Additively Manufactured Porous Ti6Al4V for Bone Implants:A Review

Ti-6Al-4V (Ti64) alloy is one of the most widely used orthopedic implant materials due to its mechanical properties, corrosion resistance, and biocompatibility nature. Porous Ti64 structures are gaining more research interest as bone implants as they can help in reducing the stress-shielding effect when compared to their solid counterpart. The literature shows that porous Ti64 implants fabricated using different additive manufacturing (AM) process routes, such as laser powder bed fusion (L-PBF) and electron beam melting (EBM) can be tailored to mimic the mechanical properties of natural bone. This review paper categorizes porous implant designs into non-gradient (uniform) and gradient (non-uniform) porous structures. Gradient porous design appears to be more promising for orthopedic applications due to its closeness towards natural bone morphology and improved mechanical properties. In addition, this paper outlines the details on bone structure and its properties, mechanical properties, fatigue behavior, multifunctional porous implant designs, current challenges, and literature gaps in the research studies on porous Ti64 bone implants.


Introduction
The history dates to the time of the late 1960s when the first use of porous metal structures in orthopedics was reported, and ever since the use of these products has continuously increased [1]. In the early 1970s, porous metal development experienced several breakthroughs, setting the stage for future in-growth materials in orthopedic surgery. Galante et al. [2] pioneered the development of fiber-metal leading to its clinical use as a porous coating in hip and knee arthroplasty. Simultaneously, works on the sintering process laid the foundation for the porous cobalt-chromium (CoCr) alloy coatings in use today [3]. Most recently, a new series of highly porous metals has been developed and released for use in orthopedic surgery, including porous tantalum, porous titanium (Ti), and their alloys [3]. This increasing trend in reconstructive surgery is apparent as solid metal implants have high stiffness and strength, however, porous implants are optimal for this use as they allow bone in-growth through the open porosities and have an improved fixation due to their high surface roughness. In addition, these porous implants have lower stiffness and prevent stress-shielding [1,4].
According to the Agency for Healthcare Research and Quality [5], more than 450,000 total hip replacements and just under 700,000 total knee replacements are performed annually in the United States. The 60+ years-old group has a global population that is expected to double from 2020 to 2050 due to both overall population growth and an expected increase in life expectancy from 73 to 77 during this period [6]. This continued increase in the elderly age population demands an increase in the efficiency of the orthopedic implant, which lowers implant failure and the need for revision surgery as they add physical, financial, and psychological burden to the patient. Different implant materials including polymers [7][8][9], ceramics [10][11][12][13][14], metals [15][16][17][18][19][20], and their composites [21][22][23][24][25] are currently being tested for manufacturing orthopedic implants. Most metals fall under the bio tolerant category, however, titanium and its alloys in certain conditions are bioinert in nature [26]. Among these implant materials, titanium is the best base material for additive manufacturing of orthopedic implants due to its excellent mechanical strength, natural good corrosion resistance, Magnetic Resonance Imaging (MRI) compatibility, acceptable in vivo biocompatibility, and bio-adhesion [27,28]. Despite these advantages, the presence of a large metallic foreign body creates obstacles such as promotion of chronic inflammation, leaching, infection and biofilm formation [29].
Titanium implants form protective layers of titania (TiO 2 ) that are a few nanometers thick when exposed to air/water [30,31]. Toxicological concerns such as yellow nail syndrome arise when any damage to this layer occurs as a result of aggravated leaching of debris and nanoparticles from Ti implants [30,32]. These leached Ti particles in surrounding tissues can range between 10 nm to 30-60 µm and the average concentration of locally leached Ti ions are 205-210 ng per mg of the implant's weight [30,33,34]. This phenomenon is more severe in the sintering-based AM process where consolidation of powder occurs below the melting point and results in disintegration/weakening of the implant structure over time, along with the release of Ti particles [35]. The release of these Ti particles has long term adverse effects such as allergic reactions, mutagenesis, carcinogenicity and hypersensitivity reactions, which ultimately result in implant rejection [32,35,36].
Among Ti alloys, commercially pure titanium (CP-Ti) and Ti64 are the most preferred Ti alloys for orthopedic implants [37][38][39]. However, due to the stress-shielding effect of the solid Ti, the research focus has shifted towards the porous Ti alloys implants. Additive manufacturing of porous Ti64 orthopedic and spinal implants with complex porous geometries is often used to achieve their osseointegration properties [40]. The bone implants are expected to have: 1. Good biocompatibility; 2. Appropriate pore sizes and porosity, which are suitable for bone cell infiltration and growth; 3. Comparable mechanical properties with adjacent bone tissue; 4. Osteoconductivity and osteoinductivity; and 5. Biodegradability, such that no traces of the original prosthesis are found after bone healing without any side effects of the degradation products on the human body [41].
The long-term clinical success of bone implants requires the fabrication of porous structures with consideration of elastic modulus, compressive and fatigue strength, wear hardness, and corrosion resistance properties together with biomaterials and biomedical engineering aspects [42]. The research focus on implants is given to specific clinical challenges, such as stress-shielding and osseointegration [43]. Furthermore, for load-bearing applications, the key focus is on maximizing osseointegration to minimize the risk of aseptic loosening as around 20% of orthopedic revisions within two years are associated with loosening. The market of these implants can be broadly divided into mass manufactured and patient-specific implants. Mass-manufactured implants and their mid-term outcomes are currently positive, however their long-term outcomes are still needed to illustrate additional clinical benefits. Patient-specific or "custom" implants are used where no alternative implant exists, and its use is justified through the last resort where the benefits outweigh the risks [43].
Studies on the effect of regular, irregular, fully random, and several types of unit cell structures on mechanical properties of AM manufactured Ti64 implants are available [44][45][46]. The results from these studies show that specimens with higher porosity tend to buckle. Strength and Young's modulus were found to be decreased with increasing porosity and pore size. Hanks et al. [46] developed a graphical user interface to provide Ashby-style plots for unit cell selection of the AM lattice structure that helps to understand the relationship of the unit cell topology and the lattice structure's mechanical properties, with the intent of guiding appropriate unit cell selection. Furthermore, research on the modification of porous structure unit cells, such as diamond-shaped (tetrahedral) lattice cells, for better integrity of mechanical properties can also be found [47]. Similarly, the topological design of lattice structures, which includes beam-based unit cells, sheet-based unit cells, and functional gradients, can be rationally designed to achieve the desired properties including mechanical, fatigue, mass transport, (e.g., permeability, diffusivity), surface area, and geometrical features, while the rate of tissue regeneration, e.g., surface curvature, are simultaneously optimized [48]. Furthermore, changing the type of unit cell in the porous structure can result in very different mechanical properties for the same porosity level [49]. Comparing the normalized, i.e., the ratio of a property in the porous structure to that of the bulk material, mechanical properties of porous non-gradient Co-Cr structures with mechanical properties of non-gradient Ti64 and pure Ti with the same unit cell designs, Hedayati et al. [50] found that topological design can cause up to a ten-fold difference in mechanical properties of AM porous structure, while up to two-fold difference when changing the material type. This indicates the usefulness of bulk material properties as scale factors to convert normalized mechanical properties of porous non-gradient structures to their absolute values.
In recognition of the benefits of utilizing AM for manufacturing orthopedic implants, review articles on AM used for metallic implants [41,43,51,52], porous metallic implants [48,49], biodegradable porous metallic implants [15], and common AM implants [26] are also available. Among the metallic implants, a systematic review on titanium and its alloys can be found [27,37,53,54]. However, review articles with focus areas only on the AM manufactured Ti64 are rarely available. Besides, the review on the fatigue behavior of the porous implants appears to be barely discussed in most of the orthopedic implants review articles. Ti64 being the most preferred titanium alloy for orthopedic applications and porous structures design is the prime focus of many recent orthopedic implants/scaffolds' studies. An article by Aufa et al. [38] focused on the Ti64 biomedical implants manufactured using only a laser powder bed fusion-selective laser melting (SLM) process. The mechanical properties and the effects of surface treatment and surface modification of SLM Ti64 implants were reviewed, however the review on the fatigue behavior of these porous implants is completely missing. Even within the L-PBF process, selective laser sintering (SLS) is also used for generating porous Ti64 implants. In addition to L-PBF, studies on porous Ti64 implants manufactured using electron beam melting, direct write process and Laser Engineered Net Shaping (LENS TM ) are widely available. Therefore, there exists a significant gap in the literature availability on porous Ti64 implants, which includes all the AM processes that are used for biomedical applications. On the other hand, Spece et al. [40] performed a systematic review on preclinical in vivo testing of 3D printed Ti64 porous orthopedic implants. Although, this review provides decades of experience and lessons learned with AM porous Ti64 implants, many novel research works have been performed on AM porous Ti64 where no in vivo testing was carried out and thus there is a need for a review that includes those in vitro, porous designs, mechanical, fatigue, multifunctionality properties, etc., of AM porous Ti64 implants/scaffolds.
With the aim to fulfill the review literature gap, this paper aims to provide in-depth knowledge and information on recent advancements of porous Ti64 implants from different perspectives: design, mechanical properties, fatigue behavior, and multifunctionality. All the processes utilized in the additive manufacturing of the porous Ti64 bone implants are reviewed and are presented in a detailed fashion to provide proper guidance for future research in the design and development of functionally porous Ti64 implants/scaffolds. The orthopedic implant is regularly subjected to loading condition and the porous implant design with its porous structure is more prone to fatigue failure as compared to its counterpart. Thus, this paper has dedicated a sub-section for the fatigue behavior of the AM porous Ti64 structures such that the existing review gaps on their fatigue behavior are addressed. Moreover, the novelty of this review comes from the holistic overview of different AM processes used in the design and development of porous Ti64 implants/scaffolds from design, process, physical, mechanical, fatigue, biological and multifunctionality perspectives. In this article, porous implant designs are classified into non-gradient and gradient porous structures. Due to its greater resemblance to natural bone morphology and improved mechanical properties, gradient porous design appears more promising for orthopedic applications. Furthermore, the human bone properties and structure, mechanical properties of porous Ti64, as well as multifunctional porous implant designs, their limitations, current challenges, and unresolved literature gaps are thoroughly discussed later in the paper.

