A nonradiographic strategy to real‐time monitor the position of three‐dimensional‐printed medical orthopedic implants by embedding superparamagnetic Fe3O4 particles

Monitoring the position of orthopedic implants in vivo is paramount for enhancing postoperative rehabilitation. Traditional radiographic methods, although effective, pose inconveniences to patients in terms of specialized equipment requirements and delays in rehabilitation adjustment. Here, a nonradiographic design concept for real‐time and precisely monitoring the position of in vivo orthopedic implants is presented. The monitoring system encompasses an external magnetic field, a three‐dimensional (3D)‐printed superparamagnetic intervertebral body fusion cage (SIBFC), and a magnetometer. The SIBFC with a polyetheretherketone framework and a superparamagnetic Fe3O4 component was integrally fabricated by the high‐temperature selective laser sintering technology. Owing to the superparamagnetic component, the minor migration of SIBFC within the spine would cause the distribution change of the magnetic induction intensities, which can be monitored in real‐time by the magnetometer no matter in the static states or dynamic bending motions. Besides horizontal migration, occurrences of intervertebral subsidence in the vertical plane of the vertebrae can also be effectively distinguished based on the obtained characteristic variations of magnetic induction intensities. This strategy exemplifies the potential of superparamagnetic Fe3O4 particles in equipping 3D‐printed orthopedic implants with wireless monitoring capabilities, holding promise for aiding patients' rehabilitation.

Slippage and dislodgement of in vivo orthopedic implants commonly exist when the patients take improper behaviors or exercises, [1,2] leading to the failure of treatments and inevitable secondary surgeries. [3]Thus, real-time monitoring of in vivo orthopedic implants is the next key step in advancing digital postoperative rehabilitation with better treatment effects. [4,5]Currently, the positions of orthopedic implants are commonly observed by radiographic methods, such as X-ray examination or computed tomography (CT). [6]Some nonradiographic composed of magnets and sensors were invented in the field of orthopedic prosthesis. [7,8]Nevertheless, the hazard of magnet exposure to human tissue cannot be dismissed lightly, and the wire connected to external power is deemed impractical.However, those radiographic investigations require special equipment and surgeons' professional experience and with a key inability to provide timely feedback in daily rehabilitation, [9] indicating the inconvenience for patients in both space and rehabilitation adjustment aspects.20] Monitoring the position/motion of orthopedic implants under the skin by a nonradiographic method is still an unresolved question.
Deep-tissue monitoring is an emerging concept, [21,22] showing great potential for medical applications but lacking available material engineering. [23]Three criteria are required for the monitoring of orthopedic implants inside patients.The first one is free of connective wires.Roriz et al. [24] investigated the mechanical condition of the sheep's lumbar intervertebral discs by directly inserting an optical fiber sensor.However, this invasive sensing method would greatly increase the possibility of infections in the exposed skin area, [25] as well as further tissue rejection occurrence. [26,27]hus, a wireless monitoring strategy is preferred. [28,29]otably, past wireless implants have predominantly concentrated on vital signal monitoring, [30] while comprehensive reports pertaining to the monitoring of implant locations have been conspicuously absent.Second, biocompatibility is of great importance for orthopedic implants.The selected material species should not contain toxic elements. [31,32]Finally, more complex systems will possess the potential for more catastrophic failure. [33]Especially in the chemically/ biomechanically unstable human internal environment, the breakage of a small part can result in failure of the whole sensing system.Thus, an all-in-one simplified design, including signal collection, signal transmission, and power supply, is the key to guarantee normal working for deep-tissue monitoring. [34]ll three of the aforementioned key features for monitoring in vivo orthopedic implants may be solved by integrating superparamagnetic Fe 3 O 4 particles to the implants.In our previous studies, the movements [35][36][37] or self-deformation of Fe 3 O 4 particle groups could cause the change of surrounding magnetic fields, yielding the generation of sensing signals based on Faraday's electromagnetic induction effect.During the whole process, the magnetic particles were kept away from the coil receiver, indicating their capability of wireless signal transmission (criterion 1).[40][41] This phenomenon means the biosafety of Fe 3 O 4 particles even after they have been implanted within human bodies (criterion 2).Finally, compared with complex and exquisite electronic devices based on microfabrication, the integration of Fe 3 O 4 particles in orthopedic implants is facile and low-cost.More importantly, its structure is simple enough to resist unpredictable destruction from an unstable human internal environment while maintaining its working capability (criterion 3). [42]ere, we demonstrate a nonradiographic strategy to real-time monitor the position of three-dimensional (3D)-printed medical orthopedic implants based on superparamagnetic Fe 3 O 4 particles.Taking interbody fusion as an example, it refers to the insertion of an intervertebral body fusion cage (IBFC) into the lumbar intervertebral disc aiming to achieve solid fusion and restore sagittal alignment. [43]We incorporated the Fe 3 O 4 /Ecoflex component onto the polyetheretherketone (PEEK) framework through the high-temperature selective laser sintering (HT-SLS) process.This integration enabled the acquisition of the superparamagnetic intervertebral body fusion cage (SIBFC), which combined both the superparamagnetic functionality and structural strength.To determine the position of SIBFC in the vertebra, the magnetic induction intensity data were measured when the SIBFC migrated to different regions, as well as the dynamic test of bending and twisting of the lumbar spine containing SIBFC.These results confirm that the integration of Fe 3 O 4 particles into SIBFC has satisfactory-mechanical properties and real-time positioning functionality.Thus, this strategy can serve for the wireless monitoring of SIBFC inside patients after percutaneous lumbar discectomy, which would facilitate their rehabilitation and recovery.
To enhance the structural configuration of the IBFC while preserving the core load-bearing prowess, a serrated external framework was devised to thwart its unexpected migration (Figure S1).A hollow semicylindrical groove was affixed to the inner wall of SIBFC to accommodate the superparamagnetic component (Fe 3 O 4 /Ecoflex composite).First, the framework of SIBFC was fabricated utilizing the self-developed HT-SLS of HK PK 125 (Figure S2).Following the completion of the PEEK framework, the whole system was gradually cooled to room temperature.Subsequently, Fe 3 O 4 /Ecoflex liquid was introduced into the hollow semicylindrical groove to replace unsintered powders.Finally, once the Fe 3 O 4 /Ecoflex had cured, the groove would undergo a post-sintering process to seal it securely (Figure 1A). Figure 1B,C display the schematic diagram depicting the operational mechanisms of the SIBFC following its implantation within the spine.The real-time monitoring system consists of the SIBFC, an external magnet, and a magnetometer.Consequently, if the SIBFC undergoes migration from its initial implantation site, the magnetometer could detect the magnetic variation and relay detected data to a smart device.Subsequently, the smart device can ascertain the real-time position of the SIBFC by analyzing the distinctive magnetic data.

