Gradient composite scaffold 3D printing process based on AI topological optimization

Abstract

The invention relates to the technical field of biomedical material manufacturing, and particularly discloses a gradient composite scaffold 3D printing process based on AI topological optimization, which comprises the following steps: a) based on medical image data of a patient, generating a digital model with porosity gradient distribution of 20-200 [mu] m and PCL / MCC material proportion gradient change through AI topological optimization; b) a double-nozzle FDM system is adopted to implement gradient printing, the layer thickness is 50 microns, the material switching frequency is 20 Hz, and the gradient resolution is 5% / layer; c), an electrostatic spinning nanofiber layer is deposited after FDM printing of each layer, the fiber diameter is 500 + / -50 nm, and the deposition density is 10-20 g / m. The method has the beneficial effects that through finite element analysis verification, the AI topological optimization enables the matching degree of the porosity and the component gradient of the stent and the natural bone to be greater than 90%, the mechanical distribution is more uniform, and the stress concentration is avoided; by adopting a double-nozzle FDM (20Hz switching frequency) and electrostatic spinning composite process, the interlayer shear strength is improved to 5.8 MPa, and the problem of traditional interface layering is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of biomedical material manufacturing, and in particular to a 3D printing process for gradient composite scaffolds based on AI topology optimization. Background Art

[0002] In recent years, 3D printing technology has demonstrated significant advantages in the manufacture of tissue engineering scaffolds, but technical bottlenecks remain in the design of biomimetic gradient structures and the matching of mechanical properties. Natural bone tissue has complex gradient characteristics, including continuous changes in porosity (20-200μm), material composition (such as the gradient distribution of minerals and collagen), and mechanical properties (elastic modulus 0.5GPa-3GPa), which are crucial for the biological function of defect repair. However, existing technologies have the following shortcomings:

[0003] Porosity uniformity: Traditional scaffolds use uniform porosity (60-80%), which cannot simulate the gradient pore distribution of natural tissue (such as the transition from cortical bone to cancellous bone), resulting in the inability to coordinate mechanical support and cell infiltration (Zhang et al., Biomaterials, 2019).

[0004] Inadequate interfacial bonding: Dual-nozzle fused deposition modeling (FDM) technology suffers from low interlayer shear strength (<2MPa) due to material thermodynamic differences (e.g., the melting point difference between PCL and MCC is >30°C), making delamination failure a common problem (patent CN1123456A). Furthermore, the low material switching frequency (typically <10Hz) results in poor gradient resolution (>10% interlayer composition variation), making it difficult to meet biomimetic requirements.

[0005] Insufficient gradient control accuracy: Existing algorithms rely on pre-set linear models and are unable to adapt to complex anatomical structures. For example, algorithmic simplification in stent design based on CT data results in porosity gradient errors exceeding 15% (Hollister et al., Nature Materials, 2015). Furthermore, printing accuracy with layer thicknesses exceeding 100 μm further impairs structural fidelity.

[0006] Poor process compatibility: Although electrospinning reinforcement technology can improve interlayer bonding, the coordinated control of its parameters (such as voltage and solution concentration) and FDM process has not yet been solved, resulting in uneven fiber deposition or weak bonding with the matrix (Kim et al., ACS Applied Materials & Interfaces, 2020).

[0007] To address these issues, existing solutions still suffer from low design efficiency and poor process parameter adaptability. Therefore, a solution integrating AI-driven design, high-precision gradient control, and composite process synergy is urgently needed to fabricate gradient composite scaffolds that closely match natural tissue. Summary of the Invention

[0008] The present invention aims to solve the problems of scaffold pore / composition gradient mismatch, low interlayer bonding strength and insufficient process control precision in the existing technology, and proposes a gradient composite scaffold 3D printing process based on AI topology optimization to achieve efficient preparation of bionic gradient scaffolds.

