Synergies Between Robotics, AI, and Bioengineering—A Narrative Review Concerning the Future of Transplants

Abstract

The critical shortage of donor organs remains the foremost challenge in transplantation medicine. Nevertheless, advancements in robotic-assisted surgery (RAS), artificial intelligence (AI)-enhanced donor–recipient matching, and bioengineering—particularly 3D bioprinting—are revolutionizing the field. Today, RAS has evolved from an innovative technique into a reliable clinical tool, with evidence indicating that it enhances surgical precision and results in better patient outcomes. Meanwhile, AI and machine learning are advancing donor–recipient matching and allocation, producing models that offer superior predictive accuracy for graft survival compared to traditional methods. Additionally, bioengineering strategies, especially 3D bioprinting and tissue engineering, are progressing from the creation of acellular scaffolds to the development of vascularized constructs, marking a significant milestone toward functional organ replacement. Despite persistent challenges such as high costs, regulatory obstacles, new structured formation programs, and the necessity for effective vascularization in engineered tissues, the integration of these disciplines is forging a new paradigm in regenerative medicine. The primary objective of this review is to synthesize multidisciplinary innovations by leveraging clinical studies and technological assessments to delineate future directions in regenerative medicine and organ transplantation.

1. Introduction

The success of organ transplantation is widely recognized as one of the most significant achievements of modern medicine, as the sole definitive therapeutic approach in cases of vital organ dysfunction [1]. Despite the significant advances achieved and the paramount importance of this surgical procedure, the persistent disparity between organ demand and supply remains a major challenge to global public health [2]. The history of transplantation has ancient origins, with the first recorded evidence of skin grafts dating back to medical texts from Ancient Egypt [3]. These texts mark the beginning of a historical journey that laid the conceptual foundation for tissue transfer.

The first transplants performed were unsuccessful due to a lack of knowledge of fundamental anatomical and physiological concepts, such as vascular anastomosis, proper organ preservation, and immune rejection [4,5,6]. A significant advancement in the technical feasibility of this procedure was marked by the first documented case of skin autograft [7,8]. In the early 20th century, the scientific community began to explore alternatives to autografts, due to the severe limitations of this method, being applicable to only a narrow category of tissues. The development of allografts (transplants between two genetically different individuals) represented a significant step forward in this field. In this context, doctors Carrel and Guthrie began to postulate the concept of “biological incompatibility” as the primary cause of failure [9]. This concept was further explored by Yuri Voronoy’s studies in 1933, which demonstrated the importance of ABO blood group incompatibility [10]. The first documented instance of clinical success in this field was not achieved until 1954, when Joseph Murray performed the first kidney transplant between identical twins, whose perfect genetic compatibility ensured survival and demonstrated the feasibility of organ transplantation [11].

Early attempts to overcome the immune barrier included toxic and ineffective techniques, such as sub-lethal X-ray exposure [12], which were quickly abandoned in favor of immunosuppressive drugs such as 6-mercaptopurine and azathioprine [8,13]. In particular, the combination of azathioprine with corticosteroids (e.g., prednisone) was rapidly established as the standard therapeutic approach, making transplants a viable treatment option for a significant number of patients. The advent of more targeted agents, such as cyclosporine and tacrolimus, significantly improved survival rates [12,14]. Despite the fact that transplantation is now a well-established medical practice, there are significant variations in the number of transplants performed in different countries around the world [15]. The United States continues to dominate the field, with the highest number of transplants performed worldwide. In the United States, the number of transplants performed in 2024 exceeded 48,000, marking an approximate 3.3% increase compared to the previous year.

Renal transplants remained the most prevalent, followed by those of the lung (+10.4%) and the liver (+7.5%) [16]. Spain has the highest number of transplants in Europe, with a record 6464 transplants in 2024 and a donation rate of 52.6 donors per million inhabitants [17]. The African region continues to exhibit the lowest number of transplants, with only 286 recorded in 2022, with the exception of South Africa [15].

2. The Revolution of Robotic Surgery in Organ Transplantation: Advantages, Limitations and Future Directions

Transplant surgery represents the culmination of a complex medical–surgical process, in which long-term success depends on highly specialized immunological and clinical management. In this context, transplant surgery is a fundamental pillar of precision medicine, integrating molecular, histological and clinical data to guide personalized therapeutic decisions [18]. The success of transplant surgery procedures also depends on the advent of new technologies that enable more accurate results and optimize surgical timing. Among the most recent technological innovations in this field, the advent of next-generation sequencing (NGS) technologies is revolutionizing donor–recipient compatibility assessment procedures, enabling matching based not only on HLA antigens but also on eplets, with direct implications for immunological risk stratification and the prevention of donor-specific antibody formation [19,20].

Similarly, the integration of next-generation pathology (NGP) approaches, which combine multiplex staining, digital imaging and automated analysis based on machine learning, overcomes the limitations of traditional histopathological assessment, which is often subjective and non-quantitative [21].

The revolution in organ transplantation also involves the adoption of minimally invasive techniques, with a gradual, albeit still partial in some areas, shift from laparoscopic to robotic methods.

Robotic-assisted surgery (RAS) has gradually established itself as a revolutionary tool in the field of living donor transplants, offering superior precision and results compared to traditional approaches, particularly in procedures such as nephrectomy for kidney donation, pancreas donation and, more recently, hepatectomy for liver donation. The main advantages and limitations of RAS are shown in Figure 1.

2.1. Kidney

Kidney transplantation is the definitive treatment for patients with end-stage renal disease, as it offers higher survival rates than conservative therapy and provides a better quality of life than dialysis [22]. For decades, the open approach has remained the standard surgical technique. However, in recent years, robotic-assisted kidney transplantation (RAKT) has emerged as a promising alternative, thanks to the typical advantages of minimally invasive surgery. RAKT has demonstrated safety and feasibility in several reference centers, with increasing use, supported by dedicated registries and structured programs [23]. Recent meta-analyses of comparative studies, such as the work by Slagter et al. [24], based on 11 studies comparing 482 RAKTs with 1316 OKTs, now provide a robust evidence base for its feasibility and specific intraoperative parameters. RAKT was indeed associated with lower a risk of surgical site infection (Risk ratio (RR) = 0.15, p < 0.001), less postoperative pain (mean difference (MD) = −1.38 points, p < 0.001), smaller incision length (MD = −8.51 cm, p < 0.001), and shorter length of hospital stay (MD = −1.69 days, p = 0.03) compared with OKT. In addition, new robotic surgical systems such as the Senhance (NC, USA) and REVO-I (Korea) surgical systems have received regulatory approval in some countries, and others will be introduced to the market shortly. The clinical importance of these technical advancements is a demonstrable improvement in patient-centered outcomes. The data from these studies translate into tangible benefits such as a lower risk of surgical site infections, reduced postoperative pain, and significantly shorter hospital stays, enhancing recovery for the living donor.

