Optimizing autoinjector devices using physics-based simulations and Gaussian processes
Source abstract
Abstract
Autoinjectors are becoming a primary drug delivery option to the subcutaneous space. These devices need to work robustly and autonomously to maximize drug bio-availability. However, current designs ignore the coupling between autoinjector dynamics and tissue biomechanics. Here we present a Bayesian framework for optimization of autoinjector devices that can account for the coupled autoinjector-tissue biomechanics and uncertainty in tissue mechanical behavior. The framework relies on replacing the high fidelity model of tissue insertion with a Gaussian process (GP). The GP model is accurate yet computationally affordable, enabling a thorough sensitivity analysis that identified tissue properties, which are not part of the autoinjector design space, as important variables for the injection process. Higher fracture toughness decreases the crack depth, while tissue shear modulus has the opposite effect. The sensitivity analysis also shows that drug viscosity and spring force, which are part of the design space, affect the location and timing of drug delivery. Low viscosity could lead to premature delivery, but can be prevented with smaller spring forces, while higher viscosity could prevent premature delivery while demanding larger spring forces and increasing the time of injection. Increasing the spring force guarantees penetration to the desired depth, but it can result in undesirably high accelerations. The Bayesian optimization framework tackles the challenge of designing devices with performance metrics coupled to uncertain tissue properties. This work is important for the design of other medical devices for which optimization in the presence of material behavior uncertainty is needed.
Source: europepmc · Quality: candidate unverified · Source record ↗
Research context
OpenAlex topics: Microfluidic and Bio-sensing Technologies · Viral Infectious Diseases and Gene Expression in Insects · Spectroscopy Techniques in Biomedical and Chemical Research
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (67)
Show source-provided references
1. Characterization of the anisotropic mechanical properties of excised human skin10.1016/j.jmbbm.2011.08.016 ↗
2. Estimation of the fracture toughness of soft tissue from needle insertion
3. Fracture mechanics model of needle cutting tissue10.1115/1.4030374 ↗
4. Feasibility of extracting tissue material properties via cohesive elements: a finite element approach to probe insertion procedures in non-invasive spine surgeries10.1007/s11517-021-02432-9 ↗
5. Prestress as an optimal biomechanical parameter for needle penetration10.1016/j.jbiomech.2012.01.049 ↗
6. Physics-informed neural networks (PINNs) for fluid mechanics: A review
7. Camanho, P.P., Dávila, C.G., 2002. Mixed-Mode Decohesion Finite Elements for the Simulation of Delamination in Composite Materials. Tech. Rep..
8. Effect of needle insertion speed on tissue injury, stress, and backflow distribution for convection-enhanced delivery in the rat brain10.1371/journal.pone.0094919 ↗
9. Dermatological complications of continuous subcutaneous insulin infusion in children and adolescents10.1016/j.jpeds.2007.10.006 ↗
10. Numerical aspects of cohesive-zone models10.1016/S0013-7944(03)00122-X ↗
11. Performance characterization of spring actuated autoinjector devices for Emgality and Aimovig10.1080/03007995.2020.1783219 ↗
12. Impact of cavitation, high shear stress and air/liquid interfaces on protein aggregation10.1002/biot.201800062 ↗
13. Bioavailability and cardiovascular effects of adrenaline administered by Anapen autoinjector in healthy volunteers10.1016/j.jaip.2017.09.021 ↗
14. Calculation of injection forces for highly concentrated protein solutions10.1016/j.ijpharm.2015.07.054 ↗
15. Fitzgibbon · 2017
16. Modeling the mechanical response of in vivo human skin under a rich set of deformations10.1007/s10439-011-0292-7 ↗
17. Food,, Administration, D., et al., 2013. Technical considerations for pen, jet, and related injectors intended for use with drugs and biological products. Md 20993 301–796.