| Structural observations of SIBFC
PEEK is a kind of high-performance engineering plastic and finds extensive application in medical engineering. [44]e HT-SLS printing method enabled the production of intricately shaped samples with precise accuracy, [45,46] as shown in Figure 2A.Since the superparamagnetic Fe 3 O 4 component was embedded inside the SIBFC, it is difficult to observe.Micro-CT serves as a nondestructive 3D imaging system that can provide structural information inside solid composites (Figure 2B). [47]Thus, Figure 2C  framework.The variations in colors are attributed to the presence of Fe 3 O 4 , which introduced differences in X-ray absorption and scattering properties compared to the PEEK polymer.
The vibrating sample magnetometry (VSM) testing curve reveals that the Fe 3 O 4 /Ecoflex composite inside the SIBFC exhibits a superparamagnetic behavior (Figure S3 and File S1). [48]Besides the micro-CT observation, the uniform dispersion of Fe 3 O 4 particles in the composite has been confirmed by the energy dispersive spectroscopy (EDS) mapping of Fe element (Figure S4).

| Mechanical property of SIBFC
Intervertebral subsidence (IS) is defined as the displacement of IBFC by more than 3 mm from its intended implantation position (IIP) during postoperative followup. [49]This indicates a compromise of adjacent vertebrae, signifying that the stress surpasses the maximum stress tolerance of endplates.A common requirement for various IBFC is to possess the appropriate stiffness to withstand the complicated loads experienced following implantation in the spine, while ensuring the endplates are intact.[52] To further evaluate the mechanical properties of SIBFC, two kinds of normative tests were conducted.
The static compression test result reveals an unusual inflection point in the load-displacement curve (displacement of 0.68 mm, Figure 2D).This can be attributed to the sawtooth shape on the surface of SIBFC, resulting in initial disruption occurred at the unstable line contact when the compression was applied.The yield load of SIBFC was beyond 3000 N, more than twice the strength for vertebral bone of approximately 1500 N. Considering activities such as standing, sitting, and walking, the estimated axial compressive stress on the lumbar spine ranges from approximately 500-1000 N. [53] Thus, the strength of the SIBFC fabricated utilizing HT-SLS falls within a safe and reasonable range.Moreover, the stiffness of SIBFC (K d ) was measured to be 3723.57± 163 N mm −1 (Figure 2D), and the lower stiffness is less likely to cause damage to the endplates and effectively prevent the stress shielding effect, while still meeting the necessary requirements. [54]ollowing the implantation of SIBFC into the vertebra, a notable disparity in stiffness may arise when compared to its intrinsic stiffness, a set of interconnected spring systems was used to replicate the spinal environment (Figure 2E).The subsidence curve demonstrated a yield load of 2236.54N. The stiffness (K s ) of the system comprising SIBFC and rigid polyurethane foams was measured to be 546 ± 39 N mm −1 .After the subsidence potential test, the SIBFC and rigid polyurethane foams are shown in Figure 2F.The SIBFC remained visually undamaged, while the significant indentation on polyurethane foam displayed subsidence evidence.In fatigue dynamic test, the SIBFC still maintained ideal mechanical property after 5 × 10 6 cycles (Figure S5).The subsidence potential of SIBFC was negatively correlated with K p (stiffness of polyurethane foams), which can be calculated using the following equation: where K p represents the stiffness of polyurethane foams, K s represents the stiffness of the system comprising SIBFC and rigid polyurethane, and K d represents the stiffness of SIBFC.
The inherent complexity and variability of spine loading, coupled with the diverse designs of IBFC, present considerable challenges in establishing a standardized set of data metrics.Consequently, the mechanical data is primarily intended for facilitating horizontal comparisons in a database among different IBFC, rather than serving as definitive performance standards.Although the K p of 641.15 N mm −1 surpassed certain IBFC submitted by manufacturers approved through the traditional 510(k) process, the risk of subsidence for SIBFC is acceptable.In summary, the fabricated SIBFC exhibited desirable mechanical properties, including the low elastic modulus and acceptable potential for subsidence, which would efficiently mitigate the stress shielding effect and IS.

