Mixed Reality‐Assisted Thoracic Endovascular Aortic Repair: A Retrospective Study on Efficacy and Safety

ABSTRACT

Objective

To investigate the efficacy and safety of Mixed Reality (MR)‐Assisted Thoracic Endovascular Aortic Repair (TEVAR).

Methods

We retrospectively analyzed 46 patients with aortic dissection who underwent TEVAR. In the Control Group, Computed Tomography Angiography (CTA) was performed on the thoracic and abdominal aorta, after which surgery was performed based on traditional 2D imaging data. In the observation group, the Star Map HoloLens Image System was used for data processing and 3D modeling, and preoperative analysis and intraoperative path guidance were conducted with the MR Microsoft HoloLens Headset. The communication time, satisfaction, and anxiety after communication, as well as the operation duration, intraoperative blood loss, postoperative complications, and rehabilitation of these two groups were analyzed.

Results

This study aimed to evaluate the perioperative outcomes of MR‐guided stent placement. A total of 46 patients were equally divided into the MR group and control group (n = 23 each), with data analyzed via non‐parametric tests. Compared with the control group, the MR group exhibited significantly shorter operation duration (70.91 ± 8.533 vs. 77.48 ± 8.474 min, Z = −2.785, p = 0.005), fewer fluoroscopy times (3.74 ± 1.214 vs. 5.61 ± 1.530, Z = −3.813, p < 0.001), and less stent adjustments (1.65 ± 1.152 vs. 2.91 ± 1.411, Z = −2.765, p = 0.006), suggesting MR guidance may optimize perioperative efficiency.

Conclusion

Multimodal MR has an important auxiliary role in patient education, surgical planning, and accurate positioning of the surgical approach.

This study demonstrates the significant clinical utility of Mixed Reality (MR) in Thoracic Endovascular Aortic Repair (TEVAR). Our findings reveal that MR‐assisted workflows enhance patient comprehension and satisfaction preoperatively and provide intuitive intraoperative navigation. This technology represents an innovative shift from traditional 2D imaging to an interactive 3D model for surgical planning and execution, potentially improving procedural precision.

1. Introduction

Thoracic Endovascular Aortic Repair (TEVAR) is one of the most common aortic interventional surgeries and is mainly used for the diagnosis and treatment of thoracic aortic aneurysm, rupture of thoracic aortic aneurysm, symptomatic aortic penetrating ulcer or intramural hematoma, traumatic thoracic aortic disconnection, Type B aortic dissection, and other diseases. TEVAR technology refers to a surgery in which a stent covered with a polymer composite film is delivered to an aneurysm through the femoral artery pathway to “reinforce” and “isolate” the blood vessels involved in the aneurysm, thereby reducing the risk of aneurysm rupture. TEVAR was first performed in patients who could not tolerate open surgery, the advantages of which include less trauma, no need for artery occlusion, less blood loss, reduced occurrence of end‐organ ischemia, and faster recovery after surgery. In a clinical study comparing TEVAR with open surgery, it was found that TEVAR is safer, with fewer complications in the heart, lung, and blood vessels, but the neurological complications were comparable to those of open surgery. Currently, TEVAR is primarily used for Stanford Type B aortic dissection in China. The placement of a covered stent to seal the primary rupture can promote the organization of the false lumen, prevent the rupture of the dissection, and improve the blood supply of the affected organs (Huang 2013).

Traditional imaging examinations include computed tomography angiography (CTA) of the aorta. Prior to surgery, relevant parameters were measured with CTA, including the length, diameter, angle, configuration of the proximal neck, presence of severe calcification (especially annular calcification) and mural thrombus, diameter, length, and angle of the aortic anchor area, maximum diameter of the tumor mass and diameter of the tumor cavity, degree of distortion and stenosis of the pathway of introduction, and opening of branch vessels. This results in higher requirements for the 3D space perception ability of the operators. Mixed Reality (MR) is an extension of Augmented Reality (AR) and a further development of Virtual Reality (VR). It generates virtual objects that do not exist in the real environment through computer graphics technology and visualization technology, and overlays virtual objects onto the real environment through sensor technology. In the new visualization environment, physical and digital objects coexist and interact in real‐time. Studies on craniofacial tumors, orthopedics, and other fields have shown that MR technology has extensive and in‐depth application value; however, its application in interventional aortic surgery is less studied, especially in type A aortic dissection. This study further explores the application value of this technology in TEVAR surgery of type A aortic dissection patients.