18. GPy · 2012
19. Gramacy · 2020
20. Cases for the nugget in modeling computer experiments10.1007/s11222-010-9224-x ↗
21. A physics-informed deep learning framework for inversion and surrogate modeling in solid mechanics10.1016/j.cma.2021.113741 ↗
22. SALib: An open-source python library for sensitivity analysis10.21105/joss.00097 ↗
23. The absorption of subcutaneously injected short-acting soluble insulin: influence of injection technique and concentration10.2337/diacare.6.5.459 ↗
24. Systematic review of device parameters and design of studies bridging biologic-device combination products using prefilled syringes and autoinjectors10.1208/s12248-020-0433-8 ↗
25. The effect of skin tension, needle diameter and insertion velocity on the fracture properties of porcine tissue10.1016/j.jmbbm.2021.104660 ↗
26. Physics-informed machine learning10.1038/s42254-021-00314-5 ↗
27. Propagation of uncertainty in the mechanical and biological response of growing tissues using multi-fidelity Gaussian process regression10.1016/j.cma.2019.112724 ↗
28. Predicting the effect of aging and defect size on the stress profiles of skin from advancement, rotation and transposition flap surgeries10.1016/j.jmps.2019.01.012 ↗
29. A poro-viscoelastic model for the subcutaneous injection of monoclonal antibodies10.1016/j.jmps.2021.104537 ↗
30. Using poro-elasticity to model the large deformation of tissue during subcutaneous injection10.1016/j.cma.2021.113919 ↗
31. When Gaussian process meets big data: A review of scalable gps10.1109/TNNLS.2019.2957109 ↗
32. Skin physiology in men and women: in vivo evaluation of 300 people including TEWL, SC hydration, sebum content and skin surface pH10.1111/ics.12068 ↗
33. Mechanics of dynamic needle insertion into a biological material10.1109/TBME.2009.2036856 ↗
34. A finite element-based machine learning approach for modeling the mechanical behavior of the breast tissues under compression in real-time10.1016/j.compbiomed.2017.09.019 ↗
35. Physics-based deep neural network for real-time lesion tracking in ultrasound-guided breast biopsy
36. Fracture behaviour of human skin in deep needle insertion can be captured using validated cohesive zone finite-element method10.1016/j.compbiomed.2021.104982 ↗
37. Detailed finite element modelling of deep needle insertions into a soft tissue phantom using a cohesive approach10.1080/10255842.2011.628448 ↗
38. Predicting failure in soft tissue phantoms via modeling of non-predetermined tear progression
39. Cavitation onset caused by acceleration10.1073/pnas.1702502114 ↗
40. Scikit-learn: Machine learning in python
41. Ranking the fracture toughness of thin mammalian soft tissues using the scissors cutting test10.1016/S0021-9290(96)00101-7 ↗
42. Injection technique and pen needle design affect leakage from skin after subcutaneous injections10.1177/1932296815626723 ↗
43. Surrogate-based analysis and optimization10.1016/j.paerosci.2005.02.001 ↗
44. Transport and lymphatic uptake of monoclonal antibodies after subcutaneous injection10.1016/j.mvr.2021.104228 ↗
45. Gaussian processes in machine learning
46. Saltelli · 2004
47. User-centric approach to specifying technical attributes of drug delivery devices: Empirical study of autoinjector-cap removal forces10.2147/PPA.S298725 ↗
48. Mechanisms of deep penetration of soft solids, with application to the injection and wounding of skin10.1098/rspa.2004.1315 ↗
49. Shergold · 2005
50. The biomechanics of autoinjector—skin interactions during dynamic needle insertion10.1016/j.jbiomech.2022.110995 ↗
51. Single-model uncertainties for deep learning
52. An adaptive finite element model for steerable needles10.1007/s10237-020-01310-x ↗
53. Development of an advanced injection time model for an autoinjector
54. Full-field bulge test for planar anisotropic tissues: Part I–experimental methods applied to human skin tissue10.1016/j.actbio.2012.11.035 ↗
55. Deep uq: Learning deep neural network surrogate models for high dimensional uncertainty quantification10.1016/j.jcp.2018.08.036 ↗
56. Needle–tissue interaction forces–a survey of experimental data10.1016/j.medengphy.2012.04.007 ↗
57. Pressure and stress transients in autoinjector devices10.1007/s13346-018-0568-7 ↗
58. Scipy 1.0: fundamental algorithms for scientific computing in python10.1038/s41592-019-0686-2 ↗
59. Age dependence of mechanical and biochemical properties of human skin. I: Stress–strain experiments, skin thickness and biochemical analysis
60. Mathematical modeling for faster autoinjector design
61. Insulin leakage value in relation to pen needle length and administered dose after subcutaneous injection10.1089/dia.2010.0050 ↗
62. An overview of overfitting and its solutions
63. Modeling cavitation bubble dynamics in an autoinjector and its implications on drug molecules10.1016/j.ijpharm.2021.121062 ↗
64. Sobol sensitivity analysis: a tool to guide the development and evaluation of systems pharmacology models
65. A framework to optimize spring-driven autoinjectors
66. An experimentally validated dynamic model for spring-driven autoinjectors10.1016/j.ijpharm.2020.120008 ↗
67. Zhong · 2022