Despite these encouraging results, scientific evidence remains limited—with only 18 studies available and a single randomized study that is not directly comparative—and RAKT still presents some technical challenges. These include potential prolongation of warm ischemia time, a steep learning curve requiring structured training programs and thorough supervision to ensure patient safety and optimal outcomes, and, finally, better management of complex cases (e.g., obesity, severe atherosclerosis, deceased donors, or pediatric recipients [25].

These challenges necessitate the implementation of RAKT reproducibility outside of reference centers, improving technical and logistical aspects and conducting large-scale multicenter clinical trials to make the procedure available to a wider range of patients [26].

2.2. Liver

Living donor hepatectomy, particularly of the right lobe, is an established approach in adult-to-adult liver transplantation, helping to overcome the shortage of organs from deceased donors [27]. However, complication rates of between 16% and 34% have been reported, often related to abdominal wall complications in the open approach [28].

The introduction of minimally invasive surgical techniques, including RAS, has opened new prospects for improving outcomes for both donors and recipients. The first case of robotic hepatectomy from a living donor was described by Giulianotti et al. in 2012 [29] using the Da Vinci Robotic Surgical System (CA, USA), in a minimally invasive fashion, during which the liver graft was safely extracted through a limited lower abdominal incision. As a minimally invasive modality, robotic-assisted surgery offers several potential advantages over conventional laparoscopic techniques. In the context of donor hepatectomy, the robotic platform provides a stable, high-magnification, three-dimensional operative field and superior instrument articulation, which collectively enhance the precision of vascular and biliary dissection around the right hepatic pedicle. These features facilitate a more accurate determination of the transection plane. Furthermore, the availability of a fourth robotic arm guarantees stable retraction of the right hepatic lobe, improving exposure and enabling precise identification and suturing of accessory hepatic veins during the caval phase. Compared with standard laparoscopy, the robotic system’s improved suture control may contribute to reduced intraoperative blood loss and potentially shorter operative times, thereby decreasing complication rates and overall perioperative morbidity [29]. Since then, evidence has accumulated that robotic liver resection is comparable in terms of safety and oncological outcomes to open and laparoscopic approaches, with advantages including reduced postoperative pain, lower morbidity, and earlier discharge, albeit with longer operating times and higher costs [30,31].

In a recent case reported by Lee et al., a 57-year-old recipient patient with alcoholic liver cirrhosis and hepatocellular carcinoma underwent successful robot-assisted liver transplantation, with total robotic hepatectomy completed in 327 min and robotic grafting in 295 min, including biliary anastomosis, with warm and cold ischemia of 55 and 229 min, respectively. The donor, who underwent pure laparoscopic hepatectomy, benefited from a standardized minimally invasive approach, with precise dissection of the hepatic hilum, cholangiography and intraoperative ultrasound [32]. Although robotic systems involve longer operating times and higher costs, RAS offers several advantages: better ergonomics, 3D vision, and articulated instrument movement. These features enable more precise dissection and may reduce complications [33]. However, due to its technical complexity and the lack of clear benefits in bleeding or liver function compared to laparoscopy, robotic hepatectomy from living donors has not yet shown economic or clinical superiority. Still, it represents a step forward toward standardizing minimally invasive liver surgery [34].

2.3. Pancreas

Pancreas transplantation is the most effective treatment for restoring physiological glycemic control in patients with insulin-dependent type 1 diabetes, especially in those with poor quality of life despite adequate insulin therapy [35]. It remains essential for preventing the progression of micro- and macrovascular complications [36]. Although advances in surgical technique and immunosuppression have improved long-term outcomes, the rate of post-operative complications still stands at around 25%, highlighting the need for less invasive surgical approaches [37]. Living pancreas donation is extremely rare and mainly limited to distal pancreatectomy with spleen preservation, aimed at reducing donor morbidity [38]. The first case of robot-assisted nephrectomy and distal pancreatectomy was described by Horgan et al. in 2007, demonstrating the feasibility of the technique, albeit with prolonged operating times and longer warm ischemia time (WIT) [39]. Robot-assisted pancreas transplantation has also been studied in complex cases such as severely obese patients, with promising results in terms of metabolic control and absence of diabetic recurrence in follow-up, as reported by Yeh et al. In this case, the robotic pancreas transplantation employed two 8 mm robotic arm ports, a 12 mm camera port, a 12 mm assistant port, and a 7 cm GelPort (CA, USA) hand-assist incision. Vascular anastomoses were performed robotically using Bulldog clamps and 5-0 e-PTFE sutures. A circular EEA stapler (Covidien) enabled duodeno-bladder anastomosis, and an Endo-GIA stapler (Ethicon) was used to close the duodenal end, ensuring precise reconstruction within a minimally invasive setup [40]. However, the adoption of RAS in pancreas transplantation is still limited by several factors, including high costs, the duration of the procedure, and the need to maintain certain key stages of the procedure (e.g., vascular complications) in open mode [41]. Currently, RAS can be considered feasible in highly specialized centers, but it is not yet a standard of care. It may complement the open approach during dissection and organ positioning, while the main anastomoses remain open to ensure safety and proper timing [42].

Despite promising results, with immediate graft function and no post-operative complications, robot-assisted pancreas transplantation remains an experimental procedure, limited by high costs, lack of tactile feedback, and the need for conversion in the event of massive hemorrhage. However, further studies are needed to define its real advantages over the open approach and to address the challenges associated with warm ischemia and intraoperative management of grafts [43].

2.4. Advantages, Limitations and Future Directions

Currently, RAS is considered an experimental procedure, mainly reserved for living donor transplants; however, the progressive refinement of robotic platforms and multicenter collaborations is expanding its potential, making it a viable future alternative.