| Static monitoring the position of SIBFC before/after its migration
This nonradiographic strategy was based on magnetic interference when the superparamagnetic component inside SIBFC was placed in the applied magnetic field.When the location of SIBFC changed, the magnetic field propagation path can be altered, thereby changing the magnetic field distribution recorded by the magnetometer.By analyzing characteristic magnetic induction intensity data of SIBFC at different positions, it's feasible to accurately identify the location of SIBFC, allowing to confirm whether or not it has deviated from the IIP.
Figure 3A,B illustrates the optical image and schematic diagram of the real-time position monitoring system for SIBFC.Importantly, the superparamagnetic component was co-implanted with the cage, originally designed for implantation.For the experiment, an artificial spine with the pelvis was used to mimic the human spine, of which the SIBFC was positioned between the L3 and L4 vertebrae (Figure S6).The center of the external magnetic field, SIBFC, and the magnetometer were positioned on the same horizontal plane.As shown in Figure 3C, the original magnetic field propagation can be interfered with by the SIBFC implanted in the vertebra due to arises from its superparamagnetic property endowed by Fe 3 O 4 /Ecoflex composite.The magnetometer was placed independently at the posterior section of the spine to monitor its surrounding magnetic data.The PNI RM3100, a printed circuit assembly with three sensor coils, [55] can provide three-axis magnetic field sensing in a high sensitivity and low noise (Figure S7).It was connected to a computer via a microcontroller unit (MCU) for data acquisition (Figure S8).
Cage migration (CM) is the most significant type of failure in interbody fusion surgery, [3] which can lead to nerve compression, vertebral disintegration, and fracture.It is defined as a displacement exceeding 2 mm from the originally intended implant position as observed in postoperative imaging, specifically into the intervertebral foramen. [56]To analyze the extent and direction of CM within vertebrae, the transverse vertebrae have been partitioned into 12 regions, measuring 2 mm laterally and 1.5 mm vertically for each region (Figure 3D).The partitioned regions encompassed both anterior/posterior displacement, as well as lateral displacement (left and right).Focusing specifically on areas of clinical concern, the posterior regions were further categorized into mild CM and severe CM.
Magnetic induction intensities (three-axis components, Bxc, Byc, Bzc) recorded by the magnetometer were used to observe changes when the SIBFC migrated from its IIP (green one in Figure 3D) to adjacent regions (red one in Figure 3D).By establishing a correlation between the positions and magnetic induction intensities data, an assessment regarding the occurrence of CM can be provided.When the SIBFC was primarily implanted at the IIP, three-axis components (Bxc, Byc, Bzc) recorded by magnetometer were 235, 24, and 555 μT, respectively.To mimic the occurrence of CM in different positions, the SIBFC was relocated to various regions within the vertebral body by using surgical forceps.The magnetic field data was then recorded after each relocation.Figure 3E-G illustrates changes in magnetic induction intensities recorded by the magnetometer when the SIBFC was positioned in different regions.Specifically, the detailed data can be found in Table 1.
The measured results revealed significant differences in the Byc and Bzc of the magnetic data when the SIBFC was located in different regions.Specifically, when the SIBFC migrated from the anterior to the posterior regions, the corresponding ΔByc and ΔBzc consistently exhibited an increasing trend.In all regions on the right (regions of C1-C4), the magnitude changes of ΔByc exceeded 30%, while less than 0.2% for ΔBzc relative to the IIP (region of B3).In the medial region (regions of B1-B4), the ΔByc changes were relatively small, all below 5%, either for ΔBzc (less than 0.11%).In all regions on the left (regions of A1-A4), the ΔByc demonstrated an approximate change ranging from 10% to 15%, while 0.05% to 0.2% for ΔBzc.For the change of magnetic induction intensity in X component (ΔBxc), the absolute percentage change of ΔBxc was less than 1%.Notably, there was a decreasing trend observed in ΔBxc from the right side to left side.Specifically, when the SIBFC migrated to severe regions (A1, B1, and C1), magnetic induction intensities recorded by the magnetometer showed higher values of ΔBxc in comparison to the IIP.This can provide an additional means to distinguish the occurrence of CM and ascertain the severity of CM (mild and severe).
By analyzing the magnitude of ΔByc and ΔBzc recorded by the magnetometer when the SIBFC was relocated in different regions, it becomes readily apparent to distinguish between the right side, medial, and left side regions.Additionally, by considering the magnitude of ΔBxc, the relocated regions of the SIBFC can be easily distinguished.Taking the example of the SIBFC being relocated in medial regions, several criteria can be used for classification.First, if the ΔByc was below 5%, it can be determined as being in the medial region (B1-B4).Next, if ΔByc was negative, indicating the migration to anteromedial region (B4).Otherwise, further judgment can be made based on the magnitude of ΔBxc.A positive value of ΔBxc indicates a severe CM region (B1), while a negative value suggests a mild CM region (B2).
The aforementioned results can be explained by the magnetic dipole model.Given that the reinforced effect in magnetic propagation induced by the SIBFC was smaller than that of the external magnetic field, and considering the fact that the SIBFC was situated at a greater distance from the magnetometer. [57]The magnetic field enhanced by superparamagnetic composite can be considered as the magnetic dipole, and this assumption is acceptable for simplifying the calculation.For a magnetic dipole, its magnetic field can be expressed as [58,59] T A B L E 1 The relative change of magnetic induction intensity data for each region to which SIBFC migrates.where B represents the magnetic induction vector in the test point, μ 0 is the permeability of the vacuum, M represents the magnetic moment vector and R represents the radius vector from the point dipole to some point, R represents the magnitude of R. The magnetic moment vector of M also satisfies the following equation [60] :

