Fully automatic volume segmentation using deep learning approaches to assess aneurysmal sac evolution after infra-renal endovascular aortic repair

The benefit of automated sac volume measurements in postoperative endovascular aortic repair surveillance

OBJECTIVE

Abdominal aortic aneurysm (AAA) shrinkage is considered a marker for success following endovascular aortic repair (EVAR). Although maximum diameter is widely used to assess sac behavior, research indicates that changes in AAA morphology do not always affect the maximum diameter. The aim of this study was to investigate if automated AAA sac volume measurements after EVAR can add more nuanced information on sac behavior compared with maximum diameter evaluation alone.

METHODS

A retrospective review of all patients treated for AAA with a standard or fenestrated EVAR at two tertiary referral centers was performed. Patients with a preoperative and postoperative computed tomography angiography ≥2 years after treatment were included. Data were collected using medical charts, radiological institutional databases, and a deep learning based method called Augmented Reality for Vascular Aneurysm. Volume and diameter assessments were automatically performed on computed tomography angiography using Augmented Reality for Vascular Aneurysm. Preoperative sac volumes and diameters were compared with those obtained at least two years after repair. Information on endoleaks (ELs) was collected. Continuous data were tested using the t test, and categorical data were tested using the χ2 or Fishers test, depending on sample size.

RESULTS

A total of 89 patients (standard EVAR n = 46; fenestrated EVAR n = 43) were included in this study. Of the 89 patients, 41 (46%) had sac diameter shrinkage, 38 (43%) had stable sac diameters, and 10 (11%) had diameter sac growth during follow-up. The distribution of sac volume behavior was different among these patients: 51 (57%) had volume shrinkage, 9 (10%) had stable volumes, and 29 (33%) had volume growth. Significantly more patients had sac growth and fewer had sac stability, when assessed with volume compared with diameter (P = .003 and P < .001, respectively). The increase in patients with volume-assessed sac shrinkage (57% vs 46%) was not statistically significant. Of the 18 patients (20%) with stable sac diameters and simultaneous volume growth, 13 (72%) had ELs (type 2 ELs, n = 9; type 1 ELs, n = 2, type 3EL, n = 1, and EL of undefined origin, n = 1).

CONCLUSIONS

This study found that volume-assessed sac behavior identifies more sac shrinkage or growth, and less sac stability than diameter does. If confirmed by larger studies, sac volume assessment should be performed routinely after endovascular repair.

Volume Measurements for Surveillance after Endovascular Aneurysm Repair using Artificial Intelligence

OBJECTIVE

Surveillance after endovascular aneurysm repair (EVAR) is suboptimal due to limited compliance and relatively large variability in measurement methods of abdominal aortic aneurysm (AAA) sac size after treatment. Measuring volume offers a more sensitive early indicator of aneurysm sac growth or regression and stability, but is more time consuming and thus less practical than measuring maximum diameter. This study evaluated the accuracy and consistency of the artificial intelligence (AI) driven software PRAEVAorta 2 and compared it with an established semi-automated segmentation method.

METHODS

Post-EVAR aneurysm sac volumes measured by AI were compared with a semi-automated segmentation method (3mensio software) in patients with an infrarenal AAA, focusing on absolute aneurysm volume and volume evolution over time. The clinical impact of both methods was evaluated by categorising patients as showing either AAA sac regression, stabilisation, or growth comparing the 30 day and one year post-EVAR computed tomography angiography (CTA) images. Inter- and intra-method agreement were assessed using Bland-Altman analysis, the intraclass correlation coefficient (ICC), and Cohen's κ statistic.

RESULTS

Forty nine patients (98 CTA images) were analysed, after excluding 15 patients due to segmentation errors by AI owing to low quality CT scans. Aneurysm sac volume measurements showed excellent correlation (ICC = 0.94, 95% confidence interval [CI] 0.88 - 0.99) with good to excellent correlation for volume evolution over time (ICC = 0.85, 95% CI 0.75 - 0.91). Categorisation of AAA sac evolution showed fair correlation (Cohen's κ = 0.33), with 12 discrepancies (24%) between methods. The intra-method agreement for the AI software demonstrated perfect consistency (bias = -0.01 cc), indicating that it is more reliable compared with the semi-automated method.

