Types of bioprinting

3D bioprinting encompasses a diverse set of technologies that enable the spatially precise deposition of living cells and bioinks to generate biomimetic structures. Each modality is based on distinct physical principles and offers different advantages and limitations. The main bioprinting approaches are described below, together with their most relevant benefits and drawbacks, as well as representative examples of in vivo success that illustrate their translational potential.

• • Extrusion-based bioprinting (EBB): This technique employs pneumatic or mechanical pressure to dispense biological inks. It is an affordable and widely used method capable of fabricating cells, tissues, organ-on-a-chip systems and even human-scale constructs, with several commercial platforms currently available (Ref. Reference Zhang, Zhao, Xiahou, Wang, Zhang and Yin24). Its main advantages include the ability to print highly viscous bioinks with elevated cell density, to deposit continuous filaments and large structures and to easily integrate multiple materials and gradients (Ref. Reference Kuss and Duan25). However, the shear stress generated in the nozzle during printing can compromise cell viability, and the resolution is generally lower than that of light- or laser-assisted systems (Ref. Reference Malekpour and Chen26). At present, extrusion-based printing remains one of the most widely applied technologies in biomedical research and development (Ref. Reference Malekpour and Chen26). A representative example is the in situ bioprinting of autologous fibroblasts and keratinocytes into a stratified skin-like structure in murine and porcine wound models, which resulted in improved wound healing (Ref. Reference Albanna, Binder, Murphy, Kim, Qasem, Zhao, Tan, el-Amin, Dice, Marco, Green, Xu, Skardal, Holmes, Jackson, Atala and Yoo27).

  Beyond conventional extrusion strategies, emerging methods such as FRESH and SWIFT have been developed to overcome current limitations and expand the potential of bioprinting:

  • ○ FRESH (Freeform Reversible Embedding of Suspended Hydrogels): This approach employs a thermoreversible support bath that acts as a temporary matrix during printing, enabling free nozzle movement and controlled bioink deposition without structural collapse. It allows the fabrication of complex 3D constructs using low-viscosity bioinks while ensuring high structural fidelity and cell viability close to 99.7% (Ref. Reference Kreimendahl, Kniebs, Sobreiro, Schmitz-Rode, Jockenhoevel and Thiebes28). FRESH is particularly advantageous for printing delicate materials such as collagen while preserving the architecture (Ref. Reference Hinton, Jallerat, Palchesko, Park, Grodzicki, Shue, Ramadan, Hudson and Feinberg29), although it requires additional steps for bath preparation and removal (Ref. Reference Brunel, Hull and Heilshorn30).

  • ○ SWIFT (Sacrificial Writing Into Functional Tissues): This strategy generates vascular channels within functional tissue constructs by printing sacrificial gelatine inks at 4 °C into organ-building blocks composed of stem cell–derived multicellular spheroids embedded in a viscous matrix. Upon heating to 37 °C, the gelatine liquefies, leaving perfusable vascular networks that confer functionality to the tissue (Ref. Reference Palaniappan, Ramesh, Maya, Shyam, Jayakumar and Rajendran31). Compared to other approaches, SWIFT enables the fabrication of organ-specific tissues with high cell density and integrated vascular channels, yielding viable and functional constructs (Ref. Reference Bertassoni32). However, to maintain high fidelity, the sacrificial filament diameter must be approximately 400 μm, about twice the size of the spheroids used (Ref. Reference Bova, Billi and Cimetta33).

• • Inkjet-based bioprinting: In this modality, the bioink – comprising a hydrogel prepolymer mixed with encapsulated cells – is loaded into a cartridge connected to a printhead. Thermal or piezoelectric actuators electronically deform the printhead to eject controlled droplets of bioink during printing (Ref. Reference Sabzevari, Pisheh, Ansari and Salati34). Inkjet printing offers high speed, precision, resolution, cost-effectiveness and the ability to deposit different cell types or biomaterials within the same construct. Nonetheless, concerns remain regarding thermal and mechanical stresses, which may affect gene expression and cellular biochemical processes in sensitive cells (Ref. Reference Kumar, Ebbens and Zhao35). Using this technology, bioprinted skin implants have been successfully produced in murine models (Ref. Reference Oropeza, Serna, Furth, Solis, Gonzalez, Altamirano, Alvarado, Castor, Cedeno, Chaparro Vega, Cordova, Deaguero, Delgado, Garcia Duarte, Gonzalez Favela, Leyva Marquez, Loera, Lopez, Lugo, Miramontes, Munoz, Rodriguez, Subia, Zuniga Herrera and Boland36).