ΔBxc
where H represents the magnetic field intensity, and the χ m represents the magnetic susceptibility of the superparamagnetic composite.Equation ( 2) can be reduced to a commonly used simple representation, as shown in below equation [58] The magnetic induction intensity B at the test point, which corresponds to the position of the magnetometer, can be regarded as consisting of three distinct components: (1) the geomagnetic field (B g ), (2) the external magnetic field (B e ), and (3) the magnetic field generated by the introduced SIBFC (B s ), which the enhanced magnetic source of SIBFC can be approximated as an equivalent magnetic dipole as above assumption.As shown in the below equation Although the geomagnetic field (B g ) and the original external magnetic field (B e ) may exhibit variability in different testing environments, their changes within a measurement cycle are typically minimal.As a an initialization process was conducted before measurements to correct the baseline (referring to the initial magnetic field or reference point), which includes environmental magnetic fields like the geomagnetic field and the external magnetic field.This initialization aided in establishing a reference point and eliminating any constant magnetic field components, thereby enabling more precise measurements of magnetic induction intensities variations attributable to the introduced superparamagnetic component.Hence, the variation of B (ΔB) at the test point before and after migration for SIBFC can be calculated as where H original and H migration represent the magnetic field intensity at SIBFC before and after migration, respectively, the R original and R migration are the distance between SIBFC and the magnetometer.
Given that the migration of SIBFC within lumbar spine was minimal, the resulting difference in the magnetic field intensity H (H original and H migration ) surrounding the SIBFC caused by its migration was negligible.Hence, this study assumed that the primary factor influencing the changes of magnetic induction intensity B (ΔB s ) at test point was the spatial migration of SIBFC, which subsequently affected the distance between SIBFC and magnetometer.Thus, the distance R between SIBFC and the magnetometer played a significant role in determining the magnetic induction intensity B recorded at the test point.
Therefore, as the decrease of distance between the SIBFC and the magnetometer, the recorded magnetic flux intensity increased.In other words, when the SIBFC migrated from anterior regions (A4, B4, and C4) to posterior regions (A1, B1, and C1), the magnitude of B displayed an increasing trend, which was in accordance with the experimental results well (Figure 3E-G).Overall, the influence of the SIBFC on the magnetic field propagation depended on its spatial position, resulting in distinct magnetic intensities observed by the magnetometer.

| Theoretical simulation for monitoring the SIBFC migration
As discussed above, the magnetic induction intensities obtained from the magnetometer can be utilized to distinguish the migration of SIBFC within the vertebrae.Figure 4A visually illustrates the SIBFC migration within vertebrae under an external magnetic field.To provide a direct observation of the SIBFC's interference on magnetic field propagation, the distribution of magnetic induction intensity surrounding SIBFC in the presence of an external magnetic field was measured (Figure 4B).Additionally, simulated calculations were performed using Ansys Maxwell to analyze the localized amplification of the superparamagnetic component (Figure 4C).The color bars indicate that a larger read area corresponded to a greater magnetic flux of the coil. [61]Both experimental and simulation results revealed that the magnetic induction intensity (B) exhibited a decrease in value as the distance from the applied magnetic field increased.In the left region of SIBFC, closer to the external magnetic field source, the equipotential lines of B shifted toward the left, indicating a reduction in the magnetic induction intensity.Conversely, on the right region of SIBFC, in the direction of magnetic field propagation, the equipotential lines shifted toward the right, suggesting an augmentation of the magnetic induction intensity in that localized region.
The relatively higher magnetic permeability of the Fe 3 O 4 superparamagnetic material within the SIBFC compared to the air medium could change the distribution of magnetic induction intensity when the SIBFC was placed in an external field.The cyclic migration tests were conducted using a plastic surgical clamp attached to a 3D platform, which allowed for precise control of the movement of the SIBFC in various directions.The SIBFC was moved back and forth between the IIP (reference point, B3 in Figure 3D) and its adjacent regions.During migration tests, the waveform of magnetic intensities (ΔBxc, ΔByc, ΔBzc) exhibited peaks and valleys (Figure 4D-F), indicating that the change of magnetic induction intensities greatly depended on the SIBFC migration.After the SIBFC reached the adjacent region, it returned along the original path back to the IIP.
Taking the anterior migration (point d in Figure 4A) of the SIBFC as an example, the changes of three components of magnetic induction intensities show a decreasing trend (Figure 4D).This is expected as the SIBFC migrated away from the magnetometer, resulting in a decrease in magnetic induction intensity.However, when the SIBFC migrated toward the posterior region (point f in Figure 4A), enhancements of ΔByc and ΔBzc can be observed (Figure 4F).These results can be explained by the simulated results of magnetic induction intensities (Figure 4D1-F1).When the SIBFC migrated to the posterior region, the distance between SIBFC and the magnetometer was closer compared to the anterior region (Figure 4D1,F1).The proximity of SIBFC to the magnetometer led to a stronger magnetic induction intensity according to Equation ( 6).The increased variation in ΔByc when the SIBFC migrated to the left side (point e in Figure 4A) can be attributed to the noncentral arrangement of the three-axis sensors in the magnetometer.Although the external magnetic field source, SIBFC, and the sensor were aligned at the same geometric center before the experiments, the internal placement of the three sensors deviated from the geometric center (Figure S7).Thus, a shorter distance in the y-axis existed between the SIBFC and the magnetometer when the SIBFC migrated to point e.These simulation results were consistent with the experimental observations and offered further insights into the variations in magnetic induction intensity resulting from the migration of SIBFC in different directions within the lumbar spine.