This technology enables surgeons to perform complex dissections and anastomoses with greater precision, dexterity, and visualization, thereby overcoming the limitations of traditional laparoscopic and open approaches [44]. Other advantages offered by the robot-assisted surgical approach include reduced donor morbidity, shorter hospital stays and better aesthetic results [45]. However, despite these benefits, the implementation of robotic surgery in living donor transplants remains limited by several key challenges, including high procedural costs, prolonged operating times, and the need for specialized training [46,47].

Furthermore, while robotic surgery has demonstrated short-term outcomes comparable to laparoscopic techniques in kidney donation, its application in more complex procedures, such as living donor liver transplantation, remains relatively limited due to technical complexities related to vascular and biliary anatomy. In addition to technical and logistical challenges, there is growing debate about the economic viability of robotic surgery in transplantation, as the marginal clinical benefits may not justify the increased financial burden, particularly in decentralized hospital settings with limited resources. Conventional laparoscopic surgery has been performed for many more years than robotic surgery and is considered standard practice, particularly for minor hepatectomy. Another aspect is that, regardless of its complexity, consumables in RAS cost, on average, USD 2000 more per procedure than conventional surgery. Added to this is the high cost of purchasing a robot, which results in an overall increase in costs of almost threefold, making wider adoption of RAS more restrictive. This combination of high costs and novelty has hindered its large-scale application [48].

Costs related to robotic applications, including acquisition, maintenance, and usage, are significant barriers to the widespread adoption of robotic surgical suites. Research indicates that robotic and laparoscopic procedures yield similar outcomes; however, the robotic method comes with higher costs. A cost-oriented meta-analysis on partial nephrectomies indicated similar effectiveness between the two approaches. While robotic surgeries often result in shorter hospital stays, these savings do not offset the operational costs. Studies also reveal that surgeons’ reluctance to adopt robotic technologies rises from concerns about ease of use, complexity, and perceived benefits [49].

However, recent advances in artificial intelligence (AI)-assisted surgery and machine learning-based preoperative planning are beginning to address some of these limitations by improving anatomical visualization and intraoperative decision-making. These innovations promise to reduce operating times and improve the accuracy of organ procurement [50,51].

Finally, it is essential to examine the bioethical dimensions of Robotic-Assisted Surgery (RAS). As several studies have emphasized [52,53], one of the primary concerns involves equitable access to robotic surgery. Due to its considerable cost, disparities in access to care may arise within specific healthcare systems. In wealthier nations, the uneven distribution of robotic platforms may lead to the centralization of such procedures, potentially influencing oncological and functional outcomes in lower-volume centers.

Another significant issue involves the necessity of specialized training and expertise to ensure competent practice and the delivery of high-quality treatment. Although the legal principles governing professional liability remain unchanged, ethical and legal controversies surrounding robotic surgery can be particularly complex. In cases of adverse outcomes, not only the physician and the healthcare institution but also the manufacturer of the robotic system may face litigation.

Moreover, ensuring the safety and reliability of the equipment, providing adequate patient information, and maintaining confidentiality represent fundamental ethical and professional responsibilities [54].

In conclusion, technological innovation and economic evaluation are critical in determining the long-term role of robotics in the future of transplant surgery. Studies demonstrate the superiority of robotic transplant surgery in a wide range of organ transplants, for both donors and recipients. Although robot-assisted living donor transplantation offers significant advantages in terms of surgical precision and donor recovery, it also presents considerable challenges related to cost, complexity, and the need for advanced training.

The future of RTS depends on the efforts of the surgical community to address challenges such as economic and bioethical implications, the need for specialized surgical training for numerous surgeons, and widespread access to robotic systems worldwide [55].

3. Artificial Intelligence Applications in Organ Transplantation

3.1. AI-Driven Donor–Recipient Matching and Allocation

Recent bibliometric analysis, covering 890 publications between 1993 and 2023, reveals that since around 2017, AI applications in kidney transplantation have accelerated, with key research trends including donor–recipient matching, deep learning-enabled post-transplant monitoring, and personalized immunosuppression strategies [56].

Artificial intelligence (AI) methods are increasingly being applied to the complex task of donor–recipient matching in organ transplantation. By integrating large, multidimensional datasets—including donor and recipient demographics, clinical lab values, imaging, histology, and graft quality metrics—machine learning models can outperform traditional allocation scores. For example, ensemble and neural network algorithms have been used to combine metrics like the Kidney Donor Profile Index (KDPI), Estimated Post-Transplant Survival (EPTS), HLA profiles and other factors to predict graft survival more accurately than conventional indices [57]. In one study, a random survival forest model blending EPTS and KDPI achieved an area under the curve (c-statistic) of ~0.64 for 5-year kidney transplant survival, surpassing the standard Kidney Donor Risk Index [57]. Machine learning models such as random survival forests have demonstrated superior performance over traditional indices like the Kidney Donor Risk Index. For instance, a random forest model trained on historical kidney transplant data (70,242 operations) identified approximately 2148 additional successful, long-term grafts compared to the standard risk index, with statistically significant improvement (p < 0.05) [58]. The practical consequence of this enhanced predictive accuracy is a more efficient and equitable allocation system. In practice, these models could reduce organ discard rates and improve long-term graft survival by identifying optimal donor–recipient pairs that conventional systems might miss.

Bioengineering, which uses engineering principles to solve biological problems, enables direct modification of organs outside the body (ex vivo) to overcome immunological barriers—obstacles created by differences in immune system compatibility between donor and recipient. For example, researchers have engineered enzymes within ex vivo organ perfusion systems, which preserve organs outside the body by circulating fluids through them, to remove blood group antigens (such as A or B antigens that trigger immune reactions if mismatched) from the vascular endothelium (the lining of blood vessels) of a donated kidney. This enzymatic modification aims to convert the organ into a functionally “universal” blood group O, potentially eliminating the need for strict blood group compatibility and greatly increasing the transplant success rate for recipients with complex immunogenetic profiles. This approach addresses a significant hurdle that traditional organ matching methods cannot overcome [59].

In general, AI-based allocation supports real-time decision-making: predictive algorithms can forecast post-transplant outcomes (e.g., graft function, delayed graft function) from donor and recipient features [57], thereby enabling dynamic matching. Recent reviews note that AI “optimizes donor-recipient matching beyond traditional metrics” and facilitates equitable allocation by personalizing risk prediction for each patient [57]. In practice, AI recommendations could rank candidates according to individualized survival benefit, complementing existing systems (e.g., MELD, LAS) and potentially reducing waiting times and organ discard rates.