• • Light-assisted bioprinting (stereolithography): This approach relies on light-induced polymerisation of a photosensitive bioink, enabling rapid fabrication of tissue structures by solidifying entire layers at once (Ref. Reference Schlauch, Ebbecke, Meyer, Fleischhammer, Pirmahboub, Kloke, Kara, Lavrentieva and Pepelanova37). A major limitation is the use of UV light to activate photoinitiators commonly added to bioinks, which can negatively impact encapsulated cells. Recent progress, however, includes the development of visible-light photoinitiators to mitigate these effects (Ref. Reference Li, Wang, Ma, Chapa-Villarreal, Lobo and Zhang38). Light-assisted bioprinting has been successfully applied to produce scaffolds for spinal cord injury repair in murine models (Ref. Reference Koffler, Zhu, Qu, Platoshyn, Dulin, Brock, Graham, Lu, Sakamoto, Marsala, Chen and Tuszynski39).

• • Laser-assisted bioprinting (LAB): This technique employs pulsed laser energy directed at a multilayer ribbon containing bioink. The laser induces vaporisation of a thin metallic layer, generating bubbles that propel bioink droplets towards a substrate with culture medium. LAB enables high-resolution deposition of cells and hydrogels, with resolutions ranging from picometers to micrometres depending on bioink properties, laser parameters and substrate conditions (Ref. Reference Ventura40). Advantages include high precision, reduced risk of nozzle clogging, deposition of high-density cell suspensions and droplet generation rates up to 5000 per second, with demonstrated potential for in situ printing in animal models and combination with other technologies (Ref. Reference Dou, Perez, Qu, Tsin, Xu and Li41). Nonetheless, LAB faces limitations in productivity, usability and sustained operation, as technical issues often arise after only a few minutes of printing. Furthermore, cellular homogeneity can increase gravitational settling, sometimes requiring additives such as glycerol for suspension stability (Ref. Reference Zoghi42). This technique has been successfully applied in bone tissue regeneration in murine models (Ref. Reference Keriquel, Oliveira, Rémy, Ziane, Delmond, Rousseau, Rey, Catros, Amédée, Guillemot and Fricain43).

• • Microfluidic bioprinting: This technology integrates engineering, physics, chemistry and biotechnology principles to manipulate minute volumes of fluids, cells and biomolecules within microchannels. It enables the fabrication of complex, heterogeneous 3D tissues that more accurately replicate native tissue organisation and functionality (Ref. Reference Aydin44). Microfluidics provides precise control of flow, mixing and bioink positioning, reduces shear stress through laminar sheath flow and enhances morphology and resolution, particularly when combined with extrusion-based printing, extending resolution beyond the ~50 μm limit of microextrusion (Ref. Reference Davoodi, Sarikhani, Montazerian, Ahadian, Costantini, Swieszkowski, Willerth, Walus, Mofidfar, Toyserkani, Khademhosseini and Ashammakhi45). Challenges remain in scaling up tissue structures, handling fragile hydrogel fibres and achieving consistent geometric outcomes (Ref. Reference Zaeri, Zgeib, Cao, Zhang and Chang46). A notable example is the development of a glioblastoma-on-a-chip model that realistically recapitulates tumour ecology, including cellular heterogeneity, ECM complexity, oxygen gradients and dysfunctional vascular interactions, thereby enabling patient-specific therapeutic resistance profiling and drug prioritisation (Ref. Reference Yi, Jeong, Kim, Choi, Moon, Park, Kang, Bae, Jang, Youn, Paek and Cho47).

• • Hybrid and multimaterial approaches: These strategies involve the sequential or simultaneous deposition of two or more bioinks, which may be single-component, homogeneous multicomponent or multiphasic composites, to impart specific properties and functionalities to different regions of the construct (Ref. Reference Yuan, Bai, Li, Fu, Wan, Guo, Zhai, Yi, Liu, Zhou and Lv48). Natural biomaterials such as collagen, fibrin, chitosan, alginate, gelatine and hyaluronic acid provide high biocompatibility but suffer from limited mechanical strength and stability, motivating their combination with synthetic materials such as PCL, PEEK, PLA, PEG or Pluronic, which enhance consistency and mechanical performance in multilayered constructs (Ref. Reference Yuan, Bai, Li, Fu, Wan, Guo, Zhai, Yi, Liu, Zhou and Lv48). Two main strategies are typically pursued: (i) pre-mixing materials to improve printability and mechanical properties, or (ii) integrating distinct structures to combine the individual advantages of each material in terms of strength, resolution, processability and cost (Ref. Reference Navara, Kim, Xu, Crafton, Diba, Guo and Mikos49). Multimaterial scaffolds can be fabricated using separate cartridges, multiscale structuring, or surface modifications with specific chemistries, resulting in improved in vitro and in vivo performance (Ref. Reference Liu, Quan, Vyas and Aslan50). Limitations include dependency on restricted combinations (extrusion with electrospinning or post-processing), limited vascularisation capacity, difficulty in finely tuning material properties during printing and challenges in scaling down to submicron or nanoscale dimensions (Ref. Reference Liu, Quan, Vyas and Aslan50). Using a multimaterial approach with a microfluidic printhead coupled to a coaxial needle, muscle tissue was successfully mimicked and implanted in murine models, achieving partial muscle organisation in vivo with tissue integration and cell viability, although challenges remain in preserving muscle architecture after implantation (Ref. Reference Costantini, Testa, Mozetic, Barbetta, Fuoco, Fornetti, Tamiro, Bernardini, Jaroszewicz, Święszkowski, Trombetta, Castagnoli, Seliktar, Garstecki, Cesareni, Cannata, Rainer and Gargioli51).