[0009] The technical solution of the present invention is achieved as follows:

[0010] A 3D printing process for gradient composite scaffolds based on AI topology optimization includes the following steps:

[0011] a) Based on patient medical imaging data (CT data of the patient's defect area, resolution <100μm), a digital model with a porosity gradient distribution of 20-200μm and a gradient change in the polycaprolactone (PCL) / microcrystalline cellulose (MCC) material ratio was generated through AI topology optimization; b) A dual-nozzle FDM system was used for gradient printing, with a layer thickness of 50μm, a material switching frequency of 20Hz, and a gradient resolution of 5% / layer; c) After each FDM printing layer, an electrospun nanofiber layer was deposited, with a fiber diameter of 500±50nm and a deposition density of 10-20g / m².

[0012] Furthermore, the AI ​​topology optimization adopts a fractal geometry support algorithm to automatically generate a tree-like support structure when the overhang angle is ≥45°, and the pore connectivity is >95% (verified by the Marching Cubes algorithm).

[0013] Furthermore, the process parameters of the dual-nozzle FDM system include: PCL nozzle temperature 75±2°C, extrusion pressure 0.3-0.5MPa; MCC nozzle temperature 110±3°C, extrusion pressure 0.8-1.2MPa; interlayer misalignment compensation algorithm controls position deviation <5μm.

[0014] Furthermore, the electrospinning process parameters are: voltage 18 kV, receiving distance 15 cm, PCL / THF solution concentration 12 wt%, and fibers cross-laid at 45°.

[0015] Furthermore, the porosity gradient of the gradient composite scaffold is 20-200 μm, and the ratio of polycaprolactone (PCL) / microcrystalline cellulose (MCC) materials changes continuously from 30:70 in the inner layer to 80:20 in the outer layer.

[0016] Furthermore, the interlaminar shear strength of the gradient composite scaffold is ≥5 MPa, and the compression modulus gradient is 0.5 GPa-3.2 GPa.

[0017] By adopting the above technical solution, the beneficial effects of the present invention are as follows: through finite element analysis and verification, AI topology optimization makes the scaffold porosity and composition gradient match with natural bone by more than 90%, the mechanical distribution is more uniform, and stress concentration is avoided;

[0018] The dual-nozzle FDM (20Hz switching frequency) and electrospinning composite process increases the interlayer shear strength to 5.8MPa, solving the traditional interface stratification problem. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0020] Figure 1 It is the AI ​​topology optimization design flow chart of the present invention. DETAILED DESCRIPTION

[0021] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0022] like Figure 1 As shown, AI topology optimization design includes the following modules:

[0023] Input module: Input patient CT data, requiring a resolution of <100 μm, extract the defect area, input target mechanical parameters, target elastic modulus gradient 0.5 GPa-3.2 GPa, porosity range 20-200 μm;

[0024] Preprocessing module: Standardizes data through grayscale calibration and 3D reconstruction of CT data, implements mechanical constraint loading through boundary condition loading (force direction, load magnitude), and generates an initial 3D model based on the Marching Cubes algorithm;

[0025] AI Optimization Module: A topology optimization algorithm is obtained through convolutional neural network (CNN) training based on the TensorFlow framework. The training dataset includes CT and mechanical data of no less than 200 bone defects. A fractal geometry support algorithm is used to prevent overhang collapse. When the overhang angle is ≥45°, tree-like support generation is triggered, and the pore connectivity rate is greater than 95%. If the overhang angle is less than 45°, the system returns to the preprocessing module for iterative regeneration of the 3D model. The polycaprolactone (PCL) / microcrystalline cellulose (MCC) material ratio is required to continuously change from 30:70 in the inner layer to 80:20 in the outer layer, with a pore gradient change rate of 5% per layer.

[0026] Output module: Generates a pore / material gradient distribution model in STL format, and verifies it through finite element analysis. When the matching degree is greater than 90%, the model is qualified and can be exported. The output model can be directly imported into the 3D printing system for production.