Artificial intelligence systems frequently learn from historical data that often reflects existing health disparities. Consequently, there is a significant risk of algorithmic bias resulting in harmful outcomes [60]. In transplantation, algorithmic bias poses a notable threat to patient outcomes. For instance, predictive models for transplant survival may misjudge benefits for particular demographic groups, potentially decreasing transplant access or resulting in inferior matches. These biases can perpetuate disparities in organ allocation and matching [61].

Artificial intelligence models, including complex neural networks and random forests [57,58], face a critical performance challenge: their predictive accuracy often declines after external validation, particularly when used in new clinical centers, with different patient populations, or under varying surgical protocols. Frequently developed from single institutions or limited registries, these models struggle to generalize to diverse real-world clinical environments [62]. This challenge is intensified by the opaque nature of many high-performance models, where the rationale for recommendations (such as donor–recipient matching or drug dosing) lacks clinical clarity. The demand for explainable artificial intelligence (XAI) is therefore vital to foster trust and enable clinicians to scrutinize and verify the logic underlying AI-driven decisions in high-risk contexts, such as organ transplantation [62,63]. A study by Alnazari et al. developed an explainable AI-based supervised model to predict 30-day readmission risk after kidney transplantation. This retrospective analysis involved 588 patients at the King Abdullah International Medical Research Centre, with the majority receiving transplants from living donors (85.2%, n = 500). The machine learning process consisted of: (1) data processing to clean and prepare records, (2) feature selection of relevant variables, (3) model development with five-level stratified cross-validation, and (4) clinical validation for applicability. Among several algorithms, the gradient boosting model performed best (AUC 0.837, 95% CI: 0.802–0.872), achieving an accuracy of 0.796 ± 0.050 and a sensitivity of 0.388 ± 0.129. Model interpretability was achieved using SHAP and (SHapley Additive exPlanations) and LIME (Local Interpretable Model-agnostic Explanations) [63].

3.2. AI in Immunogenetic and Molecular Analysis

AI techniques also enhance genetic and immunological compatibility assessment. Conventional HLA matching and crossmatch tests have finite resolution, but machine learning can integrate multi-omics data (genomic, transcriptomic, proteomic) to refine risk prediction [64]. For instance, deep learning models have been trained on pre-transplant gene expression profiles to detect immunological risk; Bayesian AI tools (e.g., Bayes-CRE) that combine causal biological networks with transcriptional data can improve robustness and reduce overfitting [64]. Machine learning (ML) has identified novel biomarkers of rejection: Shen et al. used transcriptomic data and ML to pinpoint senescence-associated genes (e.g., p21^Cip1, p16^INK4a) predictive of chronic allograft nephropathy [64]. Commercial AI-based assays like TruGraf (kidney) and kSORT (heart) use blood RNA signatures to noninvasively flag subclinical rejection, guiding immunosuppression adjustments and reducing biopsy needs [64].

In lung transplantation, the use of AI and machine learning (ML) is gaining traction. Ronen et al. (2025) [65] review multiple models developed to predict short-term outcomes such as primary graft dysfunction and time to extubation, as well as long-term survival and chronic lung allograft dysfunction. AI has also been applied to therapeutic drug monitoring, modeling tacrolimus dosing as a time-series problem [65]. Thus, deepening ML integration into clinical decision-making can improve post-transplant outcomes and enhance donor lung utilization through sophisticated predictive analytics.

3.3. Digital Twins and Organ “Avatar” Models

A cutting-edge application of AI is the development of organ “digital twins”—personalized computational models of an individual’s organ or immune system. In medicine, digital twins (high-fidelity simulations updated with patient-specific data) are already used in cardiology and diabetes management [66]. In transplantation science, digital twin models aim to simulate graft function and host response. Their application is in a transitional phase, moving from an experimental prototype with high clinical relevance to a fully integrated reality. There are already examples of functional and specific prototypes. These demonstrate the immediate clinical impact of DTs.

The study by Halder et al. integrates clinical gene expression data into a mathematical model of liver regeneration [67]. The aim is to trace transitions between quiescent, primed, and proliferating hepatocytes to create patient-specific virtual liver models. In a cohort of 12 healthy LDLT donors, researchers utilized whole-transcriptome RNA sequencing data. They identified specific gene expression patterns for liver resection using weighted gene coexpression network (WGCNA) analysis. The models were organized into distinct clusters with unique transcriptional dynamics, and these clusters were mapped to model variables using deep learning techniques. A personalized progressive mechanistic digital twin (PePMDT) was developed for LDLT donor livers. This tool predicts individual patients’ recovery trajectories by leveraging blood gene expression data to simulate regenerative responses. It provides a practical model for personalized medicine [67].

More broadly, clinical workflows like ex vivo organ perfusion offer opportunities for AI-driven twin models. A recent lung transplant study (InsighTx) used machine learning on EVLP (ex vivo lung perfusion) data to predict transplant outcomes. This model achieved high accuracy (AUC ~0.80–0.90) in identifying suitable donor lungs and retrospectively would have increased transplantation of viable lungs while preventing use of poor-quality lungs [68]. In effect, InsighTx functions as a real-time digital surrogate of the perfused lung, systematically interpreting multivariate perfusion parameters (physiological, biochemical, imaging) to guide selection. These models show that specialized AI-driven applications already serve as real-time or near-real-time virtual and predictive replicas. While they are not yet all-encompassing or routine DT systems, they play critical roles at key stages of the clinical workflow.

The concept of digital twin organs—virtual replicas that simulate patient-specific physiology—is gaining momentum across medicine. In transplantation, such models have notable early applications. For example, computational frameworks simulate liver regeneration in living donors by integrating serial transcriptomic data into a mechanistic model, enabling personalized prediction of recovery trajectory [58]. In cardiology, neural ODE-based surrogate models enable high-fidelity electromechanical heart simulations 300× faster than traditional approaches [58]. These models can perform parameter estimation and uncertainty quantification in just hours on a standard laptop [58]. While not specific to transplantation yet, such methods pave the way for real-time, patient-specific organ twins capable of supporting outcome simulation and intervention planning.

