissue, although not as hard as a scab on a bruised knee. Thanks to this, it relieves pressure on the artery wall and reduces the risk of its rupture. In abdominal aneurysms, a thrombus usually forms spontaneously and has a positive benefit here as well. It can be up to several centimeters thick and acts as a network of fibers that helps the artery wall transfer blood pressure. Therefore, thrombus in arterial aneurysms is not a scary word," Burša describes.

Unplanned experiment

Ten years of effort resulted in a mathematical model for predicting abdominal aortic rupture, which tries to best include the individual characteristics of each patient. The input is several categories of data: geometry obtained from CT images of the patient, properties of the material that the researchers examined on pig aortas and then samples taken directly during surgery, and other information about the patients, such as their blood pressure.

"We ran a seven-year study in which we followed patients who were supposed to be operated on but for various reasons were not. We created computer models of the probability of aneurysm rupture and investigated the actual development of their aneurysm over the next few years. Subsequently, we used relatively complex statistical methods to find whether our prediction was correct. It was a blinded study, so the colleagues did not know which of the aneurysms subsequently burst and which did not. The results showed us that, using our methodology, the prediction has really become more accurate," adds Burša, commenting on the details of the unique study, which involved 43 patients from two domestic hospitals and which scientists published in a prestigious journal in February this year.

A team of the clinicians of St. Anne’s University Hospital of Brno, led by surgeon Robert Staffa, is now hoping to achieve the possibility to perform a so-called preclinical experiment, which represents a necessary step to introduce the model into the routine clinical practice. But the Covid restrictions imposed this spring brought along an unexpected chance for this research as physicians applied the algorithm to patients whose surgery had to be postponed due to the pandemic. At that time, they turned to Stanislav Polzer, who started his research at BUT and now continues it at VŠB-TU Ostrava, where he founded his own research group that further develops the algorithms. "For the first time in history, dozens of patients who have been concluded to have a safe aneurysm are undergoing an unplanned experiment that will test our calculations," concludes Burša.