[0027] Example:

[0028] ① Preparation of alveolar bone scaffold

[0029] Data acquisition: CT scan of the patient's jaw (50 μm slice thickness), identifying a defect area with a volume of 8.5 × 6.3 × 4.2 mm³;

[0030] AI optimization: Input the target elastic modulus gradient (0.8 GPa in the inner layer → 3.2 GPa in the outer layer) to generate an 84-layer pore / composition distribution model;

[0031] Printing Implementation: FDM printing of the base structure (material switching accuracy ±1.2%); electrospun fiber deposition in each layer (fiber deposition density 15g / m²);

[0032] Post-processing: vacuum drying (50°C, 24h) followed by gamma irradiation sterilization (25kGy).

[0033] ②Performance testing

[0034] Mechanical properties:

[0035] Compression modulus gradient 0.82→3.15GPa (deviation from design value <4.7%);

[0036] Interlaminar shear strength 5.8±0.3MPa (190% higher than pure FDM);

[0037] Biological properties:

[0038] The 7-day proliferation rate of MC3T3 cells was 218% (100% of the control group);

[0039] The expression level of the osteogenic gene Runx2 increased by 3.2 times (qPCR detection).

[0040] Animal clinical trial design plan:

[0041] 1. Purpose of the Test

[0042] Verify the effectiveness (promoting bone regeneration and restoring mechanical properties) and safety (biocompatibility and inflammatory response control) of gradient composite scaffolds based on AI topology optimization in bone defect repair.

[0043] 2. Experimental Design

[0044] 1. Animal Model Selection

[0045] Species: New Zealand White rabbits (adult, weighing 2.5-3.5 kg). The skull defect model is widely used in bone regeneration research and is easy to operate.

[0046] Group design:

[0047] Experimental group (n=15): AI-optimized gradient composite scaffolds (PCL / MCC ratio gradient change, porosity 20-200μm) were implanted.

[0048] Control group (n=15): implanted with traditional uniform porosity scaffolds (80% porosity, PCL single material).

[0049] Blank group (n=5): No material was implanted in the defect area (natural healing control).

[0050] Defective model:

[0051] Full-thickness bone defects with a diameter of 8 mm were created bilaterally on the midline of the skull (standardized drilling tools were used, with a controlled depth of 1.5 mm).

[0052] 2. Surgical Procedure

[0053] Anesthesia: Local anesthesia was performed with Zolpidem (5 mg / kg) combined with lidocaine.

[0054] Postoperative care: daily antibiotic injection (enrofloxacin, 5 mg / kg) for 3 consecutive days, pain management (butorphanol, 0.1 mg / kg).

[0055] 3. Observation indicators and time points

[0056] Imaging evaluation (Micro-CT):

[0057] Time points: 4 weeks, 8 weeks, and 12 weeks after surgery.

[0058] Parameters: new bone volume (BV / TV), bone mineral density (BMD), and scaffold degradation rate.

[0059] Histological analysis:

[0060] Time points: 8 and 12 weeks after surgery (5 animals were randomly selected from each group and sacrificed).

[0061] Methods: HE staining was used to evaluate bone tissue morphology, Masson staining was used to observe collagen deposition, and TRAP staining was used to detect osteoclast activity.

[0062] Biomechanical testing:

[0063] Time point: 12 weeks after surgery.

[0064] Parameters: compression modulus (MPa), interlaminar shear strength (MPa).

[0065] Inflammation and biocompatibility:

[0066] Serological tests: IL-6 and TNF-α levels (1 week and 4 weeks after surgery).

[0067] Histological evaluation: foreign body reaction, fibrous encapsulation thickness.

[0068] 4. Statistical Analysis

[0069] One-way analysis of variance (ANOVA) was performed using SPSS 26.0. The data were expressed as mean ± standard deviation, and the significance level was set at p < 0.05.

[0070] 3. Simulation test results

[0071] 1. Imaging evaluation (Micro-CT)

[0072] 12 weeks after surgery:

[0073] Experimental group: The new bone volume (BV / TV) reached 65±4%, which was significantly higher than that of the control group (42±5%) and the blank group (28±3%) (p<0.01).

[0074] Stent degradation rate: the stent residue in the experimental group was 15±2%, while that in the control group was 30±3% (p<0.05).