| Dynamic bending tests of SIBFC within the lumbar spine
The aforementioned experiments and simulations have demonstrated the utility of static measurements in realtime monitoring the positions of SIBFC.To ensure the reliability of this monitoring strategy, the dynamic experiments of the artificial lumbar spine loading the SIBFC, considering various bending angles, were conducted.Figure 5A presents the photographic images of the lumbar spine during experiments, capturing bent positions that simulate real-life scenarios.Furthermore, Figure 5B elucidates three distinct states of the SIBFC in the lumbar spine: IIP (left), posterior migration (CM, middle), and IS (right).The last two states correspond to potential failure scenarios that may occur within the human body.
The motion paths of the SIBFC in the spine can be influenced by different body postures for the same motion state.Consequently, to improve recognition accuracy and minimize external interference, this study adopted a simulated scenario of an individual seated in a chair by fixing the artificial spine bottom on the table. [62]he dynamic tests involved bending and twisting of the spine at various degrees were conducted to distinguish between the CM at the vertebral plane and IS in the vertical plane.
We conducted a series of cyclic bending tests on the spine with the SIBFC at three different angles: 10°, 20°, and 30°, as depicted in Figure 5C-E.The artificial lumbar spine bottom was fixed at the table and bent by the control setup.To extract valuable information and facilitate the differentiation of signals within the collected data.We generated plots representing the relationship between the magnetic induction intensities and the different states of the SIBFC within the lumbar spine under various bending angles, as shown in Figure 5F.Furthermore, Figure 5G provides a comparative analysis of the changes in magnetic induction intensities associated with CM and IS at different bending angles, in comparison to the scenario where no migration occurs.This comparison enabled a deeper understanding of magnetic variations under different conditions and facilitated in assessing the security of SIBFC in the lumbar spine.
Taking the bending angle of 10°as an example, we conducted an analysis of the magnetic induction intensity variations in the Y component (ΔByc) for three states of SIBFC within the lumbar spine (IIP, CM, and IS).The observed differences in ΔByc recorded by the magnetometer were minimal among these states (approximately 4 μT in ΔByc), as shown by the almost flat line in Figure 5F.The same pattern holds true for bending angles of 20°and 30°, as indicated by the ratio ΔByc/ΔByc 0 of ∼1, as shown in Figure 5G (the ΔByc 0 indicates the recorded magnetic induction intensity when the SIBFC was in IIP, this parlance remained applicable in the following cases).This indicated that ΔByc was not a suitable feature for distinguishing the position of the SIBFC.When the CM of SIBFC occurred, there was a larger change of magnetic magnitude in X and Z component values (ΔBxc, ΔBzc) of −1.83 and −2.14 μT, respectively, when compared to the situation where SIBFC was positioned at the IIP, which resulted in a change of −1.06 and −1.04 μT respectively.These observations provided valuable insights for detecting SIBFC's CM, that a criterion based on the ratios of ΔBxc/ΔBxc 0 and ΔBzc/ΔBzc 0 can be established.In which, ΔBxc 0 and ΔBzc 0 were the changes of magnetic induction intensities when SIBFC was at the IIP during dynamic tests.Therefore, it can be concluded that the SIBFC migration, or named CM, exists when both of the ratios (ΔBxc/ΔBxc 0 and ΔBzc/ΔBzc 0 ) are greater than 1 (Figure 5G).
When the SIBFC experienced IS and CM in the spine, the difference in the magnetic induction intensities along Y and Z components (ΔByc and ΔBzc) were relatively insignificant (Figure 5C-E).However, there was a notable difference in the X component of magnetic field intensity (ΔBxc) between CM and IS.The ΔBxc decreased when SIBFC was at the IIP or in the presence of CM, while it showed an increase when the IS occurred.This distinct pattern allowed for distinguishing the CM and IS of SIBFC based on the value of ΔBxc.To identify IS of SIBFC, the criterion is that ΔBzc/ΔBzc 0 should be greater than 1 and ΔBxc/ΔBxc 0 should be less than 0. This criterion including CM and IS holds true for bending angles of 20°and 30°as well, where the characteristic signals can be summarized as shown in Table S2.
The distinguishing characteristics were consistent for different bending angles of 10°, 20°, and 30°.First, the deviation of the SIBFC from IIP can be determined by evaluating the ratio of ΔBzc/ΔBzc 0 .If this ratio exceeds 1, it indicates a deviation.Subsequently, based on the ratio of ΔBxc/ΔBxc 0 , the occurrence of either CM (when the ratio is greater than 1) or IS (when the ratio is less than 0) can be distinguished.Notably, for the sake of convenience in daily monitoring, the magnetic induction intensities should be measured and recorded immediately after the SIBFC was implanted into the lumbar spine, which can serve as a reference for assessing future occurrences of migration.
In scenarios where the SIBFC was at the IIP, we conducted simulations of magnetic induction intensity at the magnetometer position including the upright position and various bending angles of the lumbar spine.The region of interest was represented by a circular area, as depicted in Figure 5H. Figure 5I-L display the simulated magnetic flux passing through the equivalent coil for different lumbar bending angles.The color bars indicate that a larger read area corresponded to a greater magnetic flux of the coil.The simulated results revealed that in the upright position of the lumbar spine, the SIBFC was in closest proximity to the sensor, leading to the strongest magnetic field signal detected by the coil.As the bending motion progressed, the SIBFC gradually moved away from the sensor.As discussed earlier, the locally enhanced effect of the SIBFC on the magnetic induction intensities at the sensor location diminished.This trend is consistent with the observed practical experiments (Figure 5C-E).
Given the limitations in simulation accuracy of Ansys Maxwell and the complicated motion path of SIBFC along with the lumbar spine, achieving an exact match with actual data poses a challenge.Consequently, the simulations indicated that the enhanced impact of the SIBFC on the magnetometer position decreased as the bending angle increased.This trend became particularly pronounced at bending angles of 20°and 30°.When the bending angle of the lumbar spine reached 20°, the SIBFC was already significantly distant from the sensor.Thus, further increasing the bending angle to 30°yielded only minimal changes in magnetic flux (Φ 20 -Φ 30 equaled ∼0.85 × 10 −9 Wb).