Methodology
To find the literature related to this topic, keywords such as "additive manufacturing of porous Ti6Al4V bone implants", "porous bone implants", "fatigue behavior of porous Ti6Al4V bone implants", "leaching of titanium implants" etc., were used in online searching tools including web of science, google scholar, MDPI and ScienceDirect websites. The papers were always listed based on the descending order from the recent published date. For each section/sub-section, significant papers were selected and reviewed. More focus was given to recently published papers, i.e., after 2015. Key findings from these reviewed articles were analyzed and compiled to write this paper.

Bone Structure
Bones are categorized into two major groups, namely cortical bones, and cancellous bones. Cortical bones are compact bones and are responsible for providing mechanical strength, structural rigidity, and movement. They account for 80% of the mass of the bones in the human body. Cancellous bones, also known as trabecular bones, are soft, spongy bones and are responsible for providing structural support to the cortical bones, flexibility, and reduction in weight. Most of the active functions including blood cell production and ion exchange take place in the trabecular bones. They account for roughly 20% of the total mass of the skeleton. From the porous structure perspective, cancellous bone has porosities between 50% and 90%, whereas cortical bone has a porosity less than 10% [41].
The natural bone acts as a heterogeneous and anisotropic nanocomposite [55], of which the principal components are organized hierarchically into several structural levels, from macro to the nanometer scale as shown in Figure 1. The cortical bone thickness and the density and pore structure of the trabecular bone may vary significantly by location in the body and even within one bone, depending on the local requirements. Near the joints the trabecular bone is denser due to varying mechanical load requirements (strength and angles of loading), while in the central part of the bone there is less dense and less isotropic trabecular bone, which is more di- Macroscale arrangements of the bone involve both compact/cortical bone at the surface and spongy/trabecular bone (foam-like material with~100 µm thick struts) in the interior. Compact bone is composed of Osteons and Haversian canals, which surround blood vessels. Osteons have a lamellar structure, with individual lamella consisting of fibers arranged in geometrical patterns. The fibers comprise several mineralized collagen fibrils, composed of collagen protein molecules (tropocollagen) formed from three chains of amino acids and nanocrystals of hydroxyapatite (HA) and linked by organic phase to form fibril arrays [55].
The cortical bone thickness and the density and pore structure of the trabecular bone may vary significantly by location in the body and even within one bone, depending on the local requirements. Near the joints the trabecular bone is denser due to varying mechanical load requirements (strength and angles of loading), while in the central part of the bone there is less dense and less isotropic trabecular bone, which is more directionally aligned as shown in Figure 2a. The cortical bone thickness and the density and pore structure of the trabecular bone may vary significantly by location in the body and even within one bone, depending on the local requirements. Near the joints the trabecular bone is denser due to varying mechanical load requirements (strength and angles of loading), while in the central part of the bone there is less dense and less isotropic trabecular bone, which is more directionally aligned as shown in Figure 2a. Typical stress-strain curves of compact (Cortical) and spongy (Trabecular) bones. Reprinted with permission from ref. [57]. 2021, MDPI. Figure 2b shows a load-deformation curve for a specimen of hydrated bone loaded in tension [58]. The bone follows the normal load-deformation trend for that of metals. First, there is a rising part where the load is proportional to deformation. At the yield point, the situation changes, and the curve becomes much flatter. With a slight increase Figure 2. (a) Variation in the trabecular bone structure by location, in this example from the human femur of a 26-year-old male. Reprinted with permission from ref. [56]. 2019, Elsevier Inc. (b) Typical stress-strain curves of compact (Cortical) and spongy (Trabecular) bones. Reprinted with permission from ref. [57]. 2021, MDPI. Figure 2b shows a load-deformation curve for a specimen of hydrated bone loaded in tension [58]. The bone follows the normal load-deformation trend for that of metals. First, there is a rising part where the load is proportional to deformation. At the yield point, the situation changes, and the curve becomes much flatter. With a slight increase in load, there is a substantial increase in deformation. Then the specimen breaks. This load-deformation curve can be normalized and turned into a stress-strain curve. The total area under the curve is the amount of energy per unit volume that a material can absorb before rupturing, also known as the material toughness.