CONCLUSION

Despite some differences in AAA sac volume measurements, the highly consistent AI driven software accurately measured AAA sac volume evolution. AAA sac evolution classification appears to be more reliable than existing methods and may therefore improve risk stratification post-EVAR, and could facilitate AI driven personalised surveillance programmes. While high quality CTA images are crucial, considering radiation exposure is important, validating the software with non-contrast CT scans might reduce the radiation burden.

External Validation of Fully-Automated Infrarenal Maximum Aortic Aneurysm Diameter Measurements in Computed Tomography Angiography Scans Using Artificial Intelligence (PRAEVAorta 2)

PURPOSE

This study investigates the accuracy of fully-automated maximum aortic diameter measurements in abdominal aortic aneurysm (AAA) patients using artificial intelligence software (PRAEVAorta 2, Nurea, Bordeaux, France).

MATERIALS AND METHODS

This is a multicenter, retrospective validation study using prospectively collected data from the Zenith alpha for aneurysm Repair Registry (ZEPHYR). Automated measurements of PRAEVAorta 2 are compared with measurements of an internationally recognized core laboratory (Syntactx, New York, New York State). The reviewers at the core laboratory were measurement technologists trained to and utilizing established measurement standards, overseen by vascular surgeons and radiologists. The data set comprised 871 computed tomography angiography scans from the ZEPHYR registry with 347 patients who underwent endovascular aneurysm repair (EVAR) with the Zenith Alpha Endovascular Abdominal Graft (Cook Medical, Bloomington, Indiana) in Germany, Belgium, and The Netherlands between 2016 and 2019.

RESULTS

The analysis demonstrated excellent correlation of the measurements (r=0.97) with an intraclass correlation (ICC) of 0.972 (95% confidence interval [CI]=0.968-0.976) across all scans. For preoperative computed tomography (CT) scans, ICC was 0.953 (95% CI=0.941-0.963), and for postoperative scans, ICC was 0.979 (95% CI=0.975-0,983), respectively. In total, 95.4% of measurements were within the clinically acceptable range of 5 mm in absolute difference. In total, 10% of scans demonstrated obvious segmentation errors, mainly due to failure in detecting vessel segments (renal arteries, aortic bifurcation) or due to mis-detecting the outer border of the AAA (duodenum, inferior vena cava, aortic branches) and were excluded from the analysis.

CONCLUSION

In this study, the maximum AAA diameter could be accurately measured fully-automatically by PRAEVAorta 2 (Nurea) in most cases demonstrating that artificial intelligence (AI) software could serve as an important adjunct for research and clinical practice. However, critical review of the generated reports by an experienced observer and cautious use is warranted to identify flawed segmentations.

CLINICAL IMPACT

This multicenter, retrospective validation study assessed the accuracy of fully-automated maximum infrarenal aortic aneurysm diameter measurements. It was demonstrated, that 95.4% of measurements were within the clinically acceptable range of 5 mm in absolute difference, positioning the software as a potential adjunct for clinical practice and research. It is also highlighted however, that critical review of the measurements is obligatory, due to a 10% segmentation error rate.

The sac evolution imaging follow-up after endovascular aortic repair: An international expert opinion-based Delphi consensus study

OBJECTIVE

Management of follow-up protocols after endovascular aortic repair (EVAR) varies significantly between centers and is not standardized according to sac regression. By designing an international expert-based Delphi consensus, the study aimed to create recommendations on follow-up after EVAR according to sac evolution.

METHODS

Eight facilitators created appropriate statements regarding the study topic that were voted, using a 4-point Likert scale, by a selected panel of international experts using a three-round modified Delphi consensus process. Based on the experts' responses, only those statements reaching a grade A (full agreement ≥75%) or B (overall agreement ≥80% and full disagreement <5%) were included in the final document.