2. Material and Methods

The participants of this retrospective study were patients with aortic dissection who underwent TEVAR surgery from March 2021 to December 2021 in the Department of Cardiovascular Surgery of the Second Hospital of Hebei Medical University. This study was conducted with approval from the Ethics Committee of The Second Hospital of Hebei Medical University (Reference number: 20210978). This study was conducted in accordance with the declaration of Helsinki. Written informed consent was obtained from all participants.

2.1. Study Design

Inclusion criteria: (1) Aged 30–60 years old; (2) Thoracic aortic aneurysm > 5.0 cm or increased by 0.5 cm or more within 6 months, or acute Type B aortic dissection with end‐organ ischemia, or chronic Type B aortic dissection with degenerative changes of an aneurysm. Exclusion criteria: (1) Cases where there is involvement of the ascending aorta and the brachiocephalic artery of the aortic arch, or severe regurgitation of the aortic valve; severe lesions or diameters > 40 mm in the aortic anchoring area, as well as severe atherosclerosis in the iliac artery and abdominal aorta, are not applicable for the stent delivery system. (2) Complications with fatal organ diseases other than heart or artery. (3) Uncontrolled systemic infectious diseases and active stages of connective tissue disease.

2.2. Patients Characteristics

From March 2021 to December 2021, 46 patients underwent TEVAR surgery. The disease diagnoses of the enrolled patients included: acute Stanford type B aortic dissection (onset ≤ 14 days), chronic Stanford type B aortic dissection with aneurysmal degeneration, thoracic aortic aneurysm (degenerative/atherosclerotic), penetrating atherosclerotic ulcer, and intramural hematoma, among others. They were divided into the Observation Group (MR Group) (n = 23) and Control Group (n = 23) according to the intervention. The Control Group received CTA of the thoracic and abdominal aorta and then underwent surgery based on traditional 2D imaging data. Based on this, the Star Map HoloLens Image System was used for data processing and 3D modeling for the MR Group, and preoperative analysis and intraoperative path guidance were conducted with the MR Microsoft HoloLens Headset. All patients underwent femoral artery ultrasound. Data including age, sex, body mass index, hypertension, coronary heart disease, diabetes and other previous histories, and liver and kidney function indicators were obtained from history inquiry and relevant preoperative examinations of each patient (Table 1).

TABLE 1.

Comparison of general data of the two groups of patients.

	MR group	Control group	t/χ 2	p
n	23	23
Gender	17/6	17/6	0	1.0
Age	45.78 ± 8.847	48.43 ± 7.323	−1.108	0.274
Hypertension	16/7	15/8	0.099	0.753
Coronary heart disease	2/21	2/21	0	1.0
Diabetes	0/23	1/22	1.0	0.5
Creatinine (μmol/L)	91.52 ± 27.80	98.36 ± 42.91	−0.641	0.525
GPT a (U/L)	21.26 ± 11.02	17.18 ± 10.98	1.259	0.215
GOT b (U/L)	33.11 ± 41.76	32.18 ± 34.54	0.082	0.935

^a GPT—Glutamic Pyruvic Transaminase.

^b GOT—Glutamic Oxaloacetic Transaminase.

2.3. Thoracic Endovascular Aortic Repair

In both groups, patients were monitored for vital signs before surgery. The peripheral arterial pressure was controlled at 100–120 mmHg, and the heart rate was 60–75 beats/min. The patients were treated with analgesia and hypotension. For patients with hypertension, 5 mg of propranolol was administered intravenously intermittently, and sodium nitroprusside was administered intravenously to adjust the dripping speed and reduce the blood pressure to the clinical treatment index. It should be noted that hypertensive patients with obstruction of the main branch of the aorta should not be treated for hypotension, as hypotension can aggravate ischemia. Antihypertensive drugs should not be used in patients with low blood pressure; however, propranolol can be used to reduce myocardial contractility. Other preoperative preparations: (1) iodine allergy test; (2) routine examination of hematuria and stool, liver and kidney function, and prothrombin time; (3) skin preparation at the puncture site; and (4) fasting for 6 h before general anesthesia.

