Augmented Reality Informs Locoregional Therapy

Three-Dimensional Augmented Reality Visualization Informs Locoregional Therapy in a Translational Model of Hepatocellular Carcinoma

Clinical question
Is augmented reality (AR) visualization of pre-procedural magnetic resonance imaging (MRI) in 3 dimensions (3D) beneficial prior to transarterial embolization of hepatocellular carcinoma (HCC) in a preclinical rat model?

Take away point
3D AR visualization of pre-procedural imaging can reduce procedural fluoroscopy time and potentially total catheterization time. Further animal studies as well as prospective trials in human participants are warranted to evaluate the potential of AR implementation in endovascular interventions.

Reference
Three-Dimensional Augmented Reality Visualization Informs Locoregional Therapy in a Translational Model of Hepatocellular Carcinoma. Park, B.J., Perkons, N.R., Profka, E., Johnson, O., Morley, C., Appel, S., Nadolski, G.J., Hunt, S.J., Gade, T.P. Journal of Vascular and Interventional Radiology (JVIR), Volume 31, Issue 10, 1612-1618.

Click here for abstract

Study design
28 rats with diethylnitrosamine (DEN) induced HCCs larger than 5 mm were enrolled prospectively. 12 rats underwent transarterial embolization after 3D AR visualization while the other 16 rats underwent transarterial embolization with two-dimensional viewing only of the pre-procedural MRI. An additional retrospective cohort of 15 cases were later identified and combined with the non-3D-AR-visualization group to increase statistical power. Procedural metrics including fluoroscopy time, catheterization time, and radiation exposure were measured and compared.

Funding source
B.P. received grants from the National Institutes of Health (5T32EB004311), Society of Interventional Radiology Foundation, and Radiological Society of North America Research and Education Foundation. The authors thank the Penn Medicine Medical Device Accelerator for supporting equipment for this research.

Setting
Academic hospital. University of Pennsylvania, Philadelphia, PA.





Figure. A. Operator wearing AR headset and viewing preprocedural imaging immediately prior to the procedure. B. Captured image clip from Video 2 showing holographic 3D MRI volume in the interventional suite.

Summary

AR has seen tremendous development and promising prospects in recent years. AR headsets in particular have allowed visualization of holographic models and cross-sectional scans in true 3D space rather than flattened to 2D screens. The added depth information as well as 3D manipulation of the cross-sectional imaging may improve learning, surgical planning, and procedural efficiency. Nevertheless, these potentially beneficial impacts remain largely unstudied.

The authors designed a preclinical prospective trial using a rat model to determine if preprocedural AR visualization of MRI may improve procedural metrics. With institutional animal care approval, the authors utilized a previously established dithylnitrosamine (DEN) induced HCC rat model. 28 rats with HCCs larger than 5 mm on pre-procedural MRI scan were prospectively enrolled, 12 in the AR visualization prior to transarterial embolization arm and 16 in the control arm. Preprocedural images were always reviewed prior to intervention, in 2D plane for the control arm and in 2D+3D AR fashion for the AR visualization arm. Embolization was performed through a cut-down femoral access with 0.05 mL of 40-120 um Embosphere or 40-90 um LC Bead LUMI particles diluted in 0.9 mL of Omnipaque 300 until stasis in the selected segmental haptic artery. Total femoral artery catheterization time, total fluoroscopy time, dose-area product (DAP), and air kerma (AK) were recorded and compared. Given post-hoc power analysis showing a sample size of 52 was needed for 80% power with an effect size of 0.8 and alpha of 0.05, 15 additional cases were identified retrospectively and added to the control arm.

Analyses using only prospective data demonstrated total catheterization time reduction from 42.7 to 31.0 minutes, p=0.11, in the AR visualization arm compared to the control arm. Total fluoroscopy time decreased from 11.7 to 7.4 minutes, p=0.12, compared to the control arm. DAP and AK did not change significantly between the two arms. AR setup and viewing added 5.9 minutes of pre-procedural preparation time per case on average. 15 additional retrospective cases were identified and determined to be statistically similar to the prospective control arm. Using both combined retrospective and prospective controls, the decrease in total fluoroscopy time was statistically significant with p=0.01. Furthermore, there were statistically significant reductions in the variability of catheterization and fluoroscopy times in the AR visualization arm.

Commentary

The authors in this paper have evaluated the potential impact of 3D AR visualization in pre-procedural planning for locoregional liver therapies in a preclinical rat animal model. Prospective comparisons did not reach statistical significance. Factors including the use of human-calibrated angiographic systems on rats, different embolic agents, and small sample size may have compounded the comparisons. Addition of appropriate retrospective cases demonstrated a significant decrease in total fluoroscopy time. This reduction in total fluoroscopic times, variations in particular, suggested benefits of 3D AR visualization in diminishing excessive or unplanned fluoroscopy use. These results were compatible with prior studies showing decreased mental task load in AR simulated surgery as well as decreased needle passes and DAP on AR abdominal phantom simulations of CT-guided interventions. Preliminary data, such as the results reported in this paper, together with foreseeable benefits of 3D AR visualization for pre-procedural planning and intra-procedural guidance should prompt future animal studies, potentially with large animal models, and eventual human prospective trials. We may be on the verge of another technological breakthrough in procedural visualization after the adaptation of fluoroscopy and laparoscopy.

Post Author
Ningcheng (Peter) Li, MD, MS
Integrated Interventional Radiology Resident, PGY-4
Department of Interventional Radiology
Oregon Health and Science University, Dotter Interventional Institute 

@NingchengLi