RESULTS

One-hundred and seventy-four participants were included in the final analysis, and each voted the initial 29 statements related to the definition of sac regression (Q1-Q9), EVAR follow-up (Q10-Q14), and the assessment and role of sac regression during follow-up (Q15-Q29). At the end of the process, 2 statements (6.9%) were rejected, 9 statements (31%) received a grade B consensus strength, and 18 (62.1%) reached a grade A consensus strength. Of 27 final statements, 15 (55.6%) were classified as grade I, whereas 12 (44.4%) were classified as grade II. Experts agreed that sac regression should be considered an important indicator of EVAR success and always be assessed during follow-up after EVAR.

CONCLUSIONS

Based on the elevated strength and high consistency of this international expert-based Delphi consensus, most of the statements might guide the current clinical management of follow-up after EVAR according to the sac regression. Future studies are needed to clarify debated issues.

Assessing the diagnostic accuracy of artificial intelligence in post-endovascular aneurysm repair endoleak detection using dual-energy computed tomography angiography

Purpose

The aim of this study was to evaluate the diagnostic accuracy of an artificial intelligence (AI) tool in detecting endoleaks in patients undergoing endovascular aneurysm repair (EVAR) using dual-energy computed tomography angiography (CTA).

Material and methods

The study involved 95 patients who underwent EVAR and subsequent CTA follow-up. Dualenergy scans were performed, and images were reconstructed as linearly blended (LB) and 40 keV virtual monoenergetic (VMI) images. The AI tool PRAEVAorta®2 was used to assess arterial phase images for endoleaks. Two experienced readers independently evaluated the same images, and their consensus served as the reference standard. Key metrics, including accuracy, precision, recall, F1 score, and area under the receiver operating characteristic (ROC) curve (AUC), were calculated.

Results

The final analysis included 94 patients. The AI tool demonstrated an accuracy of 78.7%, precision of 67.6%, recall of 10 71.9%, F1 score of 69.7%, and an AUC of 0.77 using LB images. However, the tool failed to process 40 keV VMI images correctly, limiting further analysis of these datasets.

Conclusions

The AI tool showed moderate diagnostic accuracy in detecting endoleaks using LB images but failed to achieve the reliability needed for clinical use due to the significant number of misdiagnoses.

Volumetric analysis of acute uncomplicated type B aortic dissection using an automated deep learning aortic zone segmentation model

BACKGROUND

Machine learning techniques have shown excellent performance in three-dimensional medical image analysis, but have not been applied to acute uncomplicated type B aortic dissection (auTBAD) using Society for Vascular Surgery (SVS) and Society of Thoracic Surgeons (STS)-defined aortic zones. The purpose of this study was to establish a trained, automatic machine learning aortic zone segmentation model to facilitate performance of an aortic zone volumetric comparison between patients with auTBAD based on the rate of aortic growth.

METHODS

Patients with auTBAD and serial imaging were identified. For each patient, imaging characteristics from two computed tomography (CT) scans were analyzed: (1) the baseline CT angiography (CTA) at the index admission and (2) either the most recent surveillance CTA or the most recent CTA before an aortic intervention. Patients were stratified into two comparative groups based on aortic growth: rapid growth (diameter increase of ≥5 mm/year) and no or slow growth (diameter increase of <5 mm/year). Deidentified images were imported into an open source software package for medical image analysis and images were annotated based on SVS/STS criteria for aortic zones. Our model was trained using four-fold cross-validation. The segmentation output was used to calculate aortic zone volumes from each imaging study.

RESULTS

Of 59 patients identified for inclusion, rapid growth was observed in 33 patients (56%) and no or slow growth was observed in 26 patients (44%). There were no differences in baseline demographics, comorbidities, admission mean arterial pressure, number of discharge antihypertensives, or high-risk imaging characteristics between groups (P > .05 for all). Median duration between baseline and interval CT was 1.07 years (interquartile range [IQR], 0.38-2.57). Postdischarge aortic intervention was performed in 13 patients (22%) at a mean of 1.5 ± 1.2 years, with no difference between the groups (P > .05). Among all patients, the largest relative percent increases in zone volumes over time were found in zone 4 (13.9%; IQR, -6.82 to 35.1) and zone 5 (13.4%; IQR, -7.78 to 37.9). There were no differences in baseline zone volumes between groups (P > .05 for all). The average Dice coefficient, a performance measure of the model output, was 0.73. Performance was best in zone 5 (0.84) and zone 9 (0.91).