In the future, multi-scale digital twins may encompass the immune system as well. The concept of an “immune digital twin”—a dynamic model of an individual’s immune response—has been proposed for precision medicine [62]. Such a twin could simulate graft rejection or tolerance by integrating patient-specific immunological data (e.g., immune cell counts, cytokine profiles, donor-derived cell-free DNA). Preliminary digital twin efforts (e.g., in cardiology) demonstrate that patient-tailored models can improve prognosis and therapy [66]. While Halder et al.’s liver model [67] demonstrates the feasibility of organ-scale DT, the next frontier in transplantology is multiscale Digital Twins. These models encompass both the organ and the whole host. Such virtual organ models could predict complications before clinical signs appear. They can optimize immunosuppression and serve as an in silico test bed for complex clinical scenarios.

3.4. Ethical Implications of AI Applications in Organ Transplantation

As artificial intelligence (AI) becomes increasingly integrated into healthcare, awareness of its potential risks—and of the need for regulatory and ethical frameworks to mitigate them—is also rising. As AI applications in medicine grow in complexity and clinical significance, improvements in data quality and algorithmic transparency are essential to address emerging challenges. This is especially relevant in the context of healthcare’s intrinsic complexity and the high-risk nature of certain surgical procedures.

Although AI offers substantial benefits in the field of transplantation, its adoption and use raise numerous ethical challenges and moral tensions [69,70]. One major concern with embedding AI in clinical workflows is the risk of making care more impersonal, potentially reducing empathy and patient-centered communication.

Moreover, as AI technologies become more widespread, transparency in their design and application is crucial to prevent distrust among patients and clinicians. Issues of accountability also arise: in the case of an adverse outcome, it remains unclear to what extent responsibility should be assigned to the AI system itself.

AI may contribute to diagnostic errors, particularly if trained on poor-quality or biased data, which can exacerbate inequalities in diagnosis, treatment access, and outcomes. These challenges underscore the importance of ensuring fairness, inclusivity, and representativeness in AI system design.

Additionally, data-related biases and concerns about the privacy of sensitive medical information present ethical risks. Addressing these issues is crucial to sustaining trust and maintaining the ethical foundations of healthcare. Therefore, AI systems must be developed and deployed with rigorous attention to data quality, equity, and confidentiality to support accurate, fair, and ethically sound decision-making in the field of organ transplantation [71].

4. Advances in Bioprinting and Tissue Engineering: Towards Functional Organ Vascularization

The growing demand for organ transplants is driven globally by the limited availability of compatible organ donors, and in response to this, disciplines such as tissue engineering and advanced technologies like bioprinting have emerged. The ambitious and shared goal of both is to be able to recreate functional tissues and organs in the laboratory, which can then be implanted in patients who need them, reducing the thousands of deaths each year caused by the shortage of organs for transplants [72,73,74].

The concept of tissue engineering originated in the late 1980s, when Robert Langer and Joseph Vacanti used biodegradable scaffolds to support tissue cell growth [75], and in recent decades, it has become increasingly prominent as a new and innovative discipline that integrates and applies knowledge from cell biology and engineering but also materials science. Tissue engineering relies on several key strategies, including the fabrication of biocompatible scaffolds, organ decellularization and recellularization, the application of induced pluripotent stem cells (iPSCs), and advances in 3D bioprinting [76,77].

The main approaches that can be used, regarding biocompatible scaffolds, are ex vivo and in situ. In the ex vivo method, cells are isolated from the donor, often stem cells. Once isolated, maintained, and expanded in vitro, they are seeded onto biocompatible scaffolds, which in turn are enriched with bioactive molecules (such as growth factors and cytokines) that create a controlled environment for promoting cell adhesion, survival, and differentiation. Once the “engineered” tissue has been created in vitro, it is implanted into the patient who needs it and monitored for possible rejection [78,79]. On the other hand, in the in situ approach, the isolation and seeding of cells from a donor is bypassed, but prefabricated, biocompatible scaffolds are implanted directly into the site to be repaired. The goal is that the scaffold acts as a guide for the recruitment of the body’s cells, leading to the regeneration of damaged tissue. Among the advantages are certainly its simplicity compared to the previous method and the increased immunocompatibility of the technique [80,81,82]. The advantages and disadvantages of these two strategies, described in Figure 2, influence their choice in clinical practice.

Another promising strategy in the field of tissue engineering involves the decellularization and recellularization of an organ. This consists in removing all cellular components from the donor organ, which can be either human or animal, but preserving the extracellular matrix (ECM) and the native three-dimensional architecture, which will need to be maintained in the second phase of implantation. An acellular, biocompatible, and non-immunogenic scaffold is created, providing a structural environment suitable for recellularization with the patient’s cells to significantly reduce the risk of immune rejection [83]. The cells most used in recent years are iPSCs due to their ability to differentiate into virtually any cell type, making them the best candidates for restoring the structure and function of the original organ [84]. The advantages and limitations of decellularization and recellularization using iPSCs in tissue engineering are illustrated in

Table 1. Advantages and limitations of decellularization and recellularization using iPSCs in tissue engineering.

Category	Advantages	Limitations	References
Immunocompatibility	The use of patient-derived iPSCs reduces immune rejection.	Residual antigens in the ECM of the decellularized organ can trigger an immune response.	[85,86,87]
Cell source	Unlimited, autologous, and ethically safe	Maintaining and expanding iPSCs is time-consuming and expensive.
Differentiation	iPSCs can become almost any cell type	Difficult to achieve a full and functional differentiation
Scaffold functionality	ECM maintains native structure and supports cells	Decellularization may damage ECM or leave residues
Regenerative potential	Potential to regenerate whole functional organs	Functional and long-term organ regeneration is challenging
Clinical Translation	Avoids ethical issues; personalized therapy possible	Regulatory and safety challenges (e.g., tumor risk)

[85,86,87]. In summary, the use of iPSCs offers great promise for regenerative medicine due to their immunocompatibility, ethical acceptability, and broad differentiation potential. However, several challenges remain, including the complexity and cost of iPSC maintenance, difficulties in achieving full and functional differentiation, and potential ECM damage during decellularization. While iPSCs and ECM scaffolds provide a strong foundation for organ regeneration and personalized therapies, long-term functionality, safety, and regulatory concerns, particularly the risk of tumor formation, still limit clinical translation.