During the operation, the patient was positioned in the supine position, and the bilateral inguinal area and abdomen were disinfected and paved with a cloth. If the brachial artery approach was required, the arm was also disinfected. The left arm was abducted with slight rotation to better expose the aortic arch. General anesthesia, regional anesthesia, or local anesthesia was selected according to the preoperative risk assessment. Continuous intraoperative arterial pressure monitoring was required because of possible arterial perforation or tearing. Blood pressure was strictly controlled when the covered stent was placed (straight tube laminated bracket). Hypotension, which may aggravate spinal cord ischemia caused by covering the related arteries with covered stents, was avoided during the operation. Peripheral venous channels were established routinely. All stents used in both groups were 200 mm in length and 32–36 mm in diameter. The distal end of the stent was placed at the middle thoracic aorta.

An open incision was made in the femoral artery to separate the tissue layer by layer to expose the common femoral artery under the inguinal ligament. A 4 Fr sheath was inserted and fluoroscopy was performed to confirm the correct position. After confirming the establishment of the percutaneous pathway, the sheath was replaced with a 6 Fr sheath. After the passage was established, a hydrophilic‐coated guidewire and catheter were introduced. The patients were treated with heparin anticoagulation. The stiff guide wire was introduced into the aortic arch and the pigtail catheter was introduced for aortic angiography. The distance between the opening of the left subclavian artery and aneurysm or dissection was determined at the proximal and distal anchorage areas in the left anterior oblique position. The level of distal aortic anastomosis in the debranching group was between the innominate and common carotid arteries. After confirming the size of the anchoring area and its anatomical relationship with important structures, the appropriate graft was selected and the sheath was replaced. The contraction pressure drop reached 100 mmHg, and the conveying system reached the predetermined position. Adenosine was given during stent placement to obtain higher release accuracy. The balloon was introduced to allow the stent to adhere completely to the arterial wall, and angiography was conducted again after the release was completed. The femoral artery was sutured layer by layer.

2.4. Mixed Reality Technology

2.4.1. Instruments and Software

The instruments included Philips 128‐row CT machine, Star Map HoloLens imaging system, and Microsoft HoloLens Headset. The Star Map HoloLens Image System was composed of hardware including a data processing computer, USB electronic dongle, HoloLens 3D image space station, background server and router, and software including the Star Map workstation software, HoloLens client software, and Pad control terminal software.

2.4.2. 3D Modeling

Original CTA data of the thoracic and abdominal aorta was imported into 3D image processing software for 3D digital modeling (Figure 1). According to different gray values displayed by DICOM format image data, the aorta, tear and rupture, false lumen, and adjacent tissues were observed, identified, marked, and segmented in an all‐round manner, and STL files were output (Figure 2). The 3D reconstruction image and MR were combined to simulate the real image of the human body to conduct 3D imaging of the blood vessels of the patient.

FIGURE 1.

3D digital model.

FIGURE 2.

Simulated real image.

2.5. Patient Education

Wearing the Microsoft HoloLens Headset, the morphology of blood vessels can be displayed to surgeons and patients and can be rotated up and down as needed to stereoscopically present the condition. Surgeons can more specifically and vividly explain the operation process, operation risks, perioperative precautions, and other details to patients and their families (Figure 3).

FIGURE 3.

3D digital model of lesion site.

2.6. Surgical Registration and Navigation

The 3D reconstruction model was loaded onto the HoloLens Headset, and the precise targeted surgical path of TEVAR was customized. The patient was placed in a supine position, marked with bone landmarks (the same landmarks as in the preoperative image examination), and a single scannable QR Code was pasted on the bone landmarks. The operator wore the MR headset (Microsoft HoloLens Headset). By using the HoloLens Headset image recognition technology, scanning the QR Code as the spatial positioning coordinate, the real‐time spatial positioning point of the MR image was fixed with the matching point of the bone marker point, and the MR image was completely overlapped with the solid aorta, after some refining and adjustment. Under the guidance of MR technology, after general anesthesia and sterile sheets were disinfected, the operator used the 3D digital model and the MR virtual image that was observed and built using the HoloLens Headset to judge the pre‐placement section of the surgical stent (Figure 4) and avoid the location of branch vessels of important tissues and organs. With the help of an X‐ray, the vascular stent was released accurately (Figure 5), the femoral artery was sutured, and the incision was then sutured in alignment. The operation was successfully completed (eliminating the lesion without post‐operative complications).

FIGURE 4.

3D digital model after intracavitary stent implantation.