CONCLUSIONS

We describe an automatic deep learning segmentation model incorporating SVS-defined aortic zones. The open source, trained model demonstrates concordance to the manually segmented aortas with the strongest performance in zones 5 and 9, providing a framework for further clinical applications. In our limited sample, there were no differences in baseline aortic zone volumes between patients with rapid growth and patients with no or slow growth.

Multicentric clinical evaluation of a computed tomography-based fully automated deep neural network for aortic maximum diameter and volumetric measurements

OBJECTIVE

This study aims to evaluate a fully automatic deep learning-based method (augmented radiology for vascular aneurysm [ARVA]) for aortic segmentation and simultaneous diameter and volume measurements.

METHODS

A clinical validation dataset was constructed from preoperative and postoperative aortic computed tomography angiography (CTA) scans for assessing these functions. The dataset totaled 350 computed tomography angiography scans from 216 patients treated at two different hospitals. ARVA's ability to segment the aorta into seven morphologically based aortic segments and measure maximum outer-to-outer wall transverse diameters and compute volumes for each was compared with the measurements of six experts (ground truth) and thirteen clinicians.

RESULTS

Ground truth (experts') measurements of diameters and volumes were manually performed for all aortic segments. The median absolute diameter difference between ground truth and ARVA was 1.6 mm (95% confidence interval [CI], 1.5-1.7; and 1.6 mm [95% CI, 1.6-1.7]) between ground truth and clinicians. ARVA produced measurements within the clinical acceptable range with a proportion of 85.5% (95% CI, 83.5-86.3) compared with the clinicians' 86.0% (95% CI, 83.9-86.0). The median volume similarity error ranged from 0.93 to 0.95 in the main trunk and achieved 0.88 in the iliac arteries.

CONCLUSIONS

This study demonstrates the reliability of a fully automated artificial intelligence-driven solution capable of quick aortic segmentation and analysis of both diameter and volume for each segment.

One-year follow-up after active aortic aneurysm sac treatment with shape memory polymer devices during endovascular aneurysm repair

OBJECTIVE

To determine the safety and efficacy of treating abdominal aortic aneurysm (AAA) sacs with polyurethane shape memory polymer (SMP) devices during endovascular aneurysm repair (EVAR), using a technique to fully treat the target lumen after endograft placement (aortic flow volume minus the endograft volume). SMP devices self-expand in the sac to form a porous scaffold that supports thrombosis throughout its structure.

METHODS

Two identical prospective, multicenter, single-arm studies were conducted in New Zealand and the Netherlands. The study population was adult candidates for elective EVAR of an infrarenal AAA (diameter of ≥55 mm in men and ≥50 mm in women). Key exclusion criteria were an inability to adequately seal a common iliac artery aneurysm, patent sac feeding vessels of >4 mm, and a target lumen volume of <20 mL or >135 mL. Target lumen volumes were estimated by subtracting endograft volumes from preprocedural imaging-based flow lumen volumes. SMP devices were delivered immediately after endograft deployment via a 6F sheath jailed in a bowed position in the sac. The primary efficacy end point was technical success, defined as filling the actual target lumen volume with fully expanded SMP at the completion of the procedure. Secondary efficacy outcome measures during follow-up were the change in sac volume and diameter, rate of type II endoleak and type I or III endoleaks, and the rate of open repair and related reinterventions, with data collection at 30 days, 6 months, and 1 year (to date). Baseline sac volumes and diameters for change in sac size analyses were determined from 30-day imaging studies. Baseline and follow-up volumes were normalized by subtraction of the endograft volume.