| Dynamic twisting tests of SIBFC within the lumbar spine
The successful verification of the nonradiographic strategy through sagittal plane flexion and extension movements has been demonstrated.To further enhance the discriminative features and bolster the credibility of the assessment, twisting tests in the horizontal plane of the lumbar spine were conducted.The simulation involved an individual seated on a chair and sequentially performing the following actions: twisting 30°to the left, returning to an upright position, and then twisting 30°to the right, and returning to an upright position.This sequence was considered one cycle.The experimental result, as depicted in Figure 6A, reveals that when the lumbar spine twisted to the left, there was an increase of approximately 30 and 110 μT in ΔBxc and ΔByc, respectively, for the three positions of the SIBFC (IIP, CM, and IS).Conversely, when the lumbar spine twisted to the right, there was a decrease of approximately 20 and 80 μT in ΔBxc and ΔByc, respectively.This phenomenon can be attributed to the directional nature of the magnetic induction intensity vector, where the distinct directions (left and right in relation to the magnetometer) resulted in difference in the positive and negative values of ΔBxc and ΔByc.The asymmetry observed in the numerical values for different twisting directions can be attributed to the specific design of SIBFC.
In later lumbar interbody fusion surgeries, it is common for the center of the device to be aligned with the center of the vertebra.In this study, our focus was on the scenario where the SIBFC was implanted on the left side of the vertebra.Consequently, the instrument grip and the superparamagnetic component of the SIBFC were positioned on the left side of the geometric center of the vertebra, rather than directly facing the geometric center of the sensor.This characteristic was also evident in the changes observed in ΔBzc.When the SIBFC twisted in a leftward direction with the spine, ΔBzc showed a decrease of approximately 3 μT.However, during rightward twisting, it did not exhibit a single valley.Instead, it displayed a waveform pattern characterized by two peaks and one valley in between.A magnified view of this waveform can be seen in Figure 6C.This behavior may be attributed to the fact that when the spine twisted to the right, the paramagnetic component of the SIBFC initially aligned with the geometric center of the magnetometer, and then moved toward the right.
The simulated distribution of magnetic flux through the equivalent coil during different twisting motions is depicted in Figure 6D-F.When twisting to the left, there was a localized increase in the magnetic induction intensity on the left side of the coil, corresponding to the enhanced superparamagnetic component.Conversely, during the process of twisting to the right, there was a localized enhancement when the superparamagnetic component was near the center of the coil.As the twisting motion continued to the right, the superparamagnetic component exerted an enhanced effect of magnetic intensity localized to the right of the sensor, resulting in an inverse trend compared to twisting to the left.
The characteristics of the ΔBzc can serve as an indicator to distinguish whether the SIBFC has migrated.
Figure 6B presents the ratios of each magnetic field component between twisting to the left and twisting to the right.Specifically, the difference between the peak (denoted as "L" in Figure 6C) and the valley of the right twist (denoted as "R" in Figure 6C) can be utilized for the ΔBzc variation analysis.In the case of the SIBFC being the IIP, this value is approximately 1.48.It is observed that when the SIBFC migrated from IIP, the value of ΔBzc-left/ΔBzc-right decreased.For CM, it is approximately 1.13, and 1.48 for IS.This information can aid in determining whether the SIBFC has deviated from its IIP.In this study, we provide a nonradiographic strategy for real monitoring of the position of 3D-printed medical orthopedic implants embedding superparamagnetic Fe 3 O 4 Particles.It is based on the principle that the superparamagnetic component can locally enhance its surrounding magnetic induction intensity in the presence of an external magnetic field.The empirical data, simulation results, and derived formulas indicate that the distance between the superparamagnetic component, serving as an equivalent magnetic dipole, and the magnetometer is the primary determinant of the magnetic induction intensity surrounding itself.
Taking interbody fusion as an example, we fabricated the SIBFC using the HT-SLS technology.First, the as-fabricated SIBFC has satisfying mechanical properties (stiffness of 3725.57± 163 N mm −1 and K p of 641.15 N mm −1 ), which surpasses certain IBFCs approved through the traditional 510(k) process.The experiments to distinguish whether the SIBFC has undergone deviation consist of two aspects.Fortunately, the strategy described in this article extends beyond interbody fusion procedures.The compatibility of superparamagnetic components with readily available orthopedic devices enables their utilization in various regions of the human body where implants are needed, as long as the mechanical performance criteria of the implant site are satisfied.In conclusion, this innovative and user-friendly approach facilitates the regular monitoring of orthopedic implants, thereby enhancing patients' healing by promptly adjusting rehabilitation training under medical supervision.