FIGURE 5.

CT angiography of aorta after treatment.

It should be noted that this registration method is based on bony anatomical landmarks and is intended to provide the surgeon with an overall spatial orientation reference, rather than achieving pixel‐level precise overlay. During the procedure, the anesthesiologist was instructed to maintain the patient on breathing or to temporarily suspend ventilation during registration and key assessment time points to minimize the impact of respiratory motion on image matching accuracy. The final stent deployment remained entirely dependent on intraoperative digital subtraction angiography (DSA) as the gold standard guidance, with the MR model serving as a tool to enhance the surgeon's spatial awareness, rather than functioning as an independent real‐time navigation system.

2.7. Evaluation of Clinical Outcomes

Surgical all‐cause mortality was defined as death from any cause within 30 days, including death after discharge from the hospital, and any death occurring during hospitalization during the operation. Surgical parameters included operative time, number of intraoperative stent adjustments, number of intraoperative fluoroscopic exposures, total contrast medium volume, and the incidence of postoperative complications such as endoleak, spinal cord injury, cerebrovascular accident, retrograde type A aortic dissection, and stent collapse.

“Number of intraoperative stent adjustments” was explicitly defined in this study as any deliberate manipulation performed after the delivery system had been positioned, from the initial deployment attempt until complete stent release, aimed at optimizing the stent position relative to the landing zone (e.g., the ostium of the left subclavian artery). This included advancement or withdrawal of a partially opened stent‐graft, or complete recapture and repositioning for re‐deployment. Simple guidewire reshaping, catheter exchanges, or post‐dilation ballooning were not counted as stent adjustments.

In addition, the evaluation of MR technology application by the operator was taken as the observation index. There are three levels of matching degree between the large vessel disease model under 3D modeling and the vessel under DSA: good, average, and poor. The score for intraoperative navigation of MR technology is 0–10, of which 0 represents no effect or negligible effect, and 10 represents a very useful and critical effect.

2.8. Statistical Analysis

SPSS 26.0 statistical software was used for analyzing the data. The measurement data with normal distribution are expressed as mean ± standard deviation (SD), and the measurement data with non‐normal distribution are expressed as median; statistical description of counting data is expressed in frequency and percentage, and the chi‐square test was used for comparison. A p‐value of 0.05 was considered significant.

3. Results

All operations were successfully completed, and stent placement in the MR Group was smooth under the guidance of the preoperative planning route presented in the MR Headset.

The comparison of general demographic, clinical, and laboratory data between the MR group and the control group. Both groups had a sample size of 23 patients (n = 23 each) (Table 1). In terms of demographic characteristics, the gender distribution was identical in the two groups (17:6, with no specification of sex categories), and the difference was not statistically significant (χ ² = 0, p = 1.0). The mean age of patients in the MR group was 45.78 ± 8.847 years, while that in the control group was 48.43 ± 7.323 years; no significant intergroup difference was observed (t = −1.108, p = 0.274). Regarding comorbidities, 16 patients in the MR group and 15 in the control group had hypertension (16/7 vs. 15/8), with no significant difference between groups (χ ² = 0.099, p = 0.753). The prevalence of coronary heart disease was the same in both groups (2/21, χ ² = 0, p = 1.0). For diabetes, none of the patients in the MR group had the condition (0/23), whereas 1 patient in the control group did (1/22), and this difference was not statistically significant (p = 0.5). For laboratory indicators, the mean serum creatinine level was 91.52 ± 27.80 μmol/L in the MR group and 98.36 ± 42.91 μmol/L in the control group (t = −0.641, p = 0.525). The mean glutamic pyruvic transaminase (GPT) level was 21.26 ± 11.02 U/L in the MR group versus 17.18 ± 10.98 U/L in the control group (t = 1.259, p = 0.215). Additionally, the mean glutamic oxaloacetic transaminase (GOT) level was 33.11 ± 41.76 U/L in the MR group and 32.18 ± 34.54 U/L in the control group, with no significant intergroup difference (t = 0.082, p = 0.935).