Mechanical Properties
Cortical Bone: The Modulus of Elasticity of cortical bone is about 15 GPa. At the yield point, the stress is 120 MPa, and the strain is 0.008. Table 1 gives some representative values of the mechanical properties of cortical bone loaded along the length of the bone. This is an approximation table, but gives some guidance as to the properties and their potential range. Cancellous Bone: Cancellous bone is soft bone. The bone does not form closed cells, so the bone, and the marrow, are always interconnected. The Young's modulus of such bone is as low as 1.9 GPa and can go up to 6 GPa [58].

Bone Healing Mechanism
Most of the fractured bone goes through the same healing process or recovery mechanism. This holds true for the bone whether it has been operated on during the surgical procedure or fractured by being involved in an accident. Usually, the bone healing mechanism experiences an overlap of three stages, namely inflammation, bone production, and bone remodeling [59], as described below: a. Inflammation is the first step of the healing mechanism, and this starts right after the bone breaks. After the bone fractures, the blood vessels' connections in the bone are also ruptured and this results in bleeding into the fractured area. This process is called inflammation where the blood starts to clot in the fractured site and swelling is observed. Inflammation lasts for several days and thus provides the initial framework and stability to the fractured bone for developing new bone; b. The bone production step follows the inflammation step in the healing process. The clotted blood from inflammation is gradually replaced by the growth of soft callus cartilage and fibrous tissues. Eventually, when healing progresses, this soft callus cartilage is replaced with hard bone known as a hard callus. This bone is thus detectable by x-rays afterward; c. Bone remodeling is the final phase of the healing process and could last for weeks or several months. The hard callus bone continues to form and becomes compact, returns to its original shape and form, ruptured blood vessels grow, and blood circulation improves in the area. Sufficient healing of the bone after several weeks gradually remodels the bone for usual weight-bearing activities such as standing and walking, and it provides necessary support to the structure [60]. More importantly, the loading magnitude and loading frequency have significant effects on bone remodeling and thus have a major impact on the bone healing process [41].

Hindrance for Bone Healing
A wide variety of factors can slow down the healing process of a fractured bone. Movement of fractured bone fragments and weight-bearing too soon delays bone formation. The implant must be securely fixed to the host bone such that osseointegration is not inhibited. The interactions between the implant and the proteinous monolayer create the basis for osseointegration and ossification (bone formation) and are crucial to the proper healing of the wound, and ultimately, successful implantation [60,61]. In addition, smoking, medical conditions, such as diabetes, hormone-related problems or vascular disease, some medications, such as corticosteroids and other immunosuppressants, severe and complicated fractures, old age, poor nutrition or impaired metabolism, and low levels of calcium and vitamin D all hinder bone healing.

Need for Porous Structure and Its Advantages
Implants are made from biomaterials and their alloys. These materials, compared to bone, have a very high modulus of elasticity. It is well known that the stress transfer between an implant and a bone is not homogeneous when Young's moduli of the implant and the bone are different; this is defined as stress shielding. In such conditions, bone atrophy occurs and leads to the loosening of the implant and refracturing of the bone. Therefore, it is desirable if the stiffness (Young's modulus) value is as close as possible to that of the bone [62].
Porosities in the bone implants have two (2) major advantages: a. Provide space and 3D structure for bone ingrowth. b. Mimic the mechanical properties of real bone [48,63].
Based on the literature reviewed in this study, the porosity ranging from 25% to 90% and pore size ranging from 200 µm to 1500 µm were found to be mostly used in the design of porous Ti64 implants. During the porous implant design, often there exists a trade-off between the porosity percentage and mechanical strength. An increase in mechanical interlock strength of a porous implant can be observed through osseointegration, however a design with porosity greater than 80% results in a decrease in both strength and bone ingrowth [64]. Proper tailoring of porosity percentage, porosity gradient and its direction, and transition region height are required such that the biomimetic implant design can be possible without compromising the mechanical properties requirements, especially porous implant strength.
The premise of bone ingrowth is that cells adhere to the surface of materials and migrate to the inner space, thus porosities play a vital role in bone tissue engineering as it facilitates the ingrowth of mineralized tissue into a porous network. It is also suitable for seeding cells, delivering drugs, capillary tissue action, and osteoprogenitor cell migration [48,65].
In this study, we have focused on Titanium alloy, namely Ti64. This is an alphabeta titanium alloy, which has excellent corrosion resistance to chlorides, seawater, and sour and oxidizing acid media, and acts as an ideal candidate for orthopedic and dental implants [66]. The biocompatibility with living tissues and the high strength-to-weight ratio of Ti64 alloy make it an ideal candidate for bone implants. Furthermore, Ti64 shows good osseointegration property, which is defined as the functional bone adherence property where a new bone is laid down directly on the implant surface and helps the implant to exhibit mechanical stability [67,68]. More importantly, Ti64 alloy also has a low modulus of elasticity, which is roughly half that of steel and nickel alloys, making it more similar to bone in terms of mechanical properties [66]. Further, the bioinert and biocompatible nature of Ti64 with spontaneous formation of oxide layer ensures its stability in the body after implantation [69].
Titanium is generally non-magnetic in nature, generating insignificant forces in a strong magnetic field, and is considered safe to be present in clinical MRI systems. However, it has weak paramagnetic property and becomes magnetized within the MRI field resulting in either a loss of signal or distortion in the MRI image (known as susceptibility effect) due to changes in magnetic field from the presence of Ti64 implant [28,70]. This effect limits the use of MRI to visualize the tissue surrounding implants hindering future investigations in the region or diagnosis of post-operative complications. AM porous structures may be an upstream method for reducing susceptibility artifacts in Ti64 implants, helping in improving post-operative diagnosis capability [28].