Medical image acquisition and processing

The acquisition of medical images is the first step in the generation of personalised scaffolds by 3D bioprinting. This step allows detailed analysis of tissue anatomy and pathology and provides a solid basis for the design of biomimetic three-dimensional structures.

Medical imaging techniques include X-ray, CT, PET, magnetic resonance imaging (MRI), single photon emission computed tomography (SPECT), digital mammography and diagnostic ultrasound (Ref. Reference Hussain, Mubeen, Ullah, Shah, Khan, Zahoor, Ullah, Khan and Sultan52). More recently, ultrasound has been successfully used in the cardiovascular field, including 3D transthoracic echocardiography (TTE) and transesophageal echocardiography (TEE). 3D rotational angiography has also been employed. In addition, the combination of different imaging modalities, such as CT and TEE, allows the creation of hybrid 3D models that capture both structural and valvular morphology (Ref. Reference Giannopoulos and Pietila53).

The biological nature of the tissue of interest will dictate the selection of a particular imaging technique. For example, in the case of computed tomography (CT), tissue densities are directly related to pixel intensities, allowing the analysis of structures with high (e.g. bones) and low (e.g. lungs) densities, while MRI is more appropriate to appreciate differences between soft tissues (such as the white and grey matter of the brain) (Ref. Reference Bücking, Hill, Robertson, Maneas, Plumb and Nikitichev54).

CT is an imaging technique that uses iterative algorithms to reconstruct the interior of large volumes using graphics processing units (GPUs) (Ref. Reference Chghaf and Gac55). Due to its higher resolution and its ability to provide information about different tissue densities, it is a more interesting technology for medical imaging for bioprinting, as it facilitates both the generation of the 3D model and the choice of the appropriate biomaterial.

DICOM image processing

3D bioprinting based on Digital Imaging and Communications in Medicine (DICOM) images starts with the conversion of 2D medical images, which are stacked, and are later processed to create a 3D printing compatible format (.STL). This process involves the segmentation of the DICOM images into a three-dimensional computer-aided design (CAD) format. From the data obtained, initial processing is performed, such as defining and adjusting the region of interest (ROI) (Ref. Reference Kamio, Suzuki, Asaumi and Kawai56).

CT or MRI images are stored in DICOM format in the hospital’s Picture Archiving and Communication System (PACS) (Ref. Reference Quan, Wang, Du, Li and Zhou57) from where they are obtained for processing and subsequent bioprinting of the desired tissue. In case you want to use open sources, some such as DICOM Library, TCIA or NIH Chest X-ray Dataset contain numerous freely accessible DICOM files, which can be downloaded following the guidelines described by each.

The first step in converting DICOM images to .STL format is to improve the quality of the images through a thresholding and segmentation process using software that allows the CT image to be converted into a 3D model (Ref. Reference Sahai and Gogoi58).

A further improvement in the processing of DICOM images came from the use of virtual topologies combined with mesh optimisation algorithms. Recent studies have shown that these techniques make it possible to reduce the number of elements in the mesh, improving the resolution of the geometry without compromising the structural quality of the 3D model, which significantly reduces processing time (Ref. Reference Fernández-Tena, Marcos, Agujetas and Ferrera59). This approach is particularly useful in complex geometries, such as airways, where anatomical details are critical for accurate simulation.

Some suitable software for this procedure includes 3D Slicer, OsiriX, ITK-Snap, Invesalius, MeshLab or Blender. In the following, the use of 3D Slicer will be explained because of its open-source status and its widespread use in DICOM image analysis.

3D Slicer

3D Slicer is an open-source 3D data analysis and visualisation platform designed to facilitate the use of medical and scientific images. The software stands out for its ability to perform advanced segmentation and analysis, allowing users to segment medical images in two and three dimensions, which facilitates the identification and isolation of specific structures, as well as allowing the manipulation of volumetric and surface models in different planes, such as axial, sagittal and coronal.