| EXPERIMENTAL SECTION 4.1 | Materials
PEEK (330PF) powders were purchased from Jilin Joinature Polymer Co., Ltd.This commercialized PEEK underwent a preheating process at 300°C for 5 h before sintering to enhance the flowability and sintering performance of powders, meeting printing standards. [63]The particle size of PEEK after the preheating process ranged from 20 to 50 μm (Figure S9).

| Structural design of SIBFC
The lumbar vertebrae, positioned below the spine and arranged in ascending order from L1 to L5, serve as the upper (L1) connection to the thoracic vertebrae and the lower (L5) connection to the sacrum. [64]Undesired migration of IBFC usually occurred at the lumbar with a pear-shaped disc space, [65] resulting in a larger and concave space at the intended position.Thus, SIBFC was designed to resemble the protruding structure of intervertebral discs to avoid undesired migration after surgery.Moreover, the convex portion of SIBFC was designed with a serrated structure to increase frictional force during potential migration.The vertically oriented hollow structure was also necessary for SIBFC, which can provide space for bone grafting procedures to facilitate bone healing. [66,67]n this study, the SIBFC was designed with dimensions of 45 mm×16 mm × 8 mm, and was positioned between the L3 and L4 vertebrae.The property of the interfacing magnetic field was endowed by Fe 3 O 4 / Ecoflex.A hollow semicylindrical groove, incorporated to the inner side of SIBFC near the surgical instrument grasp, was designed on the inner wall of PEEK framework to accommodate superparamagnetic component (a diameter of 5 mm, and a depth of 4.5 mm, Figure S1).The size of the superparamagnetic component was sufficiently small compared to the surrounding PEEK support framework, endowing the general adaptation for various clinical implantation methods.

| Fabrication of superparamagnetic component
Fe 3 O 4 nanoparticles were mixed with viscous twocomponent Ecoflex liquid silicone (A&B, T605) in a mass ratio of 3:7.The mixed fluid mixture was placed under vacuum of 0.01 atmosphere for 10 min to remove bubbles, yielding a homogeneous fluid mixture. [68]

| 3D printing of the SIBFC
The framework of SIBFC was fabricated utilizing the selfdeveloped HT-SLS of HK PK 125 (Figure S2).Before the sintering process, PEEK powders were dried at 70°C for 24 h, and the dried powders were added into HK PK 125 for preheating.The PEEK framework of SIBFC was then processed.After processing completion of the PEEK framework, the whole system was cooled down to room temperature and an aspiration nozzle was used to extract unsintered PEEK powders in the hollow semicylindrical groove.Following that, the superparamagnetic fluid mixture, comprising 30 wt% Fe 3 O 4 , was then injected into the hollow semicylindrical groove using an injector.Subsequently, the temperature in the working area was increased to 80°C and maintained for 0.5 h to facilitate curing.Then, the temperature was increased to the processing condition for sintering PEEK powders onto the groove for the sealing process, resulting in the final SIBFC.Detailed comprehensive processing parameters can be found in Table S1.

| Characterizations of the SIBFC
The dispersion of the PEEK framework and Fe 3 O 4 /Ecoflex composite was observed by in situ X-ray micro-CT technology (Xradia 515 Zeiss).The surface morphology of Fe 3 O 4 /Ecoflex composite was characterized by field emission scanning electron microscopy (Sirion 200; FEI) equipped with an EDS.
The static axial compression test and subsidence test of SIBFC were conducted by the electronic universal testing machine (AG-IC 100kN; SHIMAD-ZU).Notably, two stainless steel blocks were contacted with the SIBFC to transfer static axial load per ASTM F2077, while two grade 15 rigid polyurethane foam blocks (intended to mimic the compression properties of trabecular bone) for subsidence test per ASTM F2267. [50]he magnetic property of Fe 3 O 4 /Ecoflex composite was characterized using a VSM (VSM, LakeShore 7404) under an applied magnetic field sweeping from −20 to 20 k Oe.The magnetic induction intensity distribution surrounding SIBFC in the applied magnetic field was measured by a multidimensional magnetic field test system (F-30; CH-Magnetoelectricity Technology) at 1 mm apart from the top surface of SIBFC.