Porosity Designs
Porous implant structures involve porosity, pore size, etc., as major morphological design parameters [71]. The level of these factors has a significant impact on the mechanical  Tables 2-4. Bone implant porosity and stiffness/strength require optimized trade-off to improve long-term load sharing while simultaneously promoting osseointegration [64]. Porous Ti64 implants have been developed with a biomimetic design attempt towards natural bone morphology. The porous Ti64 implants can be categorized into non-gradient and gradient porous implants. Non-gradient implant design has uniform porosity throughout the implant structure, whereas the gradient design may have porosity variation along x, y, and/or z-direction. Details on these porosity design types are explained later in this section. During the literature review, most of the studies on porous design were identified to be performed on samples fabricated by Laser Powder Bed Fusion (L-PBF) and Electron Beam Melting (EBM). Furthermore, some pieces of literature were found on porous Ti64 alloy implants fabricated using direct ink writing (DIW) and Laser Engineered Net Shaping (LENS TM ) processes. In the powder bed fusion (PBF) process, the parts are fabricated by heating powder particles at selective areas until they are sintered or melted. Based on the source of energy utilized, PBF can be categorized into laser powder bed fusion or electron beam melting. In the DIW process, two-dimensional or three-dimensional structures are created directly onto the flat or conformal surfaces through jetting or dispensing, or extruding of material such as the micro-dispensing of ink. LENS TM is an example of directed energy deposition (DED) AM process. The DED process is mostly used for part repair where the material is deposited mostly on the existing substrate to generate new surfaces by using powder feedstock or wire feedstock. LENS TM , originally developed by Sandia National Labs in 1996, uses Nd: YAG/fiber laser to melt co-axially delivered powder feedstock and deposits the material into the substrate.
The evaluation of L-PBF and EBM AM process conditions are most commonly performed using volumetric energy density, defined as [72]: where P is the laser/electron beam power, v is the scan speed, h is the hatch distance, and d is the layer thickness. The PBF process takes place inside an enclosed chamber filled with inert gas to minimize degradation and oxidation of powered material [73]. First, a layer typically 0.1 mm thick of powder material is spread over the build platform. The powder is then preheated and maintained at an elevated temperature below the melting point or glass transition temperature of the powdered material. A focused heat source (laser or electron beam) is then directed onto the powder bed and is moved using galvanometers in a way that it thermally fuses the material to form the slice cross-section completing the first layer. In this process, the energy beam is scanned over the powder surface (scan speed) often in a hatched stripe pattern (hatch distance) using controlled mirrors until a layer is processed. After completing this layer, the build platform is lowered by one layer thickness, and a new layer of powder is spread over the build platform using a recoater blade mechanism. The heat source again scans the subsequent slice cross-section. This process repeats until the complete part is built. Finally, the part is removed from the powder bed, loose powders are cleaned off the part, and further finishing operations, if necessary, are performed [73]. The schematic of the PBF process is shown in Figure 3. A study of the effect of laser processing parameters showed that mechanical properties increased with increases in either laser power or exposure time [44]. Furthermore, the surface topography and roughness of implants manufactured using L-PBF and EBM are affected by build orientation [74,75]. The lowering of the build angle lowers surface roughness and hydrophobicity, and the number of partially melted particles [75]. This reduction is reported to be directly correlated with significantly lower biofilm coverage for a lower build angle without compromising cell viability and attachment. Since the porous AM implants (both gradient and non-gradient) have struts and walls oriented at different angles, they may require careful consideration during the design for biofilm restriction. A review article by Aufa et al. [38] states that a rough surface (R a values) of 3 µm-5 µm is favorable for osteoblasts attachment both in vivo and in vitro, and an implant roughness between the 1 µm and 2 µm range is effective for developing a titanium dioxide (TiO 2 ) layer that is resistant to bacterial infection.
Furthermore, the surface topography and roughness of implants manufactured using L-PBF and EBM are affected by build orientation [74,75]. The lowering of the build angle lowers surface roughness and hydrophobicity, and the number of partially melted particles [75]. This reduction is reported to be directly correlated with significantly lower biofilm coverage for a lower build angle without compromising cell viability and attachment. Since the porous AM implants (both gradient and non-gradient) have struts and walls oriented at different angles, they may require careful consideration during the design for biofilm restriction. A review article by Aufa et al. [38] states that a rough surface (Ra values) of 3 µm-5 µm is favorable for osteoblasts attachment both in vivo and in vitro, and an implant roughness between the 1 µm and 2 µm range is effective for developing a titanium dioxide (TiO2) layer that is resistant to bacterial infection.

Non-Gradient (Uniform) Porous Structure
One of the major design parameters of bone implants is porosity. Non-gradient or uniform porosity is one of the porosity design parameters where the size of pores does not change throughout the implant. This makes the implant density uniform along x, y,

Non-Gradient (Uniform) Porous Structure
One of the major design parameters of bone implants is porosity. Non-gradient or uniform porosity is one of the porosity design parameters where the size of pores does not change throughout the implant. This makes the implant density uniform along x, y, and/or z-direction. Figures 4 and 5 represent some of the non-gradient porous structures found in the literature [77][78][79][80][81]. The design optimization study on uniform Ti lattice structure by El-Sayed et al. [82] identified the most significant design parameter to be strut length where the porosity of unit cell increased with increasing length, but the effect reverses on strength and elasticity. Goto et al. [83] studied two different designed implants: one with a smooth surface without any porosity and the other with uniform pores of 800 µm and 1400 µm. The characterization of these EBM fabricated porous implants showed average porosity of 57.5% from binary images of 30 slices. Balci et al. [84] obtained bone samples from the femoral and vertebral regions of a sheep, scanned them using Micro-CT and produced trabecular bone models with two different porosities with the purpose of increasing biocompatibility by imitating bone. The vertebra model had a distribution between 102 µm and 1719 µm, and the femur model had a distribution between 90 µm and 1434 µm. Ødegaard et al. [77] manufactured scaffolds with 800 µm void spacing and lattice thickness based on the Schwarz P triply periodic minimal surface-based unit cell and had a designed porosity of 50%. Biomimetic porous Ti bone scaffolds with a high level of interconnectivity between pores are beneficial for tissue growth and osseointegration [85]. Tsai et al. [86] used a highly interconnected pore network structure with a high-volume porosity of 60% for better bone growth. Fluid permeability is a function of pore characteristics (i.e., pore shape, size, and porosity) and shows a strong correlation with bone ingrowth and thereby served as an indication of scaffold osteointegration capabilities [87,88]. Accordingly, Arjunan et al. [88] investigated the uniform porous L-PBF scaffold designs with targeted porosity of 68.46%-90.69% considering permeability, stiffness, strength, and stress concentration factor. The results showed the primary dependency of maximum stress on pore shape rather than relative density. Required stiffness and strength of bone scaffold can be achieved by tailoring the stress concentration factor, irrespective of the targeted porosity. Therefore, in addition to porosity, pore shape and permeability, the stress concentration should also be carefully considered during the design of functional porous scaffold.
The porous Ti64 cage was produced by the L-PBF (Selective Laser Melting, SLM) process. Chen et al. [78] performed a study on three different shapes of Ti64 samples with a porosity of 70% and a pore size of 600 µm as shown in Figure 5. In the study by Hameed et al. [89], the decrease in mechanical strength with an increase in pore size was identified and a scaffold with a pore size of 250 µm showed the optimum combination of overall mechanical and biological properties for practical applications. Zhang et al. [90] studied the spacing of the laser scan lines (hatch spacing) on pore characteristics and set the scan line spacing from 200 µm to 700 µm by a step size of 100 µm to produce the porous implants. Ahmadi et al. [67] investigated the effects of two different heat treatment regimens (below and above β-transus) on the mechanical properties and microstructure of porous Ti64 specimens. Many similar studies on the uniform gradient porous Ti64 can be found in the literature. Among them, some of the significant articles are listed in Table 2.      In vivo study on Goats for 3, 6, and 12 months identified that depth of bone ingrowth increased over time and no implant dislocation was observed; Porous Ti64 scaffold with an intended pore size of 300 to 400 µm are most suitable for cell adhesion and proliferation.
Melancon et al. [71] L-PBF Tetrahedron and Octet -truss unit cells were studied over the entire design space, which includes porosity, strut thickness, unit cell size, pore size, height, width, and depth.
Depend on the unit cell design parameter Depend on the unit cell design parameter Proposed a systematic approach integrating computed tomography, mechanical testing, and statistical analysis of geometric imperfections to generate statistical-based numerical models of high strength additively manufactured porous biomaterials; The proposed methodology reduced the error between predicted and tested mechanical properties from 49% to 11% (elastic modulus) and 41% to 7.6% (yield strength) Thin-layer CT and 3D reconstruction bone models are effective in evaluating periprosthetic bone loss preoperatively; Effective reconstruction of periprosthetic bone defect and restoration of knee prostheses; Stable Protheses even after 3 years.