It supports multiple data formats, such as DICOM, Neuroimaging Informatics Technology Initiative (NIfTI) or sections in .dcm format, allowing the integration of images from different modalities such as CT, MRI and ultrasound. In addition, 3D models can be exported in printable formats such as STL, facilitating the creation of physical models from medical images.

Using 3D Slicer to process medical images

Image processing begins by selecting the Data button and Choose file(s) to Add in the main 3D Slicer interface, from where it is possible to import individual DICOM files or entire folders containing multiple studies. In this case, to load a lung CT in .dcm format, all files corresponding to the sections must be selected. Once the files are uploaded, patient data such as patient name, ID, gender or year of birth, among others, can be viewed. The loaded file is then selected, and the Load option is chosen and the images automatically appear in the corresponding views (axial, coronal and sagittal), allowing a detailed visualisation of the different anatomical sections. In the case of Figure 1, a lung CT file has been chosen.

Figure 1. Detailed visualisation of the different anatomical sections.

Next (Figure 2), the desired area must be cropped using the ROI, choosing the Volume Rendering module, with the Crop option activated. By clicking on the Display ROI option (the eye) it is possible to delimit the area to be processed.

Figure 2. Selection of the desired area using the ROI (region of Interest).

Subsequently, by searching the Crop Volume module, the delimited area can be accessed to facilitate segmentation, and after determining the cutting options, the Apply option is chosen. With this step, the area of interest is obtained for processing.

The next step involves adjusting contrasts to define areas that are not visible using the default settings via the Volumes module. The programme allows to analyse the images using default options, so in this case the CT-Lung option will be selected.

Next, the segmentation of the anatomical structures of interest must be carried out. To do this, access the Segment Editor module, where a new segment is created by choosing the Add option, on which the available segmentation tools will be applied (Table 1):

Table 1. Description of the tools available in 3D Slicer

One of the most widely used tools is Threshold, which allows selecting and isolating anatomical structures based on the density or intensity values of the images, such as bone or soft tissues. To do this, in the generated segment, the density of the structure of interest (in this case, the lung) must be selected with this tool, dragging until the desired area is selected, and allowing it to be differentiated from the rest of the image. The selection will be validated with the Apply option. Threshold values can be interactively adjusted to ensure accurate selection of relevant structures. If necessary, manual adjustments can be made using tools such as Paint, Erase or Scissors to ensure optimal segmentation.

Using the Show 3D option, the result of the insulation of the structure can be observed after the necessary adjustments have been made (Figure 3).

Figure 3. Isolation of the desired anatomical structure using the Threshold tool.

Once the segmentation is finished, it is possible to improve the quality of the 3D model by applying smoothing algorithms through the Smoothing tool of the segment editor. This procedure reduces the irregularities present at the edges of the structures, improving the representation of the segmented surfaces. The level of smoothing can be adjusted according to the specific requirements of each case, providing precise control over the quality of the final model.

When the segmentation has been refined, the next step is to convert it into a 3D model (Figure 4). This is done using the Export to files option within the Segmentations module. The STL format is one of the most common formats for this purpose.

Figure 4. Export of the isolated 3D object obtained from the DICOM file to 3D printable .STL format.

Although 3D Slicer provides robust tools for STL model segmentation and generation, in some cases it can be useful to perform additional file optimisation prior to printing. This can be done using post-processing tools such as Meshmixer or Blender, which can repair errors in the mesh, simplify complex geometries and improve the structure of the model to ensure a successful print.

Slicing and its impact on the final structure

For 3D models to be bioprinted, .STL files must go through two processes: slicing and optimisation of the bioprinting parameters. In 3D printing, slicing refers to the process of converting a 3D model into a series of 2D layers or slices that will be sequentially printed, using algorithms specifically designed for this function (Refs Reference Neel, Mesto and Hascoet60, Reference Zhao, Yan, Zhang, Zhang, Pan, Yuan, Xiao, Jiang, Wei, Lin and Chen61).

Types of slicing

The different slicing methods can be initially grouped into two families: planar slicing algorithm (flat) and non-planar slicing algorithm (out of plane or curved). Each of the different slicing techniques has advantages and disadvantages depending on geometry, type, amount of material and purpose.

Planar slicing algorithms

Printing occurs layer by layer, in one plane. They are the most common method of slicing because of their simplicity and because they are accessible to more types of technologies. However, they offer poor surface quality and require the use of supports so that the bioprinted structure does not collapse during printing, which increases material usage and decreases the speed with which we can obtain the final product (Ref. Reference Nayyeri, Zareinia and Bougherara62).