Despite these advances, one of the main unresolved challenges of tissue engineering applied to organ transplantation remains the creation of adequate vascularization and innervation within the transplanted structure. In fact, the lack of an extensive and functional vascular network leads to limited diffusion of nutrients and oxygen if the new tissue exceeds a few hundred micrometers, just as the lack of adequate innervation compromises the restoration of physiological functions [88]. 3D bioprinting technology helps to overcome this limitation of previous techniques. It is based on the principle of precise deposition of bio-inks: hydrogel-based formulations that can incorporate living cells and ECM components as well as bioactive molecules for cell support and growth. Bioprinting allows the formation of complex, three-dimensional tissue architectures where it is possible to design the vascular and nervous networks necessary to obtain a functional transplanted organ [89].

The classic 3D bioprinting strategies that have become established over the last two decades are extrusion, droplet jetting, laser-assisted, and selective photopolymerization (SLA/DLP). All four techniques use a layer-by-layer production strategy that enables control, which varies depending on the technique used, of the main components of the final product, meaning cells, scaffolds, and biochemical signals [90]. Extrusion-based bioprinting, like fused filament fabrication in conventional 3D printing, involves the continuous deposition of bioink through a nozzle usually controlled by pneumatic systems (piston or screw). The advantage of choosing this method is that it supports a wide range of possible bioinks, including alginate, gelatin, and decellularized extracellular matrix. However, the disadvantages include relatively low resolution compared to other methods and the potential mechanical stress induced by the nozzles on the cells [91]. On the contrary, drop-on-demand bioprinting offers greater spatial precision thanks to the generation of discrete drops that can be produced using different mechanisms, including thermal, piezoelectric, electrostatic, or laser-induced actuation. These systems are suitable for low-viscosity bio-inks because they operate in a non-contact, drop-on-demand (DOD) mode, characteristics that also minimize stress on the individual cell [92]. Similarly, laser-assisted bioprinting also operates in non-contact and DOD mode. In the latter case, however, laser energy is used, which is absorbed by a donor substrate to induce the localized vaporization of microdroplets of bioink on the target surface. The final spatial resolution is exceptional, with precision positioning of individual cells, but obviously requires complex instrumentation that is more expensive than the previous methods [93]. Finally, the last of the classical methods is selective photopolymerization bioprinting, which includes stereolithography (SLA) and digital light processing (DLP). It uses both photopolymerizable bioinks and patterned light exposure to crosslink materials within a kind of reservoir. The structures produced using this method are certainly high-resolution and robust, with the added advantage of relatively fast printing times. However, the limitation of this method lies in its use of specialized bioinks and limited cell loading capacity [94].

An emerging strategy is volumetric bioprinting (VBP), which differs from the layer-by-layer approach of conventional bioprinting described above. Instead, tomographic projections of light or other energy fields are used to solidify an entire three-dimensional construct in a few seconds within a cell-laden hydrogel. The fabrication of centimeter-scale architectures is thus rapid and has been used to create miniaturized cardiac and perfusable liver tissues. The significance of this breakthrough for transplantation is its potential to overcome the critical scaling bottleneck. By enabling the rapid fabrication of centimeter-scale, perfusable tissues, VBP moves the field closer to the practical goal of creating implantable, patient-specific organ patches [95,96,97]. In addition to VBP, other innovative methods are advancing, including acoustic and magnetic bioprinting for cell modeling, melt electro writing (MEW), and hybrid biofabrication platforms that combine printing with organ-on-chip technologies. These techniques aim to overcome current barriers in vascularization, scalability, and functional integration of engineered tissues [98,99,100,101].

Below there is a summary of the most recent 3D bioprinting applications across different organs and tissues, including the external ear [102,103,104,105,106,107], skin [108,109,110,111,112,113,114,115], bone [116,117,118,119,120,121], cornea [122,123,124,125,126], cartilage [127,128,129,130], liver [131,132], heart [133,134], and airways/lungs [135,136,137,138,139,140]. Columns indicate the target organ or tissue, the bioprinting method employed, the specific application, and the corresponding references (

Table 2. Most recent 3D bioprinting applications across different organs and tissues.

Organ/Tissue	Bioprinting Method	Application	References
External ear (auricle)	CAD workflow + extrusion-based bioprinting (hydrogel/cartilage)	Outer ear model (auricle)	[102]
	Pneumatic extrusion bioprinting with cartilage bioink	Auricular cartilage regeneration	[103]
	Extrusion bioprinting with sterile bioink from segmentation	Patient-specific ear implants	[104]
	Extrusion bioprinting with biocompatible/cartilage bioink	Personalized ear reconstruction frameworks	[105]
	Extrusion bioprinting of polymers and auricular bioink	Auricular models reinforced with nanoparticles	[106]
	Extrusion-based bioprinting to create and assemble ears	Patient-specific ear implants for reconstruction surgery	[107]
Skin	Extrusion bioprinting	Human bilayered skin	[108]
	Extrusion/DLP bioprinting	Fibrinogen-based skin	[109]
	Extrusion bioprinting	Vascularized bilayered skin	[110]
	DLP 3D bioprinting	Artificial skin model	[111]
	Extrusion bioprinting (Alg-Gel hydrogel)	Vascularized bilayered skin (MSCs + HUVECs)	[112]
	3D prestress bioprinting	Skin with muscle and endothelial cells	[113,114]
	DLP/extrusion bioprinting	Skin for wound healing	[115]
Bone	3D gel printing	Biphasic calcium phosphate (BCP) scaffolds	[116]
	3D bioprinting (hydrogel)	Hydrogel scaffolds with vessel-like structures	[117]
	3D bioprinting (hydrogel)	Vascularized bone regeneration	[118]
	Inverted light-curing 3D printing	Composite piezoelectric bone scaffolds	[119]
	Low-temperature condensation deposition 3D printing	PLLA + pearl composite scaffolds	[120]
	3D printing (photosensitive resin)	Preoperative bone models	[121]
Cornea	Laser-assisted bioprinting (ADSCs + laminin + collagen I)	Human cornea mimic (epithelium + stroma)	[122]
	Visible light-based stereolithography (GelMA + corneal stromal cells)	Corneal stroma regeneration	[123]
	Digital 3D bioprinting (sodium alginate + gelatin type B + bovine collagen)	3D-printed corneal equivalents for in vitro models	[124]
	Digital Light Processing (DLP) with polyglutamic acid (PG) and calcium carbonate (CC) hydrogel	Artificial cornea with aligned fibrous structure	[125]
	3D bioprinting with collagen-based bioink	Corneal model using human corneal stromal cells	[126]
Cartilage	Inkjet-based printing, extrusion-based printing and laser-assisted printing	Nasal cartilage regeneration	[127]
	3D bioprinting with tissue-specific photoreticulable bioinks	Trachea reconstruction	[128]
	Polycaprolactone/graphene oxide (PCL/GO) scaffolds fabricated using 3D bioprinting	Meniscus	[129,130]
Liver	Spheroid-based bioprinting	Functional hepatocyte organoids	[131]
	Omnidirectional printing embedded network (OPEN)	Hepatic extracellular-matrix and liver (HEAL) construct	[132]
Heart	Hydrogel-based 3D bioprinting	Application in the treatment of congenital heart disease	[133]
	Bioprinting-assisted tissue assembly (BATA)	Tissue mimicking left ventricular myocardial fiber orientation	[134]
Airways and lungs	Bioprinted airway epithelium + vascular network	Modeling asthma, allergen-induced exacerbation, airway inflammation	[135]
	Inkjet-based bioprinting model	Advanced biomimetic in vitro airway models	[136]
	Collagen scaffolds	Enhancing microenvironment for lung regeneration	[137]
	Inkjet bioprinting	Alveoli structural model	[138]
	Bioprinted patient-derived lung cancer organoids + perfusable vessels	Tumor modeling; testing cancer therapies in a vascularized system	[139]
	Light-based bioprinting using food dyes as photoabsorbers	Mimicking oxygenation, ventilation and airway distension	[140]