Gradient (Non-Uniform) Porous Structure
Although non-gradient porous implant structure tends to mimic the natural bone in terms of mechanical properties, the non-gradient designs do not structurally mimic the natural bone morphology. Looking at an example of the human tibia bone, it consists of an outer cortical bone with relatively higher density and inner cancellous bone with lower density. This density variation implies higher mechanical strength at the outer cortical when compared to inner cancellous bone [96]. The variation in bone structure is not only in the radial direction but also along the longitudinal direction. Serious bone density variation in bones such as mandible and vertebrae require the implant to meet these bones' Young's modulus to prevent stress shielding. Thus, in vivo implantation of non-gradient porous structures can still be susceptible to serious stress shielding effects [97]. Therefore, to account for these structural differences and inhomogeneity of the natural bone, gradient porous structure designs can be achieved through average porosity % level variation along the longitudinal direction (longitudinal gradient porous structure) and/or the radial direction (radial gradient porous structure) [96,97], as shown in Figure 6. Spatial variation in porosity of gradient lattice structures changes the stiffness and makes it comparable to cortical or trabecular bone, however, the relationship between the mechanical properties and biological properties of the gradient structure is not quite clear [98]. Compared to longitudinal gradient structures, radial-gradient structures appear to have a higher mechanical strength as the outer region with lower porosity protects the inner region to avoid destructive localized deformation [96,97]. Furthermore, even within the radial-gradient design, different unit cell structures (Figure 6(c) and Figure 7) [99], region thickness, and region volume fraction [96,97] can significantly impact the mechanical properties of the gradient porous implant. For radial design, interestingly, mechanical properties of gradient structures with the same thickness regions increase with an increase in the average porosity percentage level variation, whereas mechanical properties remain approximately the same for gradient porous structures with the same volume regions. Wu et al. [97] showed that the gradient porous structure with transition region is superior to those with a sharp interface in terms of mechanical properties such as Young's modulus, yield strength, and maximum compressive strength. Furthermore, the failure of samples with transition region occurs towards the higher porosity region as opposed to the interface in sharp interface samples [96]. Another study conducted by Wu et al. [96] found that Young's modulus, Yield Stress, maximum compressive strength, and strain at failure values of radial gradient samples were within the individ- Compared to longitudinal gradient structures, radial-gradient structures appear to have a higher mechanical strength as the outer region with lower porosity protects the inner region to avoid destructive localized deformation [96,97]. Furthermore, even within the radial-gradient design, different unit cell structures (Figures 6c and 7) [99], region thickness, and region volume fraction [96,97] can significantly impact the mechanical properties of the gradient porous implant. For radial design, interestingly, mechanical properties of gradient structures with the same thickness regions increase with an increase in the average porosity percentage level variation, whereas mechanical properties remain approximately the same for gradient porous structures with the same volume regions. Wu et al. [97] showed that the gradient porous structure with transition region is superior to those with a sharp interface in terms of mechanical properties such as Young's modulus, yield strength, and maximum compressive strength. Furthermore, the failure of samples with transition region occurs towards the higher porosity region as opposed to the interface in sharp interface samples [96]. Another study conducted by Wu et al. [96] found that Young's modulus, Yield Stress, maximum compressive strength, and strain at failure values of radial gradient samples were within the individual single porosity non-gradient samples values, however, the elastic limit was significantly lowered as a downside. Studies with both radial and longitudinal designs combined into a single porous structure will be an interesting topic of study as these designs will better mimic the natural bone. Additionally, the topological design of functionally graded porous structures exhibits superior comprehensive mechanical properties to uniform porous structures by a combination of low density, moderate Young's modulus, high yield stress, high maximum stress, and favorable ductility [100]. Studies in bone tissue engineering indicated that a range of mean pore sizes (96-150 µm) facilitate optimal attachment, whereas large pores (300-800 µm) are necessary for successful bone growth in scaffolds [101,102]. On the other hand, an in vitro study by Van Bael et al. [103] showed that pore occlusion occurs on porous structures with the hexagonal pores of pore size 500 µm and such pore occlusion should be avoided in the design of porous implants. The gradient design can take advantage of high porosity for fluid permeability without compromising the mechanical strength, implying the superiority of gradient porous implants, compared to uniform porosity implants.
In vivo studies are desirable for the performance analysis of implants; however, these in vivo studies are time-consuming and expensive. One of the approaches for the biological and mechanical performance evaluation is Finite Element Analysis (FEA) simulation. Vee San et al. [104] used FEA to simulate bone ingrowth and verified FEA results against histology results for an ovine condylar critical-sized defect model. Furthermore, the FEA analysis performed on longitudinal gradient implant predicted that higher bone ingrowth can be expected on implants made from lower Young's modulus, such as Titanium-tantalum alloy. With the lower Young's modulus of these implants, their performance and safety may be improved. Another study by Davoodi et al. [99] compared deformation and failure mechanisms of hip implants under compression with experimental results. Both FE simulation and experimental results exhibited similar failure mechanisms with failure more sensitive towards the unit cell shape. These studies indicate the feasibility of FEA analysis in identifying the stress-concentration, implant failure mechanism, and expected bone ingrowth if implanted in vivo. Some of the significant research articles on gradient Ti64 implants identified during the literature review are listed in Table 3. Studies in bone tissue engineering indicated that a range of mean pore sizes (96-150 µm) facilitate optimal attachment, whereas large pores (300-800 µm) are necessary for successful bone growth in scaffolds [101,102]. On the other hand, an in vitro study by Van Bael et al. [103] showed that pore occlusion occurs on porous structures with the hexagonal pores of pore size 500 µm and such pore occlusion should be avoided in the design of porous implants. The gradient design can take advantage of high porosity for fluid permeability without compromising the mechanical strength, implying the superiority of gradient porous implants, compared to uniform porosity implants.
In vivo studies are desirable for the performance analysis of implants; however, these in vivo studies are time-consuming and expensive. One of the approaches for the biological and mechanical performance evaluation is Finite Element Analysis (FEA) simulation. Vee San et al. [104] used FEA to simulate bone ingrowth and verified FEA results against histology results for an ovine condylar critical-sized defect model. Furthermore, the FEA analysis performed on longitudinal gradient implant predicted that higher bone ingrowth can be expected on implants made from lower Young's modulus, such as Titanium-tantalum alloy. With the lower Young's modulus of these implants, their performance and safety may be improved. Another study by Davoodi et al. [99] compared deformation and failure mechanisms of hip implants under compression with experimental results. Both FE simulation and experimental results exhibited similar failure mechanisms with failure more sensitive towards the unit cell shape. These studies indicate the feasibility of FEA analysis in identifying the stress-concentration, implant failure mechanism, and expected bone ingrowth if implanted in vivo. Some of the significant research articles on gradient Ti64 implants identified during the literature review are listed in Table 3. Optimal gradient structures should have small pores in their core (~900 µm) to increase mechanical strength and large pores (~1100 µm) in outer surface to enhance cell penetration and proliferation while avoiding pore occlusion; Static mechanical properties of gradient structure follow the rule of mixtures and are in the range of their uniform structures.   High improvement of fluid permeability (about 140% for D unit cell structure and 277% for P unit cell structure);