RESULTS

Of 34 patients treated with SMP devices and followed per protocol, 33 patients were evaluable at 1 year. Preprocedural aneurysm volume was 181.4 mL (95% confidence interval [CI], 150.7-212.1 mL) and preprocedural aneurysm diameter was 60.8 mm (95% CI, 57.8-63.9 mm). The target lumen volume was 56.3 mL (95% CI, 46.9-65.8 mL). Technical success was 100% and the ratio of SMP fully expanded volume to estimated target lumen volume was 1.4 ± 0.3. Baseline normalized sac volume and diameter were 140.7 mL (95% CI, 126.6-154.9 mL) and 61.0 mm (95% CI, 59.7-62.3 mm). The adjusted mean percentage change in normalized volume at 1 year was -28.8% (95% CI, -35.3 to -22.3%; P < .001). The adjusted mean change in sac diameter at 1 year was -5.9 mm (95% CI, -7.5 to -4.4 mm; P < .001). At 1 year, 81.8% of patients (95% CI, 64.5%-93.0%) achieved a ≥10% decrease in normalized volume and 57.6% of patients (95% CI, 39.2%-74.5%) achieved a ≥5 mm decrease in diameter. No device- or study procedure-related major adverse events occurred through 1 year after the procedure.

CONCLUSIONS

Treatment of AAA sacs with SMP devices during EVAR resulted in significant sac volume and diameter regression at 1 year with an acceptable safety profile in this prospective study.

Importance of sac regression after EVAR and the role of EndoAnchors

The initial success and widespread adoption of endovascular aneurysm repair (EVAR) for the treatment of abdominal aortic aneurysms have been tempered by numerous reports of secondary interventions and increased long-term mortality compared with open repair. Over the past decade, several studies on postoperative sac dynamics after EVAR have suggested that the presence of sac regression is a benign feature with a favorable prognosis. Conversely, increasing sacs and even stable sacs can be indicators of more unstable sac behavior with worse outcomes in the long-term. Endoleaks were initially perceived as the main drivers of sac behavior. However, the observation that sac regression can occur in the presence of endoleaks, and vice versa - increasing sacs without evidence of endoleak - on imaging studies, suggests the involvement of other contributing factors. These factors can be divided into anatomical factors, patient characteristics, sac thrombus composition, and device-related factors. The shift of interest away from especially type 2 endoleaks is further supported by promising results with the use of EndoAnchors regarding postoperative sac behavior. This review provides an overview of the existing literature on the implications and known risk factors of post-EVAR sac behavior, describes the accurate measurement of sac behavior, and discusses the use of EndoAnchors to promote sac regression.

Editor's Choice -- European Society for Vascular Surgery (ESVS) 2024 Clinical Practice Guidelines on the Management of Abdominal Aorto-Iliac Artery Aneurysms

OBJECTIVE

The European Society for Vascular Surgery (ESVS) has developed clinical practice guidelines for the care of patients with aneurysms of the abdominal aorta and iliac arteries in succession to the 2011 and 2019 versions, with the aim of assisting physicians and patients in selecting the best management strategy.

METHODS

The guideline is based on scientific evidence completed with expert opinion on the matter. By summarising and evaluating the best available evidence, recommendations for the evaluation and treatment of patients have been formulated. The recommendations are graded according to a modified European Society of Cardiology grading system, where the strength (class) of each recommendation is graded from I to III and the letters A to C mark the level of evidence.

RESULTS

A total of 160 recommendations have been issued on the following topics: Service standards, including surgical volume and training; Epidemiology, diagnosis, and screening; Management of patients with small abdominal aortic aneurysm (AAA), including surveillance, cardiovascular risk reduction, and indication for repair; Elective AAA repair, including operative risk assessment, open and endovascular repair, and early complications; Ruptured and symptomatic AAA, including peri-operative management, such as permissive hypotension and use of aortic occlusion balloon, open and endovascular repair, and early complications, such as abdominal compartment syndrome and colonic ischaemia; Long term outcome and follow up after AAA repair, including graft infection, endoleaks and follow up routines; Management of complex AAA, including open and endovascular repair; Management of iliac artery aneurysm, including indication for repair and open and endovascular repair; and Miscellaneous aortic problems, including mycotic, inflammatory, and saccular aortic aneurysm. In addition, Shared decision making is being addressed, with supporting information for patients, and Unresolved issues are discussed.