| Real-time monitoring the position of SIBFC
The real-time monitor device system consists of an external magnetic field source, a SIBFC, and a magnetometer at the fixed point for recording the magnetic induction intensities.The chosen RM3100 magnetometer (PNI) with high sensitivity and low noise properties has been widely used in aerospace and automotive fields.The data recorded by the sensor at the test point was transmitted to the computer via a MCU for reading and storing magnetic induction intensities data.
demonstrates the uniform distribution of Fe 3 O 4 particles inside the PEEK framework.The distribution of Fe 3 O 4 /Ecoflex is represented by the color gray-white, and gray-black for the PEEK F I G U R E 2 The distribution of the superparamagnetic component in SIBFC and its mechanical properties.(A) The optical image of SIBFC, (B) the schematic diagram of micro-CT, (C) the cross-sectional result of micro-CT, (D) the result of axial static compression per ASTM F2077, (E) the result of subsidence test per ASTM F2267, and (F) optical images of SIBFC and the compressed rigid polyurethane foam.CT, computed tomography; SIBFC, superparamagnetic intervertebral body fusion cage.

F I G U R E 3
The distribution of static magnetic intensities recorded by magnetometer when the SIBFC migrates to different regions in the horizontal plane of the vertebra.(A) Optical image and (B) schematic illustration of the SIBFC fixed in the artificial lumbar spine, (C) the schematic diagram of the monitoring system and the three-axis components of magnetic induction intensity, (D) the transverse partitions of the vertebra.The changes of magnetic induction intensity (E) in X component (ΔBxc), (F) Y component (ΔByc), and (G) Z component (ΔBzc) when the SIBFC migrates to the corresponding region in comparison to no deviation of SIBFC.The figure was partly generated using Servier Medical Art, provided by Servier, licensed under a Creative Commons Attribution 3.0 unported license.MCU, microcontroller unit; SIBFC, superparamagnetic intervertebral body fusion cage.

F
I G U R E 4 The simulations of magnetic induction intensity surrounding the SIBFC.(A) The schematic diagram of the SIBFC in an external magnetic field, (B) the actually measured and (C) the simulation of localized, magnified magnetic induction intensities distribution surrounding the SIBFC.(D-F) The recorded three-axis magnetic induction intensities and (D1-F1) simulations surrounding the magnetometer when SIBFC migrates toward corresponding points from the intended implantation position.SIBFC, superparamagnetic intervertebral body fusion cage.

F I G U R E 5
The dynamic bending tests and their corresponding magnetic induction intensities.(A) The bent photographic images of the SIBFC within a lumbar spine, (B) the schematic diagram of three distinct stats of SIBFC within the lumbar spine, (C-E) the measured magnetic induction intensities at various bending angles and different states of SIBFC within the lumbar spine, (F) the summary of magnetic induction intensity variations for different bending angles and SIBFC states, (G) the comparison of magnetic induction intensity variation between the occurrence of deviation and nondeviation of the SIBFC, (H) the schematic diagram of magnetometer and equivalent coil.(I-L) The magnetic induction intensity modulus distribution at bending angles of 0°, 10°, 20°, and 30°, respectively.All error bars represent standard deviation based on 20 replicated data points under the same test condition.SIBFC, superparamagnetic intervertebral body fusion cage.

F
I G U R E 6 The dynamic twisting tests and their magnetic induction intensities.(A) The measured magnetic field intensity when the lumbar spine twists, (B) the magnetic induction intensity comparative statistical graph between the lumbar spine twisting to the left and to the right (ΔB left twist /ΔB right twist ), (C) the magnified curve of magnetic induction intensities in Z component (ΔBzc) during tests, and (D-F) the magnetic induction modulus distribution of coil corresponding to various degrees of lumbar twisting.All error bars represent standard deviation based on 20 replicated data points under the same test condition.
(1) Static tests, the vertebral plane is divided into 12 regions, the results can provide a reference for the position of SIBFC in the horizontal vertebrae.Denotatively, the changes of magnetic induction intensities data in the Y component (ΔByc) can provide a significant partition, which is 10%-15% of lateral-left regions, <5% of medial regions, and >30% of lateral-right regions.Combining the values of ΔBxc and ΔBzc, the more detailed regions can be distinguished.(2) Dynamic tests, bending and twisting tests of lumbar spine offer valuable reference data to determine if the SIBFC has undergone horizontal migration or vertical subsidence.During the bending test, if both the ratio of ΔBxc/ΔBxc 0 and ΔBzc/ΔBzc 0 are greater than 1, it indicates the occurrence of CM within the horizontal vertebral plane.If the ratio of ΔBzc/ΔBzc 0 > 1 while the ratio of ΔBxc/ΔBxc 0 < 0, it indicates the occurrence of IS within the vertical vertebral plane.The numerical asymmetry observed in ΔBzc during dynamic twisting tests can also serve as a reference for distinguishing the position of SIBFC within the lumbar spine.
Fe 3 O 4 nanoparticles with an average diameter of 20 nm were purchased from Beesley new materials (Su zhou) Co., Ltd.Ecoflex (T605#A&B) was provided by Huizhou hongyejie technology Co., Ltd.