This type of slicing considerably limits the performance of additive manufacturing (AM) systems, causing staggered surfaces, massive support structures, non-conformity with curved substrates and reduced strength in thin structures (Ref. Reference Shan, Gan and Mao63). The most common types of Planar Slicing are:

• • Uniform slicing or Constant Layer Thickness Planar Slicing (Figure 5): the most common method in 3D printing, in which the 3D model is cut into thin flat layers of the same thickness, which are deposited one by one by the printer (Ref. Reference Shan, Gan and Mao63). The height of the layers is a key factor in the quality of the printed object, as the thinner the layers, the higher the resolution and detail of the final object. However, this flat layer-based cutting considerably limits the performance of AM systems, leading to staggered surfaces, massive support structures, non-compliance with curved substrates and reduced strength in thin shell structures (Ref. Reference Hu64).

• • Adaptive Planar Slicing (Figure 6): individual layers of the same part can have different layer heights, so details can be produced with high resolution, while non-detailed areas are produced with higher layer heights, improving production speed (Ref. Reference Pelzer and Hopmann65) although it only improves the representation of curved surfaces. In addition, because the part is still made of flat layers, the staircase effect is not eliminated. It represents an optimisation compared to the previous method, as it allows a better surface finish to be obtained by reducing the staircase effect, with an optimised printing time (Ref. Reference Pelzer and Hopmann65), and requires a much longer planning and adjustment stage than the previous techniques described (Ref. Reference Lettori, Raffaeli, Borsato, Pellicciari and Peruzzini66).

Figure 5. Uniform slicing by flat layers of the same thickness.

Figure 6. Adaptive planar slicing with layers of different heights.

Almost all fused deposition modelling (FDM) bioprinters on the market are capable of printing from planar sliced 3D models, as these are the simplest slicings. The minimum requirements are that the extrusion can be performed in an XY plane and that they have 3 Degrees of Freedom (DOF). Examples of models that allow printing from planar sliced 3D models are NovoGenTM MMX Bioprinter, developed by Organovo and BIO X by CELLINK.

Non-planar slicing algorithms

Compared to planar slicing algorithms, non-planar slicing algorithms allow more DOF at the cost of being much more complicated in terms of procedure and computational processing (Ref. Reference Zhao and Guo67) because, as the printing is not limited to a single plane, it is necessary to determine the toolpath of the print head very well to avoid collisions in its path (Ref. Reference Nayyeri, Zareinia and Bougherara62). All these techniques require the 3D bioprinter hardware to be able to print in more than one plane and to have at least 5 DOF with computerised robotic assistance, in addition to higher computing power (Ref. Reference Bhatt, Malhan, Shembekar, Yoon and Gupta68). The most common types of non-planar slicing are:

• • Curved-Layer Slicing (Figure 7): the 3D model is interpreted in curved layers instead of flat layers, achieving a higher contour fidelity of the final object and reducing the need to include structural supports (Ref. Reference Zhao and Guo69).

• • Conformal Slicing (Figure 8): the cuts are conformal to the preconstructed layer/substrate, so they can be flat or non-flat (Ref. Reference Kapil, Negi, Joshi, Sonwane, Sharma, Bhagchandani and Karunakaran70). It allows printing a 3D structure on a free-form surface on which slices and layering are performed, as the z-axis coordinate changes continuously on the same layer, depending on the complexity and topology of a slice surface (Ref. Reference Alkadi, Lee, Bashiri and Choi71). It requires the use of more sophisticated algorithms and GPU demanding software to translate 2D patterns to non-planar 3D surfaces (Ref. Reference Alkadi, Lee, Bashiri and Choi71).

• • 5-Axis Dynamic Slicing (Figure 9): uses a multitude of planes to do the slicing for the different sections of the 3D model, selecting the cutting plane that best suits the needs, presenting a method that significantly reduces the support structures and is not very costly in terms of time and processing capacity that optimises the final result (Ref. Reference Wang, Zhang, Hu, Liu and Lammer72).

• • Helical Slicing (Figure 10): forms a single continuous 3D toolpath and eliminates seam defects, where the geometry is initially cut into flat slices and then, using two consecutive flat slices, direction vectors are constructed from the current layer to the next layer (Ref. Reference Yigit and Lazoglu73).

• • Mixed-Layer Adaptive Slicing (Figure 11): combines planar and non-planar slicing methods to optimise the printing process for different target subvolumes. This method takes advantage of the benefits of planar splicing and non-planar slicing to present a much higher quality result following a much more suitable manufacturing process (Ref. Reference Zhao and Guo69).

Figure 7. Curved-layer slicing.

Figure 8. Conformal slicing where the cuts are carried out on a previously existing surface.

Figure 9. 5-Axis dynamic slicing where you can see the constant change of planes to generate the structure in the most optimal way.

Figure 10. Helical slicing where it can be seen that the trajectory is continuous and ascending.

Figure 11. Mixed-layer adaptive slicing where the combination of uniform slicing and 5-axis dynamic slicing can be observed.