CONCLUSION

The ESVS Clinical Practice Guidelines provide the most comprehensive, up to date, and unbiased advice to clinicians and patients on the management of abdominal aorto-iliac artery aneurysms.

The Role of Aortic Volume in the Natural History of Abdominal Aortic Aneurysms and Post-Endovascular Aortic Aneurysm Repair Surveillance

There has been a debate about whether maximum diameter can be solely used to assess the natural history of abdominal aortic aneurysm. The aim of the present review is to collect all the available evidence on the role of abdominal aortic aneurysm (AAA) volume in the natural history of AAAs, including small untreated AAAs and AAAs treated by EVAR. The current literature appears to reinforce the role of volume as a supplementary measure for evaluating the natural history of AAA, in both intact AAAs and after EVAR. The clinical impact of AAA volume measurements remains unclear. Several studies show that volumetric analysis can assess changes in AAAs and predict successful endoluminal exclusion after EVAR more accurately than diameter. However, most studies lack strict standardized measurement criteria and well-defined outcome definitions. It remains unclear whether volumetry could replace diameter assessment in defining the risk of rupture of AAAs and identifying clinically relevant sac growth.

A Comparison of Suprarenal and Infrarenal Fixation and Renal Volumetric Analysis after Endovascular Aneurysm Repair

OBJECTIVE

Clinical and experimental studies of the stent-graft fixation impact on renal volume after endovascular abdominal aortic aneurysm repair have focused on glomerular filtration rate (GFR) and the results were controversial. The aim of this study was to analyze and compare the impact of the supra (SRF group) and infra (IRF group) renal stent-graft fixation on the renal volume.

METHODS

Between December 2016 and December 2019, all patients treated with EVAR were retrospectively analyzed. Patients with atrophic or multicystic kidney, renal transplantation, ultrasound or incomplete follow-up were excluded. Renal volume in both groups was extracted with a semi-automatic segmentation from contrast-enhanced CT-scan performed before the procedure, at 1 month and at 12 months follow-up. A subgroup analyze of the SRF group was performed in order to study impact of the stent struts position relative to the renal arteries.

RESULTS

63 patients were analyzed (SRF: 32, IRF: 31). Demographic and anatomical characteristics were similar between the groups. Procedure contrast volume was higher in the IRF group (P= 0.01). At 12-months, the renal volume decreased of 1.4 % in the SRF group and 2.3 % in the IRF group (P=0.86). The SRF sub group analysis showed only 2 patients with no stent struts crossing renal arteries. In the remaining cases, struts crossed one renal artery in 60% of cases (19 patients) and 2 renal arteries in 34% of cases (11 patients). The renal volume decrease was not correlated to the presence of stent wires struts, crossing renal artery.

CONCLUSIONS

Stent-graft with supra renal fixation seems not to be correlated with renal volume deterioration. A randomized clinical trial with a higher effective and longer follow-up is needed to assess the impact of SRF on the renal function.

Abdominal Aortic Aneurysm Diameter versus Volume: A Systematic Review

Recently, AAA volume measurement has been proposed as a potentially valuable surveillance method in situations when diameter measurement might fail.

OBJECTIVE

The aim of this systematic review was to analyze the results of previous studies comparing AAA diameter and volume measurements.

METHODS

A systematic search in PubMed, Cochrane, and EMBASE databases was performed to identify studies investigating the use of diameter and volume measurements in AAA diagnosis and prognosis in English, German, and Russian, published until December 2022. The manuscripts were reviewed by three researchers and scored on the quality of the research using MINORS criteria.

RESULTS

After screening 752 manuscripts, 19 studies (n = 1690) were included. The majority (n = 17) of the manuscripts appeared to favor volume. It is, however, important to highlight the heterogeneity of methodologies and lack of standardized protocol for measuring both volume and diameter in the included studies, which hindered the interpretation of the results.

CONCLUSIONS

The clinical relevance of abdominal aortic aneurysm volume measurement is still unclear, although studies show favorable and promising results for volumetric changes in AAA, especially in follow-up after EVAR.