The perioperative indicators between the two groups, analyzed using non‐parametric tests with Z‐statistics (Table 2). The mean operation duration in the MR group was 70.91 ± 8.533 min, which was significantly shorter than the 77.48 ± 8.474 min in the control group (Z = −2.785, p = 0.005). The MR group also had significantly fewer fluoroscopy times (3.74 ± 1.214 vs. 5.61 ± 1.530 times, Z = −3.813, p < 0.001) and fewer stent adjustments (1.65 ± 1.152 vs. 2.91 ± 1.411, Z = −2.765, p = 0.006) compared to the control group. However, there was no significant difference in postoperative hospital stay between the two groups: the mean stay was 2.52 ± 0.593 days in the MR group and 2.65 ± 0.647 days in the control group (Z = −0.677, p = 0.498).

TABLE 2.

Comparison of perioperative indicators between the two groups.

Index	MR group	Control group	Z	p
Duration of operation (min)	70.91 ± 8.533	77.48 ± 8.474	−2.785	0.005
Fluoroscopy times (Times)	3.74 ± 1.214	5.61 ± 1.530	−3.813	< 0.001
Number of stent adjustments	1.65 ± 1.152	2.91 ± 1.411	−2.765	0.006
Postoperative hospital stay (D)	2.52 ± 0.593	2.65 ± 0.647	−0.677	0.498

It should be emphasized that this score reflects the surgeon's assessment of the technology's usability from the physician's perspective, rather than a direct measurement of patient comprehension.

Pre‐discharge CTA follow‐up revealed no evidence of type I, type II, or type III endoleaks in either group, nor were there any severe complications such as spinal cord injury, cerebrovascular accident, stent migration, retrograde type A aortic dissection, or stent collapse. Although no endoleaks were observed in this cohort, given the relatively small sample size and the fact that follow‐up was limited to the hospitalization period, these findings warrant further validation in studies with larger cohorts and longer‐term follow‐up.

4. Discussion

MR is a further development of VR and AR. It builds an interactive feedback information loop between the real world, the virtual world, and users by presenting virtual scene information in the real scene to enhance the realism of the user experience (Mitsouras et al. 2020; Olivieri et al. ²⁰¹⁵). With the help of MR, real‐time comparison and verification of anatomical data can be obtained, and the accuracy and strategy of surgery can be controlled throughout the entire process. Currently, this technology is widely used in adjuvant treatment for tumor surgery. For complex diseases, such as porta hepatis cholangiocarcinoma and head and neck tumors, MR can clearly show the vascular structure around the tumor, improve the protection of important blood vessels, and greatly help the complete resection of tumors (Yan et al. 2021; Tang et al. ²⁰¹⁷). The discussion on the visible application of MR in the field of complex congenital heart disease shows that it has obvious advantages in the preoperative and intraoperative evaluation of complex congenital heart disease, and can optimize the operation strategy, shorten the operation duration, and reduce surgical trauma (Zhang et al. 2019). In this study, MR was used for digital 3D modeling through CTA image data of the patient's aorta, and an individualized and accurate surgical treatment plan was formulated. The 3D anatomical structural model was quickly mapped and restored in a real‐world operating room. During the operation, a gesture operation model was used to conduct a multidimensional and omnidirectional quantitative 3D spatial analysis of organs and lesions. The key quantitative parameters of the surgical site and 3D information related to incisions were measured in real‐time to guide the incision operation in real‐time during the operation, to match and fuse the real tissue and preoperative image data during the operation, and ensure the individualization and accuracy of TEVAR surgery.

TEVAR surgery is safer, with fewer postoperative cardiac, pulmonary, and peripheral vascular complications and with the advantages of less trauma, less blood loss, and faster postoperative recovery (Zeeshan et al. 2010; Torsello et al. ²⁰¹⁰; Brunt et al. ²⁰¹¹).

At present, commonly used clinical treatment strategies are as follows: (1) “Chimney” technology, that is, the technology of opening branches with another stent after covering some branches of the aortic arch with a covered stent, is mainly used for reconstructing the covered left subclavian artery or left common carotid artery in TEVAR, and it has also been reported to be used for the innominate artery (Zhu et al. 2013). (2) “Fenestration” technology, including in situ window opening technology and customized window opening bracket (Boersma et al. 2012). (3) “Scallop” technology. (4) Branched stent placement, can be single‐branch or multi‐branch (Wang and Li 2005). (5) “Hybrid” surgery, that is, endoaortic repair combined with surgical bypass surgery (Koullias and Wheatley 2010); is mainly used for patients who cannot undergo traditional surgery or TEVAR treatment, especially elderly patients with severe heart, lung, kidney, and other multiple organ diseases. The emergence of “hybrid” surgery has led to such patients regaining the opportunity of treatment, and greatly improved their prognosis.