Some of these algorithms require a bioprinter of at least 5 DOF, such as the RoboPrint developed by Acoustic Robotic System Lab and Maxon, or 6 DOF, such as BioAssemblyBot by Advanced Solutions.

The choice of each type of slicing depends on both the technology available and the requirements of the result.

Parameter optimisation in biological models bioprinting

Once the image has been processed by slicing in the most appropriate way and considered what bioprinter is available, it is necessary to adjust the technical parameters that influence both the structural precision and the viability and functionality of the bioprinted cells and tissues. Thus, bioprinting parameters are defined as those settings or firmware inputs necessary to produce bioprinted structures in an appropriate way (Ref. Reference Sánchez87).

The range of values suitable for bioprinting depends mostly on the type of bioink used and their physicochemical properties. Therefore, it is important to extensively characterise it before determining which parameters to adjust in the bioprinter before starting the bioprinting process. In the case of extrusion bioprinters, the most widely used due to their versatility, relatively low cost and capacity to generate thicker and bulkier structures, the main parameters to be adjusted are:

1. 1. Extrusion pressure: is defined as the force applied to the bioink to push it through the nozzle of a bioprinter during the printing process. During the extrusion process, cells are exposed to mechanical forces of different types, which can play an important role in their differentiation, motility and growth control (Ref. Reference Wong, Chan, Kamm and Tien88). Of the different forces to which the cells are exposed, the one that has the greatest impact on them is shear stress, and is the main cause of cell death (Ref. Reference Cidonio, Glinka, Dawson and Oreffo8). Therefore, because a higher pressure triggers a higher shear stress in extrusion bioprinting, causing more cell death (Ref. Reference Han, Kim, Jin and Kim89) it is advisable to reduce the pressure to the minimum allowed by the rheological properties of the bioink (Ref. Reference Thakare, Pei, Qin and Jerpseth90).

2. 2. Printing speed: is the speed at which the nozzle moves over the bioprinter bed and is defined as the flow rate divided by the feed rate. It is a dominant factor in printing results, affecting the height, width and thickness of the printed structures (Ref. Reference Gillispie, Han, Uzun-Per, Fisher, Mikos, MKK, Yoo, Lee and Atala91). It is directly related to the previous parameter, as higher extrusion pressure requires higher printing speed, so speed adjustments must be made to minimise shear stress and maintain cell integrity (Refs Reference Fakhruddin, Hamzah and Razak92, Reference Webb and Doyle93).

3. 3. Nozzle diameter: nozzle diameter affects bioprinting by affecting the shear stresses caused by the internal pressure of the extruder nozzles, which affects printability and cell viability (Ref. Reference Magalhães, de Oliveira, Dernowsek, Casas and Casas94). Thus, reducing the nozzle diameter in bioprinters increases the shear stress, which can lead to cell death due to mechanical damage (Ref. Reference Jentsch, Nasehi, Kuckelkorn, Gundert, Aveic and Fischer95).

4. 4. Nozzle geometry: nozzle geometry, including radius, length and angle of convergence, significantly influences shear stress, which, in turn, influences cell viability (Refs Reference Gómez-Blanco, Pagador, Galván-Chacón, Sánchez-Peralta, Matamoros, Marcos and Sánchez-Margallo96, Reference Chand, Muhire and Vijayavenkataraman97). For example, cylindrical nozzles tend to generate the lowest shear stress, but this stress is maintained over a longer period, which can reduce cell viability (Ref. Reference Chand, Muhire and Vijayavenkataraman97) while conical and truncated conical nozzles can increase the flow rate and reduce the dispensing pressure (Ref. Reference Gómez-Blanco, Pagador, Galván-Chacón, Sánchez-Peralta, Matamoros, Marcos and Sánchez-Margallo96). On the other hand, coaxial nozzles allow the creation of microchannels within the printed structures, improving cell viability by facilitating nutrient delivery (Ref. Reference Gao, He, Fu, Liu and Ma98).

5. 5. Temperature: influences bioink viscosity, cell viability and structural accuracy. Proper temperature control ensures that the bioink maintains a suitable viscosity for extrusion, guarantees cell survival (usually around 37 °C) and regulates the gelation of the biomaterials used, which is essential to maintain the printed structure without deformation. It also affects the final resolution and quality of the bioprinted tissues (Ref. Reference Hospodiuk, Dey, Sosnoski and Ozbolat99).

6. 6. Layer height: critical in determining the resolution and accuracy of printed structures, as well as the cell viability and mechanical integrity of the tissue. Thinner layers allow for greater precision and definition in detail. (Ref. Reference Murphy and Atala19), but may increase printing time and generate higher mechanical stresses on cells, while thicker layers speed up the process and reduce cellular stress, but may compromise resolution and homogeneity of tissues (Ref. Reference Ozbolat and Hospodiuk100). The balance between layer height, resolution and cell viability is key to the success of the process.