Fatigue Behavior of Porous Ti64 Structures
The bone implants in the human body bear cyclic loading while walking, running, and performing other physical activities over a long period of time [102,106]. The porous Ti64 bone implants have a structural function and are expected to have a lifespan of several decades [107]. Therefore, a fundamental understanding of the fatigue behavior of porous Ti64 implants plays an important role in the design and implementation of porous Ti64 implants for orthopedic applications [102,[106][107][108]. Similar to the features controlling conventionally processed products, the fatigue performance of AM Ti64 implants is governed mostly by surface finish/roughness, residual stress, internal manufacturing defects, microstructure, and loading conditions [109]. Several papers [107,[110][111][112] listed in Table 4 show that the normalized fatigue strength in their study is lower than the fatigue strength of the solid material of the same microstructure for porous Ti64 structures. This drop in the fatigue strength is mainly associated with the inherent defects coming from the fabrication process itself, namely, internal defects and poor surface roughness. Furthermore, pore design/geometry and strut thickness were also found to have an influence on the fatigue strength of these porous Ti64 samples [63,110].

Fatigue Behavior of Porous Ti64 Structures
The bone implants in the human body bear cyclic loading while walking, running, and performing other physical activities over a long period of time [102,106]. The porous Ti64 bone implants have a structural function and are expected to have a lifespan of several decades [107]. Therefore, a fundamental understanding of the fatigue behavior of porous Ti64 implants plays an important role in the design and implementation of porous Ti64 implants for orthopedic applications [102,[106][107][108]. Similar to the features controlling conventionally processed products, the fatigue performance of AM Ti64 implants is governed mostly by surface finish/roughness, residual stress, internal manufacturing defects, microstructure, and loading conditions [109]. Several papers [107,[110][111][112] listed in Table 4 show that the normalized fatigue strength in their study is lower than the fatigue strength of the solid material of the same microstructure for porous Ti64 structures. This drop in the fatigue strength is mainly associated with the inherent defects coming from the fabrication process itself, namely, internal defects and poor surface roughness. Furthermore, pore design/geometry and strut thickness were also found to have an influence on the fatigue strength of these porous Ti64 samples [63,110].  Fatigue endurance ratio at 10 6 cycles for L-PBF specimen decreased from 0.5 to 0.15 with an increase in porosity from 33 vol.% to 84 vol.%. However, for the HIPed samples with different porosities, the fatigue endurance ratio was higher than 0.45 at 10 6 cycles Mechanical properties from static testing and fatigue behavior of 45 • build orientation struts were better than that of 90 • vertical struts; Abrupt decrease in the fatigue curve was observed in low cycles, obtaining the fatigue strength for 10 6 cycles between 15% and 25% of the flexural strength Günther et al. [118] L-PBF Figure 8 Resonant testing machine with symmetrical push-pull loading at R = −1 and 105 Hz Razavi et al. [108] developed non-porous and induced microporous samples with internal voids and defects across each layer of parts using LENS technology. When comparing non-porous and porous specimens under uniaxial tensile loading fatigue test with a stress ratio of R = 0.01 (stress or load ratio (R) is the ratio between the minimum compressive fatigue stress (σ min ) and the maximum fatigue stress (σ max )), porous samples showed 75% reduction in stress range at one million cycles. This implies porosity to be the main contributor to a significant reduction in fatigue life. Studies by Li et al. [102,119] used vacuum diffusion bonding of alloy meshed to develop a non-gradient porous Ti64 structure with a designed porosity of 50%, 60%, and 70% and found the normalized fatigue strength at 10 6 cycles to be in the range of 0.50-0.60 under compression-compression fatigue test at a load ratio of 0.1. The normalized fatigue strength was higher than the normalized fatigue strength of other porous Ti64 alloys developed using powder sintering and L-PBF.
Furthermore, the fabrication and post-processing conditions also highly influence the Ti64 porous implant material, its mechanical properties, and even its microstructure. In L-PBF, inherent heating and cooling speeds induce directionally dependent grain growth, internal porosity, and anisotropic microstructure [63]. Typically, additional post-processing heat treatment such as stress-relieving or hot isostatic pressing (HIP) are utilized to reduce internal porosity, to control the microstructure such that it improves mechanical properties, particularly ductility [63]. With HIP treatment, Wu et al. [114] reported that the strut porosity of L-PBF specimens were reduced to less than 0.001 vol.%. Regarding the microstructure, the α'martensite phase, which is a common phase on the L-PBF specimens, can be transformed into the tougher α + β mixture phases by using the HIP process [106,114]. This phase transformation helps in resisting the propagation of fatigue cracks by crack blunting and thereby improving fatigue performance [106,114,120]. Furthermore, the fatigue life of porous Ti64 structures can also be improved using a sandblasting process as it induces compressive stress on the strut's surface and removes partially melted particles [121].
In many load-bearing applications, compression is the primary loading mechanism, however, implants can sometimes experience tensile loads due to the stochastic nature of forces on the human body [63]. Pure compression loading does not usually result in fatigue failure in most continuous materials as compressive loads may suppress crack growth and prevent fatigue failure from happening [49]. When compressive fatigue loads are applied to porous structures at a macro-scale, tensile stress may be developed in the individual struts. Depending on the porous design, some unit cells can result in much more tensile stress than others [49]. However, a few literature studies focused on fatigue behavior at tensile loading are available. The detailed fatigue behavior of porous Ti64 implants under different testing conditions is presented in Table 4. (a) (b) Figure 9. (a) S-N curves of batches with different internal channels as shown in Figure 8, partly recompiled from their previous studies [122]. (b) S-N curves with different batches compared to the results by Leuders et al. [123] and Kasperovich et al. [124] designated as Reference 1 and 2, respectively. Reprinted with permission from ref. [118]. 2018, Elsevier Ltd.

Multifunctional Implant Design and Development
Innovative multifunctional porous implant design and development can be a possible approach towards better implant life and a reduced number of revision surgeries. As indicated in Figure 10, with additive manufacturing and tissue engineering, porous titanium implants can be custom fabricated to match patients' anatomical structure with desirable porous geometry and multifunctionality for osseointegration. Poor implant integration with native bone at the implant interface is one of the major causes of implant (a) (b) Figure 9. (a) S-N curves of batches with different internal channels as shown in Figure 8, partly recompiled from their previous studies [122]. (b) S-N curves with different batches compared to the results by Leuders et al. [123] and Kasperovich et al. [124] designated as Reference 1 and 2, respectively. Reprinted with permission from ref. [118]. 2018, Elsevier Ltd.

Multifunctional Implant Design and Development
Innovative multifunctional porous implant design and development can be a possible approach towards better implant life and a reduced number of revision surgeries. As indicated in Figure 10, with additive manufacturing and tissue engineering, porous titanium implants can be custom fabricated to match patients' anatomical structure with desirable porous geometry and multifunctionality for osseointegration. Poor implant in-  Figure 8, partly recompiled from their previous studies [122]. (b) S-N curves with different batches compared to the results by Leuders et al. [123] and Kasperovich et al. [124] designated as References [1,2], respectively. Reprinted with permission from ref. [118]. 2018, Elsevier Ltd.