In this study, the “number of stent adjustments” was explicitly defined as repositioning maneuvers performed prior to stent deployment. Although this metric is itself a “process measure” rather than an “outcome measure”, our study found that the MR group had a significantly reduced number of stent adjustments, which directly led to two clinically meaningful improvements: first, a reduction in intraoperative fluoroscopy, thereby decreasing radiation exposure risk for both patients and medical staff; and second, a significant reduction in intraoperative contrast medium volume.

Given that contrast medium volume is an independent risk factor for predicting contrast‐induced acute kidney injury (CI‐AKI), this finding carries clinical relevance. Although there were no statistically significant differences in postoperative renal function indicators between the two groups in this study, MR technology, by optimizing deployment precision and reducing the need for repeated angiographic confirmation, theoretically has the potential to reduce the incidence of CI‐AKI. This advantage may be particularly pronounced in high‐risk patients with pre‐existing renal insufficiency.

Therefore, although stent adjustment serves as a surrogate endpoint, its close correlation with patient safety (radiation exposure and contrast burden) makes it a valuable metric for assessing intraoperative efficiency and safety.

TEVAR significantly reduced the incidence of early complications and mortality when compared with open surgery (Bavaria and McCarthy ²⁰¹⁹). However, endovascular operation of the diseased aorta still has a high risk, including a series of fatal and disabling complications, such as endoleak, spinal cord injury, cerebrovascular accident, retrograde Type A dissection, and stent collapse, which are closely related to the conditions of the aortic tube wall, as well as complications in the selection of stent size and the correct release of the stent during operation. The branched arch device model depicted in the figures of this study represents an application scenario of MR technology in planning such complex anatomies. The patient ultimately underwent successful treatment via TEVAR, demonstrating the important role of the MR model in helping surgeons understand the spatial relationships of branch vessels and plan the optimal projection angle.

Endoleak was one of the most common complications and was the main observation index in this study. Endoleak refers to blood flowing into the aneurysm cavity after endovascular repair of the thoracic aorta. There are five types of endoleaks after TEVAR—Type I is anchored area endoleak; Type II is finger blood flows back into the tumor through the collateral artery; Type III is endoleak caused by support structure failure; Type IV refers to the blood flow entering the tumor cavity through the mesh on the fabric; Type V is of internal tension‐type, which refers to the increase of pressure in the tumor cavity and expansion of cyst cavity, but there is no obvious leakage of contrast agent (Sattah et al. 2018). Morales et al. reported that the incidence of endoleak after endovascular repair of the thoracic aorta for thoracic aortic aneurysm was 19.5%, and mainly included Type I endoleak and Type II endoleak (Morales et al. 2008).

In this paper, 23 patients with various aortic surgical diseases who underwent MR‐assisted intraoperative navigation are discussed. In the process of surgical assistance, MR technology could not only overlap the virtual image with the surgical field to achieve an accurate display of important anatomical structures, but also could be used to move, enlarge, shrink, and rotate the 3D projection to enrich the intraoperative observation angle, thereby further improving the visualization effect.

Based on the evaluation of the application of MR technology by the operators, the operators agreed that MR technology can help surgeons determine the location of aortic dissection rupture quickly and accurately; however, there is still some potential for improvement in the intraoperative navigation process, including virtual images that interfere with the visual effect of surgeons in the real environment, hand gestures that are difficult to master, and overly heavy equipment that cannot be worn for a long time during the operation. In general, MR technology has a high diagnostic value in the field of large blood vessels, as it can help comprehend their morphology. The operator clearly recognized the valuable role of MR technology in patient education and navigation in large‐vessel interventions. It is believed that MR technology played a very good auxiliary role in diagnosing and treating large vessel diseases.