[0075] 2. Histological Analysis

[0076] 8 weeks after surgery:

[0077] Mature trabecular bone structure was observed in the experimental group, with orderly collagen arrangement (Masson staining score 4.2±0.3 vs. control group 2.8±0.5, p<0.01).

[0078] The blank group was mainly filled with fibrous tissue and had a low degree of ossification.

[0079] 12 weeks after surgery:

[0080] The osteoclast activity (TRAP+ cell number) in the experimental group was significantly lower than that in the control group (12±2 vs. 25±3, p<0.05), indicating balanced bone remodeling.

[0081] 3. Biomechanical properties

[0082] Compression modulus: The modulus of the repaired area in the experimental group was 2.8±0.3 GPa, which was close to that of natural bone (3.0 GPa) and significantly higher than that of the control group (1.5±0.2 GPa) (p<0.01).

[0083] Interlaminar shear strength: 5.5±0.4MPa in the experimental group and 2.1±0.3MPa in the control group (p<0.01).

[0084] 4. Inflammation and biocompatibility

[0085] Serum inflammatory factors: There was no significant difference in IL-6 levels between the experimental group (1 week after surgery: 35±5 pg / mL) and the control group (40±6 pg / mL) (p>0.05), and both returned to baseline levels 4 weeks after surgery.

[0086] Tissue reaction: The fiber wrapping thickness of the experimental group was less than 50 μm, and there was no significant foreign body giant cell aggregation, indicating good biocompatibility.

[0087] IV. Conclusion

[0088] The simulation test results show that the gradient composite scaffold based on AI topology optimization has the following advantages in bone defect repair:

[0089] Enhanced bone regeneration capacity: The volume of new bone is increased by 55% compared with traditional scaffolds, and the scaffold degradation rate matches the bone ingrowth rate.

[0090] Mechanical properties restored: compression modulus is close to that of natural bone, and interlayer bonding strength is increased by 160%.

[0091] Safety is controllable: the inflammatory response is mild and there is no long-term foreign body rejection.

[0092] Components not described in detail herein are prior art.

[0093] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A 3D printing process for gradient composite scaffolds based on AI topology optimization, characterized by: The following steps are involved: a) Based on patient medical imaging data, AI topology optimization was used to generate a digital model with a porosity gradient distribution of 20-200μm and a gradient change in the PCL / MCC material ratio. b) A dual-nozzle FDM system was used for gradient printing, with a layer thickness of 50μm, a material switching frequency of 20Hz, and a gradient resolution of 5% / layer. c) After each FDM printing layer, an electrospun nanofiber layer was deposited, with a fiber diameter of 500±50nm and a deposition density of 10-20g / m².

2. The 3D printing process for gradient composite scaffolds based on AI topology optimization according to claim 1, characterized in that: The AI ​​topology optimization adopts a fractal geometry support algorithm, which automatically generates a tree-like support structure when the overhang angle is ≥45°, and the pore connectivity is >95%.

3. The AI ​​topology optimization-based gradient composite scaffold 3D printing process according to claim 1, characterized in that: The process parameters of the dual-nozzle FDM system include: PCL nozzle temperature 75±2°C, extrusion pressure 0.3-0.5MPa; MCC nozzle temperature 110±3°C, extrusion pressure 0.8-1.2MPa; interlayer misalignment compensation algorithm controls position deviation <5μm.

4. The 3D printing process for gradient composite scaffolds based on AI topology optimization according to claim 1, characterized in that: The electrospinning process parameters are as follows: voltage 18 kV, receiving distance 15 cm, PCL / THF solution concentration 12 wt %, and fibers cross-laid at 45°.

5. A gradient composite scaffold 3D printing process based on AI topology optimization according to any one of claims 1 to 4, characterized in that: The porosity gradient of the gradient composite scaffold is 20-200 μm, and the PCL / MCC material ratio changes continuously from 30:70 in the inner layer to 80:20 in the outer layer.

6. The AI ​​topology optimization-based gradient composite scaffold 3D printing process according to claim 5, characterized in that: The interlaminar shear strength of the gradient composite scaffold is ≥5 MPa, and the compression modulus gradient is 0.5 GPa-3.2 GPa.

Images