). Extrusion-based bioprinting emerges as the most widely used technique, particularly for the fabrication of complex, large-scale tissues such as the external ear, skin, and bone, where hydrogels and cartilage-derived bioinks enable the creation of patient-specific and anatomically accurate structures. On the other hand, light-based and laser-assisted techniques have improved precision for delicate tissues like the cornea and cartilage. Studies on liver and heart bioprinting demonstrate the ability to reproduce specific organ microarchitectures and functions, but clinical applications are still limited by the need for complex vascular networks and functional maturation. Finally, airway and lung bioprinting represents one of the most innovative areas, combining cellular complexity with perfusable structures to model diseases such as asthma and cancer. These models are valuable for drug testing and mechanistic studies, but translating them into functional, implantable organs remains an ambitious goal.

Despite these technological advances, quantitative indicators of performance and translational potential remain limited. Recent preclinical studies have reported encouraging results: for example, volumetric bioprinting of hepatic constructs has achieved over 80% cell viability after 14 days and a 60–70% recovery of liver-specific metabolic activity [131,132], while cardiac tissue models reproduced more than 75% of native contractile force [97]. However, clinical applications are still rare, and only early-phase human trials have tested bioprinted skin and cartilage grafts, showing reduced surgical time (−25% compared with manual reconstruction) and lower complication rates (10–15% reduction in infection or graft failure) [107,110,115].

On the other hand, from an economic point of view, there is no evidence, but while bioprinting platforms could initially increase production costs (equipment and bioinks), once the techniques are standardized, operating times and the number of surgical procedures would be reduced, resulting in lower public healthcare costs.

5. Future Frontiers: Xenotransplants, Ethical Challenges, and the Rise of Human–Machine Hybrids

Xenotransplantation, also called heterogeneous transplantation, is a key aspect of transplantation’s future. It involves transplanting cells, tissues, and organs from a non-human donor to a human recipient.

The possibility of xenotransplantation, however, raises public debate at ethical, religious and legal levels.

As defined by the US Public Health Service: “Xenotransplantation is now defined as any procedure that involves the transplantation, implantation, or infusion into a human recipient of live cells, tissues, or organs from a nonhuman animal source or human body fluids, cells, tissues, or organs that have had ex vivo contact with live cells, tissues, or organs from nonhuman animals. In addition, xenotransplantation products have been defined to include live cells, tissues or organs used in xenotransplantation” [141].

The first xenograft in history was performed in Russia in 1682, where a dog skull fragment was used to replace a missing human skull fragment in a wounded soldier. Some sources claim instead that the first dog-to-human bone transplant was performed in 1501 in Iran [142]. Around the early 1920s, the study of xenotransplantation began, including the use of animal grafts. It is estimated that 70 animal-to-human transplants were performed between 1906 and 1995 [143].

Talking about xenotransplantation also involves focusing attention on the ethical and moral limitations underlying this practice. Principlism, a theoretical and moral framework developed by Beauchamp and Childress, groups duty-based concerns using four principles: respect for autonomy, non-maleficence, justice, and beneficence [144]. Regarding the religious question, the three monotheistic religions, such as Christianity, Judaism and Islam, accept xenotransplantation to contribute to the well-being of humanity [145].

One of the main post-operative problems associated with xenotransplantation is the risk of acute rejection. Careful selection of recipients is necessary, as well as adequate screening, to avoid postoperative complications. A retrospective study of pig kidney xenograft cases in baboons revealed that preformed antibodies detectable before transplantation, particularly those high in IgG, were independently associated with HAR. These findings led to the development of a screening algorithm integrating complement-dependent cytotoxicity and antibody binding assays to predict and manage rejection risks [146].

Playing a key role in cellular xenograft rejection is macrophages. These cells have a physiological role in the removal of senescent or damaged cells but have been seen to be the cause of rejection of xenografts, including islet or hematopoietic cell xenografts. In 2022, the first pig-to-human heart transplant was performed on a compassionate basis, and xenotransplantation experiments with pig kidneys in deceased human recipients provided encouraging data. Many advances in the field of xenotransplantation have been achieved through improvements in the ability to genetically modify pigs using CRISPR-Cas9 and other methodologies [147].

In recent years, the development of interfaces capable of replacing or restoring motor functions has had a major impact on human health. Current technologies enable the generation of motor commands from muscle activations and movements of the individual, as in the case of myoelectric prostheses, or from neuronal signals [148]. A study conducted on the use of artificial limbs evaluated the remote control of artificial limbs using robotic components. The aim was to determine whether the pressure sensation provided by controlling an additional robotic finger, known as the Third Thumb, could be utilized to enhance the device’s functionality. Using a local anesthetic and a placebo on two different groups of subjects, they demonstrated that reducing sensitivity in the area controlling the robotic finger (e.g., the numb big toe) makes it harder for people to control the movement and learn to use it over time [149].