It should be particularly noted that the MR technology applied in this study is not a real‐time imaging system; its three‐dimensional model is constructed based on preoperative CTA data, reflecting the patient's static anatomy at a single point in time. Consequently, a certain degree of discrepancy inevitably exists between the static MR model and the dynamic real‐time anatomy during surgery. In the procedural workflow of this study, the role of MR imaging was not to provide millimeter‐level real‐time “navigation,” but rather to serve as an enhanced spatial reference tool. Its value was primarily manifested in two aspects: ① Preoperatively, it helped surgeons intuitively comprehend complex anatomical structures and optimize surgical plans (e.g., selecting the optimal stent size and projection angle); ② Intraoperatively, the virtual model overlaid onto the surgical field enhanced the surgeon's perception of the spatial relationship between the target vessels and surrounding tissues, enabling faster and more confident positioning of guidewires, catheters, and stents under fluoroscopy, thereby reducing exploratory manipulations and repeat angiography. We acknowledge that the current registration method is based on bony anatomical landmarks and does not address the issue of real‐time displacement caused by respiratory motion. Therefore, expressions such as “precise positioning” in this manuscript should be understood as relatively accurate operational decisions achieved through enhanced spatial awareness, rather than real‐time pixel‐level fusion with dynamic anatomy. Future technological development should focus on integrating MR with intraoperative real‐time image fusion technologies, as well as respiratory gating or tracking systems, to achieve truly dynamic and precise navigation.

The results of this study demonstrate that MR‐assisted TEVAR was associated with shorter operative time, fewer fluoroscopic exposures, and reduced stent adjustments, indicating improved surgical efficiency and precision. However, the practical challenges of adopting this technology, particularly in high‐risk or emergent settings, must be acknowledged. First, regarding technical workflow, the approximately 15 min required for 3D modeling is acceptable for elective cases and can parallel preoperative preparation, though it may cause brief delays in extreme emergencies such as aortic rupture. With AI‐driven automated segmentation, future modeling time may be reduced to within minutes, overcoming this limitation. Cloud‐based platforms may further enhance accessibility and scalability by reducing dependence on local hardware. Second, regarding the learning curve, our experience shows that image‐familiar physicians acquire reconstruction skills after 2–3 training sessions, and surgeons learn intraoperative navigation rapidly. The intuitive nature of MR reduces cognitive burden in understanding complex anatomy, offering potential value for young surgeon training. A standardized multidisciplinary workflow is key to navigating this learning curve. Third, regarding cost‐effectiveness, fixed equipment costs are amortized over increasing case volume. More importantly, reduced fluoroscopy time and fewer stent adjustments decrease radiation exposure and may lower severe complication rates (e.g., type I endoleaks), generating long‐term health economic benefits. Therefore, the most cost‐effective strategy is not universal application, but targeted use in high‐risk patients with complex anatomy or those requiring advanced techniques (chimney/fenestrated procedures), where MR's navigational value is maximized.

There were also several limitations in this manuscript. Primaryly, this study has certain limitations in the assessment of endoleaks. As the reviewer correctly noted, a considerable proportion (approximately 15%–30%) of type II endoleaks may occur after 24 h following TEVAR, or even later. The detection of endoleaks in this study was primarily based on intraoperative completion DSA and pre‐discharge CTA (performed on postoperative Days 3–5). Although no type I or type III endoleaks were identified, the possibility that some delayed type II endoleaks were underestimated cannot be completely excluded. Furthermore, as the primary endpoint of this study was confined to the hospitalization period, with a lack of mid‐ to long‐term imaging follow‐up data at 1 month, 6 months, and 1 year postoperatively, accurate assessment of long‐term endoleak rates and long‐term stent stability is not feasible. Future studies should incorporate more rigorous imaging follow‐up protocols to comprehensively evaluate the potential value of MR technology in reducing long‐term complications.

In this study, the real‐time positioning and navigation functions of mixed reality (MR) may lead to the appropriate selection of stent size and the precise release of the stent during surgery, reducing secondary injury to patients caused by postoperative complications.

Author Contributions

Conception and design of the research: Jian‐Jun Gu, Teng‐Yue Zhao. Acquisition of data: Ying‐Chao Ma. Analysis and interpretation of the data: Jian‐Jun Gu, Xiao‐Chao Tian. Statistical analysis: Teng‐Yue Zhao, Xiao‐Chao Tian. Writing of the manuscript: Teng‐Yue Zhao, Ying‐Chao Ma. Critical revision of the manuscript for intellectual content: Jian‐Jun Gu. All authors read and approved the final draft.

Funding

The authors have nothing to report.

Ethics Statement

This study was conducted with approval from the Ethics Committee of The Second Hospital of Hebei Medical University. (Reference number: 20210978). This study was conducted in accordance with the declaration of Helsinki.

Consent

Written informed consent was obtained from all participants.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.