Multifunctional Implant Design and Development
Innovative multifunctional porous implant design and development can be a possible approach towards better implant life and a reduced number of revision surgeries. As indicated in Figure 10, with additive manufacturing and tissue engineering, porous titanium implants can be custom fabricated to match patients' anatomical structure with desirable porous geometry and multifunctionality for osseointegration. Poor implant integration with native bone at the implant interface is one of the major causes of implant failure [99]. Table 5 shows that depending on the objectives of studies, multifunctional porous implants have been developed for the improvement of mechanical properties, bonding on the boneimplant interface, bone volume fraction, bioactivity, osteointegration, antibacterial implant design, etc.
Metals 2022, 12, x FOR PEER REVIEW 25 of 36 Figure 10. Schematic of an iterative optimization process for hybrid additively manufacture (AM) patient-specific Ti implant fabrication including (A) critical-size bone defect data collection by CT scanning followed by (B) modeling through rational design, which is (C) additive manufactured with precisely controlled architectures. After necessary surface modification, implants can be (D) bio-functionalized through cell/growth factor incorporation with the assistance of bio-fabrication strategies and subsequent transplantation to repair the defect or (E) incubated for in vitro maturation in a bioreactor and then transplanted, leading to (F) rapid bone and vasculature formation. Reprinted with permission from ref. [54]. 2020, Elsevier Ltd. Micro-nanoporous MAO layer stimulated cell adhesion and considerably increased the cell number and improved cell morphology while subsequent nano-silver immobilization, inspired by the mussel's adhesive versatility, yielded potential antibacterial surfaces Figure 10. Schematic of an iterative optimization process for hybrid additively manufacture (AM) patient-specific Ti implant fabrication including (A) critical-size bone defect data collection by CT scanning followed by (B) modeling through rational design, which is (C) additive manufactured with precisely controlled architectures. After necessary surface modification, implants can be (D) bio-functionalized through cell/growth factor incorporation with the assistance of bio-fabrication strategies and subsequent transplantation to repair the defect or (E) incubated for in vitro maturation in a bioreactor and then transplanted, leading to (F) rapid bone and vasculature formation. Reprinted with permission from ref. [54]. 2020, Elsevier Ltd.
Ti-alloys have poor tribological behavior owing to their high friction coefficient and can cause significant adhesive wear generating debris resulting in inflammation and bone resorption [53]. In addition, the performance of 'as-built' Ti-alloy implants by AM processes are limited due to their low surface quality, surface imperfections, and defects introduced due to layer-by-layer deposition processes [53]. To resolve such issues, surface treatment can be used to enhance the resistance to tribology or corrosion of the as-fabricated AM parts [53]. Surface treatment alters surface roughness and modifies implant topography to improve biological response before implantation [38,125]. Studies show that surface treatment can be performed by sandblasting (S), electropolishing, acid-etching (E), anodization oxidation (AO), anode spark oxidation (ASO), heat treatment, and laser treatment (laser polishing, laser shock peening, laser remelting) [38,53,125]. Surface modification technique adds material of various thicknesses superficially on the implant surface to reduce roughness, mask cracks, improve surface strength, and improve tribological and/or corrosion resistance. Surface modification can be conducted by adding silver nanoparticles (AgNPs), silver nitrate (AgNO 3 ), antibiotic mesoporous bioactive glass (MBG), calcium carbonate (CaCO 3 ), and hydroxyapatite [38,53,125].
Multifunctional porous implant designs are mostly found to be subjected to surface modification treatments for property enhancement. The majority of the surface modification studies focus on coating Hydroxyapatite (HA) or other calcium phosphate compounds such as Tri-calcium phosphate (TCP) for improving bioactivity, biocompatibility, and cellular activities, such as cell adhesion, proliferation, differentiation, osteointegration, etc. [38,53]. Chudinova et al. [126] performed electrophoretic deposition of near-spherical calcium phosphate nanoparticles (CaPNPs) on the surface of AM Ti64 scaffolds and found homogeneous deposition of CaPNPs on Ti64 scaffolds surface with an improvement of both cell adhesion and growth. Besides HA surface modification, Jia et al. [127] developed a novel porous Ti64 implant with antibacterial properties via Electron Beam Melting combined with micro-arc oxidation (MAO). The release of silver nanoparticles (AgNPs) after implantation within the body acts as an antibacterial agent, which prevents biofilm formation at the bone-implant interface, thus enhancing the implant success rate. Another study by Fazel et al. [39] developed a multifunctional layer on Ti64 implants, a hierarchically structured surface including an interconnected microporous TiO 2 layer containing silver nanoparticles and hydroxyapatite nanocrystals through plasma electrolytic oxidation (PEO) followed by hydrothermal treatment (HT), where they found that HT and spindle-like HA nanocrystals showed sustained antibacterial activity, a higher level of pre-osteoblasts metabolic activity, and a higher level of alkaline phosphate activity as compared to PEO-treated implants without HT. Hengel et al. [128] designed a novel functionality-packed implant using a surface biofunctionalization technique to prevent aseptic loosening through stimulation of the osteogenic differentiation of stem cells, as well as septic loosening through the short-and long-term delivery of antibacterial agents from the entire volume of porous structure (detailed multifunctionality design listed in Table 5). The multiple active agents were incorporated into the micro-and nano-topographical structure uniformly throughout the entire surface. In addition to the surface modification, Li et al. [129] mixed Ti64 powder with β-TCP to create interconnected porous structures and evaluated implants' in vivo tissue reaction and osteoinductive potential. Detailed information on the functionalization of additively manufactured Ti alloys for orthopedic implants can be found in the review paper by Jing et al. [27].
The use of Mg alloys in implants helps in reducing the relative elastic modulus, makes implant bioresorbable, and more importantly, helps in creating bioactive systems [15,130]. Perets et al. [130] designed biodegradable Titanium-Magnesium (Ti-Mg) non-gradient porous implants. Even though the utilization of bioresorbable material with Ti64 has high potential in future implant materials, detailed studies on the in vivo corrosion rate of these novel porous implants are required, otherwise, the multifunctional porous implants may end up losing their mechanical integrity before bone healing, demanding for revision surgery. Furthermore, the potential of the NiTi-Ti64 multi-material cellular structured hip implant has been studied by Bartolomeu et al. [131] with a broader objective of creating implants with customized stiffness, superior wear resistance, and a controlled NiTi outer region volume change. Multifunctional porous implant studies are mostly found to be focused on evaluating the performance (mechanical and biological) of as-prepared porous implants. Micro-nanoporous MAO layer stimulated cell adhesion and considerably increased the cell number and improved cell morphology while subsequent nano-silver immobilization, inspired by the mussel's adhesive versatility, yielded potential antibacterial surfaces Hengel et al. [128] L-PBF (Ti64-ELI) Miniaturized implant geometry optimized for murine femora implantation; Hexagonal unit cell, Total Length 4 cm and diameter 0.5 mm. A surface-to-volume ratio of 35.6 Biofunctionalized using plasma electrolytic oxidation to locally release both osteogenic (i.e., strontium) and anti-bacterial (i.e., silver ions) agents, along with the incorporation of hydroxyapatite into the surface of the implant in a same single-step In vitro and Ex vivo (mouse) In-vitro: 1, 3, 7 and 11 days Ex-vivo: 24 h Tremendous potential use of synergistic antibacterial behavior between silver ions and strontium allowing for simultaneously reduced required dose of silver ions by 4-32 folds while inducing osteogenic behavior In-vitro 1, 3, and 7 days The overall distribution of cells on regular porous scaffolds was similar to the orderly arrangement of cultivated crops in the field, and the microporous structure of scaffolds with an aperture close to cell size was more suitable for cell proliferation and adhesion Kayacan et al. [133] L-PBF Transitive porous implants of semispherical and elliptical pores of 500 µm at outer region and cell development channels of 200 µm at the inner region