Technological development has reached significant milestones in electronic devices that support individuals with disabilities. The human–machine interface (HMI) facilitates the evaluation of assistive devices, enabling more intuitive interaction. Examples include speech recognition for people with disabilities and electromyography (EMG) to control motorized wheelchairs. In 2022, Bouyam and Punsawad presented an innovative HMI system using piezoelectric sensors to translate facial and tongue movements—a promising solution for motorized wheelchair control. Their system addresses two key issues. First, six piezoelectric sensors capture facial muscle signals, allowing for the observation of sensor positions and characteristics during specific movements, such as winking and tongue actions. This precise data capture ensures responsive device interaction. Second, the system was tested for simulated online wheelchair control. In the experiment, twelve volunteers participated and achieved a maximum classification accuracy of 98.0% using linear discriminant analysis and closest K-neighbor algorithms. This high accuracy is essential for the correct interpretation of facial and tongue commands. The algorithm enabled translation models with an average ranking accuracy over 95% and a command creation window of 0.5 s, supporting smooth wheelchair control. Simulated tests confirmed high efficiency under time constraints. Using blinks and tongue actions, users achieved steering times comparable to those of joystick-based control, with response times that were less than half those of a traditional joystick. Thus, this system marks a significant advancement in HMI technology and offers potential application in electric wheelchairs for quadriplegic patients who retain facial or tongue muscle control. This innovation could substantially improve their quality of life by enhancing autonomy and independence [150].

One of the most recent frontiers of scientific research is the development of micro- and nanorobots. Their functionality has improved through the implementation of the use of intelligent materials, adapted to respond to certain conditions, and the integration of artificial intelligence. This makes them one of the emerging tools in the field of biomedical applications, from invasive medicine to targeted drug delivery [151] (

Table 3. Technological advances in medicine.

Technology/Device	Primary Function	Details/Operational Mechanisms	Exemplified Applications
Human–Machine Interfaces (HMI)	Evaluating and implementing assistive devices; replacing or restoring motor functions.	Generate motor commands from muscle activations (e.g., myoelectric prostheses) or neuronal signals.	Voice recognition, Electromyography (EMG) for controlling motorized wheelchairs [148].
Advanced Prosthetics/Controls	Restoration of motor functions; remote control.	Study on the control of artificial limbs using additional robotic components. Sensitivity is crucial for learning and control.	Control of the “Third Thumb” using pressure sensation [149].
HMI System with Piezoelectric Sensors	Device control via facial movements.	Uses piezoelectric sensors to translate facial and tongue movements.	Controlling a motorized wheelchair [150]
Micro- and Nanorobots	Advanced biomedical applications (invasive medicine, drug delivery).	Functionality improved by intelligent materials (adapted to respond to certain conditions) and integration with Artificial Intelligence (AI).	Targeted drug delivery [151].

).

6. Conclusions

Our narrative review has charted significant and independent advancements within robotics, artificial intelligence, and bioengineering, each demonstrating immense potential to address specific facets of the organ transplantation crisis. Robotics enhances surgical precision and reduces donor morbidity, AI optimizes donor–recipient matching and immunological risk prediction, and bioengineering drives innovation in tissue fabrication through 3D bioprinting and scaffold development. Yet the most profound opportunity lies not in parallel development, but in the deliberate and synergistic integration of these fields, creating a new paradigm of transplant engineering. At the core of this synergy is a virtuous cycle of data and precision. In this vision, artificial intelligence serves as the “command center,” processing patient-specific anatomical and immunogenetic data to design personalized organ scaffolds for 3D bioprinting and dynamically matching recipients with both human and bioengineered organs. AI-guided robotic platforms then enhance surgical accuracy and intraoperative decision-making. Robotics act as specialized effectors, translating AI-generated blueprints into reality—whether fabricating complex cellular architectures via bioprinting or performing flawless anastomoses in transplantation surgery. Bioengineering provides the essential building blocks—biocompatible scaffolds and functional tissues—that these systems design and implant. This closed-loop pipeline moves beyond isolated tools towards an interoperable platform for organ replacement. However, the trajectory toward this integrated future is fraught with significant technical, ethical, and clinical risks. From a technical standpoint, the increased reliance on digital infrastructure and data flow makes AI-driven allocation systems and robotic platforms vulnerable to cyberattacks, which could disrupt transplant logistics or, in a worst-case scenario, lead to catastrophic surgical outcomes. Similarly, the failure of a robotic mechanism during a critical phase of vascular anastomosis could compromise the entire graft. In the realm of bioengineering, the long-term immunological profile of bioprinted organs remains a profound unknown. While the use of patient-derived iPSCs aims to minimize rejection, the complex, novel biomaterials and the potential for immature cell phenotypes could trigger unforeseen immune responses or lead to functional failure post-implantation. Ethically, the concentration of these advanced, costly technologies risks exacerbating global health inequities, creating a two-tiered system where access to life-saving transplants is determined by geographic and economic privilege. Therefore, the successful integration of these disciplines is contingent not only on technological breakthroughs but also on the parallel development of robust cybersecurity protocols, fail-safe engineering, long-term safety studies, and equitable implementation frameworks.

To realize this collaborative vision clinically, future research must be strategically directed. Beyond technology, fostering multidisciplinary collaboration among clinicians, engineers, biologists, and data scientists is essential. Addressing scalability, training, and cost-effectiveness is paramount for democratizing access to these advanced therapies worldwide. Ethical and regulatory frameworks must also evolve to ensure equity, safety, and public trust in AI-driven allocation and lab-grown organs.

Critical priorities include:

▪

Conducting large-scale multicenter clinical trials to establish long-term efficacy, cost-effectiveness, and standardized protocols for procedures such as robotic-assisted kidney transplantation, while validating AI-based allocation models against current standards.

▪

Developing integrated AI-robotic platforms that fuse real-time surgical data with preoperative plans, enabling semi-autonomous task execution, intraoperative decision support, and robotic bioprinting of patient-specific constructs.

▪

Advancing bioengineering to replicate not only vascularization but also innervation and immune-modulatory features in bioprinted organs, ensuring long-term physiological integration and survival.

▪

Establishing robust ethical and regulatory frameworks to govern clinical translation, ensuring safety, fairness, and societal acceptance.

In conclusion, the convergence of robotics, AI, and bioengineering constitutes a paradigm shift in regenerative medicine. This integration offers new avenues to overcome organ shortages and improve transplant success beyond reliance on donor availability. By forging strategic synergies and pursuing targeted research, the global transplant community can overcome persistent challenges and usher in an era where organ scarcity no longer limits life-saving treatment (Figure 3).