Enhancing liver surgery and transplantation: the role of 3D printing and virtual reality

3D Printing for preoperative planning and simulation

Modern cross-sectional imaging techniques such as computed tomography (CT) and magnetic resonance imaging (MRI) allow detailed visualization of patient anatomy. Several studies have demonstrated the feasibility of using such images to create accurate 3D printed models of the liver vasculature, lesions, and segmental anatomy^[3-7].

In an early report, Zein et al. described 3D printing liver models for three living liver donors and recipients undergoing liver transplantation. Individual arteries, veins, and biliary radicles were differentially colored, with anatomical accuracy confirmed after explantation^[1]. Printed models allowed enhanced the preoperative simulation and intraoperative navigation compared to standard CT/MRI. In the multicenter LIV3DPRINT study, Lopez-Lopez et al. validated the accuracy of 3D liver models with average errors of only 0.31 mm compared to the original CT/MRI scans^[3]. Although the study did not demonstrate improvements in surgical outcomes, it highlighted the significant utility of these models for educational purposes, enhancing the understanding of complex hepatic anatomy, and aiding in precise surgical planning. This underscores the importance of 3D printing technology in medical training and preoperative preparation, despite the need for further evidence on clinical outcomes.

Joo et al. evaluated personalized 3D models for detecting small liver lesions < 10 mm^[4]. Of 98 focal lesions (mean 11.5 mm) identified on MRI, printed models enabled the detection of 99.0% compared to 82.7% by conventional review, showing particular value for diminutive metastases and subcentimeter hepatocellular carcinoma. Multiple studies have concluded that 3D models improve spatial comprehension, facilitating planning for complex surgical anatomy for hepatectomy^[3,4].

Applications in surgical education

3D printed models allow trainees to visualize and manipulate anatomical structures with high fidelity, offering a hands-on approach to understanding complex surgical anatomy. Valls-Esteve et al. devised an economical 3D printing workflow to fabricate low-cost surgical simulators from patient scans, which trainees used to practice complex resections prior to operating^[5]. Surveyed residents reported enhanced procedural understanding and rated 3D models as excellent educational tools. These 3D models provide unparalleled visualization of subsurface tumors and vascular anatomy not otherwise visible, offering analytic training opportunities that are superior to animal labs or virtual simulators alone^[5].

The combined use of 3D printing and VR technologies has shown significant benefits in neurosurgical preoperative planning, as highlighted by González-López et al. (2024) in their study on neuro-oncology^[8]. They found that integrating these technologies transforms the paradigm of surgical planning, ensuring safer procedures and enhancing surgical preparedness. The detailed 3D models created from patient scans can be experienced through VR systems for comprehensive preoperative rehearsal, and 3D printed models provide hands-on training opportunities, thereby improving the overall safety and efficiency of neurosurgical procedures. Moreover, the advantages of integrating 3D printing and VR are not limited to neurosurgery. These technologies are equally applicable and beneficial in liver surgery and transplantation. By creating accurate 3D models of liver anatomy from patient scans, surgeons can plan and rehearse complex procedures in a virtual environment, leading to better understanding and anticipation of potential challenges. This approach can significantly enhance surgical outcomes, improve educational opportunities for trainees, and increase patient safety in hepatic surgeries as well. Moreover, these technologies address some of the limitations of traditional surgical training methods, such as animal labs or cadaveric dissection, by providing a more ethical and logistically feasible alternative^[5]. The adoption of 3D printing and VR in surgical education is poised to expand as these technologies become more accessible and affordable. As educational tools, they hold the potential to revolutionize the way surgical skills are taught and acquired, ultimately improving patient outcomes through better-prepared surgeons.

Patient education with 3D models

Structured preoperative education and informed consent discussions are integral in surgery, but conveying complex 3D relationships is difficult with standard imaging for patients. Giehl-Brown et al. investigated whether personalized 3D models improved patient education for hepatobiliary procedures^[9]. Among 40 subjects randomized to standard vs. 3D consent processes, the 3D cohort showed significantly higher satisfaction postoperatively (90% vs. 65%, P = 0.052). Critically, 94.4% of patients in the 3D model cohort were able to correctly report their complication status compared to just 68.4% of controls, indicating enhanced retention and comprehension.