Materials and cells used in bioprinting

The appropriate selection of materials and cells depends on the tissue to be bioprinted. Choosing the right components is essential for the successful reproduction of complex biological structures. Biomaterials used as bioinks must meet specific requirements, such as being biocompatible, promoting cell adhesion and possessing mechanical and functional properties comparable to the target tissue. These materials can be derived from natural sources, synthetic sources or combinations of both (Ref. Reference Groll, Boland, Blunk, Burdick, Cho, Dalton, Derby, Forgacs, Li, Mironov, Moroni, Nakamura, Shu, Takeuchi, Vozzi, Woodfield, Xu, Yoo and Malda101). In addition, selected cells must be able to proliferate, differentiate and organise themselves into 3D architectures that mimic the functions of the tissue or organ to be bioprinted. To this end, the interaction between bioinks and cells is key, ensuring that the printed structure maintains its viability and functionality, both during the printing process and after implantation or maturation (Ref. Reference Jang, Park, Gao and Cho102). The selected bioink should be tested for its compatibility with the growth and function of a particular cell type and, therefore, for the development of a particular tissue. Table 2 exemplifies successful combination of recreated tissues with cell types.

Table 2. Bioprinting of tissues and the types of bioinks and cells used

Functional test of bioprinted constructs

Testing cell viability and functionality of bioprinted constructs is an essential step to ensure the generated structures perform as expected. During the bioprinting process, cells are exposed to changing environmental conditions (e.g. temperature, viscosity, mechanical forces) that can impact their viability and function (Refs Reference Willson, Ke, Kengla, Atala, Murphy and Crook103, Reference Zhang, Haghiashtiani, Hübscher, Kelly, Lee, Lutolf, McAlpine, Yeong, Zenobi-Wong and Malda104). There exist different methods to assess cell viability. However, they were originally designed for 2D conditions and they do not often perform well in 3D conditions. We propose the combination of methods for an adequate assessment of cell viability in these conditions, including (Refs Reference Lee, O’Connell, Onofrillo, Choong, Di Bella and Duchi105, Reference O’Brien, Haskins, Taylor, Haskins and Giuliano106, Reference Bikmulina, Kosheleva, Efremov, Antoshin, Heydari, Kapustina, Royuk, Mikhaylov, Fomin, Vosough, Timashev, Rochev and Shpichka107):

• - Live/Dead staining: a methodology for the assessment of cell viability in both 2D and 3D cultures. It is widely employed through the combination of two fluorogenic compounds: calcein acetoxymethyl ester (calcein AM) and propidium iodide (PI). Differentiation between viable and non-viable cells is based on cell metabolic activity and plasma membrane integrity. The major component of this assay is Calcein AM, a non-fluorescent molecule hydrolysed in viable cells by intracellular esterases into calcein, a highly fluorescent substance. This process occurs exclusively in cells with active metabolism and intact membranes, as only living cells possess the requisite enzymatic activity to release the acetoxymethyl group from Calcein AM and convert it into a fluorescent compound (Ref. Reference Miles, Lynch and Sikes108). The emission wavelength of calcein is approximately 515 nm, which permits its detection through the use of fluorescence microscopy or flow cytometry (Ref. Reference So, Sallin, Zhang, Chan, Sahni, Schulze, Davila, Strome and Jain109).

  In contrast, PI is a marker that only permeate to cells with compromised plasma membranes, which is indicative of non-viable cells. PI is positively charged, rendering it unable to cross intact cell membranes. However, in damaged cells, it is capable of passing through and binding to cellular DNA and RNA (Ref. Reference Liu, Gao, Li and Sun110) via electrostatic interactions with the phosphate groups of nucleic acids. The intercalating bond between propidium iodide and DNA markedly enhances the fluorescence of the latter, resulting in the emission of light at a wavelength of approximately 617 nm upon excitation.

  The live/dead staining technique is widely used to assess viability in conventional cell cultures. Its use in 3D cell cultures can be compromised due to limited penetration of the compounds, therefore to obtain a more accurate quantitative assessment of viability in 3D models, it is recommended to combine this technique with other complementary methods to improve the accuracy of in-depth detection (Ref. Reference Bikmulina, Kosheleva, Efremov, Antoshin, Heydari, Kapustina, Royuk, Mikhaylov, Fomin, Vosough, Timashev, Rochev and Shpichka107).