Sheep 12 weeks
Semispherical porous structure resulted in the formation of a strong and durable bond between tissues and implant Due to the transitive porous structure, a strong and durable tissue bond could be formed at the bone-implant interface Cells on AM discs exhibited expedited proliferative activity when compared to titanium plasma spray (TPS) coated samples; 48% higher calcium production on AM discs than on TPS discs; No significant enhancement of cellular activities with HA-coating; Additive manufactured porous titanium surfaces can be alternative to TPS-coating for orthopedic spinal implants

Current Challenges and Literature Gaps
New bone formation is normally facilitated by spaces in pores for bone growth and larger porosity of over 75% as they improve the bodily fluid permeability [99]. The higher the porosity of implants, the weaker the mechanical strength it would be. A higher degree of control over the macro-level-porosity pore sizes and strut thickness can be achieved through lattice/unit cell designs, however, control over the microlevel porous design can be challenging, especially for processes such as L-PBF. The use of additive manufacturing for the fabrication of medical devices has two major advantages of 'batch-size-indifference' and 'complexity-for-free' [48]. Batch-size-indifference provides the feasibility of producing patient-specific biomaterials, implants, and surgical instruments that exactly match the complex and highly variable anatomy of individual patients. Whereas, complexity-forfree provides freedom to designers to use complex geometries that give rise to favorable properties and advanced functionalities [48]. Implants made of Ti64 alloy are biocompatible in nature, however, the bioactivity of Ti64 implants can be enhanced both in vitro and in vivo by surface coating them with bone-like apatite for better bone ingrowth performance and success rate of the implants [139]. Orthopedic implants can be patient-specific and the need for additional surface coating/surface activation adds to the manufacturing costs of the implants [41,53].
In addition to the mechanical properties and porous structure, permeability, fatigue behavior, surface area and many other properties of additively manufactured porous implant structures can be adjusted through the application of geometrical designs [39]. Even with the optimum geometrical design, it is difficult to address all the challenges associated with the orthopedic implants and there exists a tradeoff between porosity and stiffness/strength while simultaneously promoting osseointegration and avoiding pore occlusion [39,64,103]. Muthaiah et al. [53] suggested that a new functionally graded porosity implant with bioceramic coating should be developed to mimic natural bone and provide a good interfacial bond between bone tissue and metallic implants. Currently, the primary challenges in the clinical use of AM porous implant structures are in guaranteeing the consistency of material produced and its optimum performance [43].
It is especially challenging to evaluate periprosthetic bone loss prior to an operation based on the X-rays and computed tomography (CT) and makes it difficult to design the implant accurately [92]. Furthermore, the proneness of porous implants towards fatigue failure as compared to their solid counterparts can be a challenging factor that requires careful consideration, but limited studies have been conducted on the fatigue behavior of porous structures. In addition, the orthopedic implants are often subjected to cyclic tensile load as discussed in Section 4.3, which demands more studies on fatigue tensile behavior of these porous structures.
Many researchers lack the resources and manpower for in vivo testing, although these in vivo studies form an essential part of novel implant performance studies. Furthermore, the in vivo studies are performed for a limited time duration. Extensive and rigorous animal studies for longer durations is required for novel and innovative multifunctional porous implant designs, as the porous implants' strength/performance may deteriorate differently than what was observed in shorter duration studies, such as for Ti64 implant infiltrated with biodegradable material. Despite the closeness of gradient designs to bone morphology, compared to non-gradient designs, most of the articles in the literature were found to be focused on non-gradient Ti64 structures. More comprehensive studies on gradient porosity design are required so that the findings of varying research approaches in AM implant development converge into addressing current challenges, such as insufficient bone, vascularization, contiguous infection, and implant longevity [54]. Furthermore, as most of the additive manufacturing processes demand post-processing, studies on the effect of post-processing treatments on internal porosity, surface roughness, mechanical properties, fatigue behaviors, and biocompatibility properties can be of particular interest.
The application of conventional surface finishing techniques to AM porous implants is insignificant mainly due to porous structures, partial adhesion of powder feedstock and variable roughness with the orientation of the AM implants and possible needs for variable finishes across implants, and therefore demands the need of new advanced surface finishing techniques [43]. Although there has been huge progress in the field of surface engineering on AM manufactured Ti-implants, there is still significant scope for continued research such that a thorough understanding of the process involved in each surface treatment, systematic experimental studies, advanced analytical solutions, and computational simulations can be achieved [53]. The regulating agencies around the world are quite aware of the importance of this issue, however, with the rapid development and technological advancement in AM of medical devices, these directions have not been necessarily kept up and still pose some unique challenges in meeting requirements and establishing regulatory standards associated with quality assurance in manufacturing [54]. More importantly, there is a need for re-scrutinizing the regulatory position on the breadth of the 'design envelope' covering mass-manufactured AM porous implants [43].

Conclusions
Additive manufacturing technology is gaining rapid interest in orthopedic implant design and development due to its design freedom, which allows for patient-specific custom implant designs. In this systemic review, besides the porous design, mechanical properties, in vitro and in vivo biological properties, fatigue behaviors and multifunctionality of AM porous Ti64 implants were carefully reviewed and presented thoroughly for all the AM processes currently utilized in the design and development of these porous Ti64 implants/scaffolds. This review article helps in mitigating the current gap in the existing review articles on AM porous Ti64 related to the review of different AM processes, fatigue behavior and critical design aspects of these porous implants. To reduce the stress-shielding effects of solid titanium implants, researchers are currently focusing on the design and development of porous Ti64 alloy implants along with the aim of creating novel multifunctional implants. Most of the porous Ti64 implants were fabricated using either a laser powder bed fusion or an electron beam melting process. Research studies on porous Ti64 implants suggest that they can be tailored to mimic the mechanical properties (Young's modulus, mechanical strength, etc.) similar to natural bone. Different macroporous cellular structures can be utilized to improve the fluid permeability/cellular activity within the Ti64 bone implants. Furthermore, bioactive, and antibacterial surface modification, infusion of bioactive materials within the Ti64 matrix, and surface treatment to form functionally packed porous implants have a perspective towards future multifunctional orthopedic Ti64 alloy implants. Since gradient porous design mimics closest to the natural bone and has promising potential as orthopedic implants/scaffolds, more studies on these gradient designs are necessary. Further in-depth studies on design, mechanical, fatigue, biological, and multi-functionality aspects of porous Ti64 along with the convergence of future study findings, can turn the porous Ti64 implants into feasible alternatives for the currently expensive Tantalum implants. The comparison of AM porous Ti64 with other manufacturing techniques for orthopedic implant/scaffold can also be an interesting topic for future study, such that their respective advantages and disadvantages can be identified.