Limitations of 3D printing

Despite holding immense potential to transform surgical planning in liver surgery, 3D printing has several limitations. For instance, producing accurate anatomical replicas requires considerable modeling expertise and advanced software/hardware, thus limiting widespread adoption^[4,5]. Moreover, in addition to costs, multi-day segmentation and printing introduce production delays, which may constrain their utility for urgent cases. Witowski et al. noted that static preoperative models cannot reflect intraoperative deformations from pneumoperitoneum or tissue manipulation, necessitating improved biomechanical modeling^[10]. Ongoing technological progress will increase the availability and functionality of 3D printing techniques.

VR training systems and procedural rehearsal

Immersive 3D environments provide invaluable foundations for elaborate multifaceted simulation systems. Lefor et al. developed a virtual reality simulator tailored for robotic liver resection, specifically modeling tissue physics and cautery dissection^[14]. Realistic ultrasonic shears approximate fine motor control, procedural kinetics, and pedagogical objectives for robotic training. Validation testing showed significant metric differentiation between novice and experienced surgeons over five domains, including time (P = 0.004), economy of motion (P = 0.043), and path length (P = 0.008). The system demonstrated utility in assessing skill level and progress through graduated complexity scenarios. VR circumvents the limitations of classical animal laboratories and cadaveric training, while generating quantitative performance data that are superior to observational clinical grading systems^[14].

Clinical outcomes

While abundant evidence has confirmed the technical accuracy of 3D modeling and immersive realism of VR simulation, few studies have linked these technologies to improvements in concrete perioperative outcomes. Given the multiple concurrent advances in hepatic surgery over the past decade, high-quality randomized controlled trials are essential for determining the impacts on safety, efficiency, complications, recovery, hospitalization duration, and long-term oncologic results^[9,10,12-13].

Present clinical data derive primarily from small prospective case series. However, in a 3D planning pilot study for complex oncologic resections, Lyuksemburg et al. reported increased concordance between preoperative models and ultimate procedures performed over standard CT/MRI review (92% vs. 54% for operating surgeons, 69% vs. 23% for consulting surgeons)^[15]. Furthermore, the 3D models enhanced preoperative understanding of respectability and vascular relationships in this initial experience. Experience is accruing, with improvements in planning, operative time, and reproduced margins, portending future randomized evaluations^[16].

AR navigation for liver segment identification

Beyond localizing lesions and avoiding vascular injury, precise liver segmentation is integral for oncologic resections^[22]. Recently, fluorescence imaging with indocyanine green (ICG) has become ubiquitous for identifying tumor-bearing portal segments and facilitating anatomic dissection. Selective puncture of tumor-bearing portal branches (positive staining method) or pedicle ligation with illuminated remnants (negative staining method) enhances intra-operative liver segmentation, but procedures of both positive and negative staining sometimes present technical difficulty in laparoscopy^[22,23]. Therefore, AR has shown early promise by compensating through projected portal mapping.

Kasai et al. devised a hybrid strategy consolidating 3D prints, AR projection, and real-time fluorescence for superimposed visible navigation^[24] [Figure 1]. In further evolution, the incorporation of artificial intelligence (AI) markedly improved auto-registration accuracy over algorithms without AI^[25] [Figure 2]. Furthermore, AI-based algorithms, such as those developed by Qin et al., significantly enhance the realism and accuracy of 3D models used in surgical simulations^[26]. AI-assisted AR navigation systems are expected to advance alongside maturing imaging and graphics modalities, potentially matching the stability of expensive robotic platforms.

Figure 1. Three-dimensional printing combined with augmented reality. Three-dimensional portal vein, hepatic vein and relationship of liver tumor with intrahepatic anatomy with segmental mapping. (A) Portal vein, hepatic vein, tumor in right liver lobe; (B-D) liver portal segment with augmented reality of segments 6, 7, dorsal segment 8. Image courtesy of Kasai et al.^[23].

Figure 2. Laparoscopic segmental mapping with augmented reality with artificial intelligence. This figure shows the pipeline of our AI-assisted AR navigation system, illustrating the process from training the AI on liver images to displaying 3D models in augmented reality. (A) Liver silhouette induced by AI; (B) three-dimensional models of liver; (C and D) superimposition of 3D models in laparoscopic surgical videos. Image courtesy of Kasai et al.^[24]. AI: Artificial intelligence; AR: augmented reality.