• - Alamar Blue (AB) assay: a widely used technique for assessing the metabolic activity of cells based on the reduction of the non-fluorescent compound resazurin to the fluorescent compound resorufin by enzymatic reductases using reducing metabolites such as NADH, NADPH or FADH (Refs Reference Lavogina, Lust, Tahk, Laasfeld, Vellama, Nasirova, Vardja, Eskla, Salumets, Rinken and Jaal111, Reference Vieira-da-Silva and Castanho112). The resorufin produced can be quantified by absorbance, measured at a wavelength of 570 nm, or by fluorescence, using an excitation wavelength of 540 nm and an emission of 590 nm (Ref. Reference Lavogina, Lust, Tahk, Laasfeld, Vellama, Nasirova, Vardja, Eskla, Salumets, Rinken and Jaal111), allowing sensitive and quantitative detection of cell activity.

  In bioprinting applications, cell encapsulation within bioprinting gels can impact the assay’s performance. Key variables that may affect assay effectiveness include the gel’s concentration, thickness and water content. For instance, in hydrogels with high water content, such as those based on collagen, AB dye dilution can occur, reducing assay performance by up to 25% (Ref. Reference Bonnier, Keating, Wróbel, Majzner, Baranska, Garcia-Munoz, Blanco and Byrne113). To mitigate these effects, it is recommended to calibrate assay conditions according to the specific properties of the bioprinted material being used.

• - PicoGreen Assay: this technique measures the DNA content of cells encapsulated in the gel, which can be used to allow an estimation of their proliferation. The technique involves breaking the cell membranes to release the DNA and measuring its concentration by fluorescence. Therefore, is not suitable for continuous monitoring of cell proliferation.

Therefore, for optimal gel/3D structure characterisation, the effects of hydrogel structure, water content and porosity have to be considered when selecting a cell viability assay.

Challenges and limitations for the clinical translation of 3D bioprinting

The clinical application of bioprinting still faces numerous challenges that must be addressed. With regard to immunological safety and scaffold immunogenicity, it is known that hydrogels and scaffolds can trigger innate immune responses and capsular fibrosis, impairing graft integration, perfusion and function (Ref. Reference Mariani, Lisignoli, Borzì and Pulsatelli114). In certain natural bioinks, impurities are critical determinants. For example, alginate is often contaminated with endotoxins, proteins and polyphenols capable of inducing inflammation and cellular overgrowth within capsules, making purification essential to mitigate these effects (Ref. Reference Dusseault, Tam, Ménard, Polizu, Jourdan, Yahia and Hallé115). Similarly, decellularised ECM-derived matrices provide biomimetic signals, but residual immunogenicity largely depends on meeting minimal decellularisation criteria, including the content and size of nuclear and mitochondrial DNA fragments, the presence of reactive oxygen species (ROS), and fragmented ECM components such as hyaluronic acid or fibronectin (Ref. Reference Kasravi, Ahmadi, Babajani, Mazloomnejad, Hatamnejad, Shariatzadeh, Bahrami and Niknejad116).

Sterilisation is another critical factor in the development of new bioinks for clinical use, as it can significantly alter key properties. For instance, GelMA sterilisation by autoclaving or ethylene oxide reduces stiffness, whereas gamma irradiation increases stiffness, alters pore size, and may compromise sol–gel transition, printability and 3D cell viability (Ref. Reference Rizwan, Chan, Comeau, Willett and Yim117). Consequently, sterilisation procedures for each bioink must be systematically evaluated alongside other characterisation tests to fully understand the modifications introduced during processing.

The viability of bioprinted tissues also remains constrained by limited oxygen and nutrient diffusion in the absence of a fully functional vascular network. Although printing strategies that rely on diffusion can produce perfusable channels and enhance molecular transport within constructs, maintaining stable perfusion and effective vascular integration in vivo remains a major hurdle for the development of volumetric tissues (Ref. Reference Cai, Kilian, Mejia, Rios, Ali and Heilshorn118).

The choice of cell source also represents a key translational barrier. Autologous cells minimise the risk of immune rejection but complicate logistics and extend application timelines. Allogeneic cells offer greater scalability but may elicit donor-specific antibodies and consequent immune responses (Refs Reference Sanabria-de la Torre, Quiñones-Vico, Fernández-González, Sánchez-Díaz, Montero-Vílchez, Sierra-Sánchez and Arias-Santiago119, Reference Chen, Lv, Chen and Cui120). Induced pluripotent stem cells (iPSCs) present another alternative, although potential links to tumorigenicity have been suggested, underscoring the need for further studies and rigorous genomic and functional quality controls before clinical implementation (Ref. Reference Zhong, Liu, Pan and Zhu121).

Finally, one of the most critical barriers to clinical translation is the lack of standardisation, quality control and reproducibility of bioinks. Batch-to-batch variability in natural polymers and ECM-derived materials, the degree of functionalisation, the type of crosslinking agent, the light source used in photopolymerisation and rheological properties, among others, directly affect viability, printing fidelity and mechanical performance. Therefore, the establishment of robust bioink standardisation protocols and reproducible quality control systems is essential to achieve successful clinical translation of bioprinted constructs.