Additive manufacturing of elastomeric PDMS materials with tunable longitudinal and transverse relaxation times for MRI phantom applications in MRI guided radiotherapy

Abstract

Magnetic resonance imaging (MRI) is a non-invasive, non-ionising medical imaging technique that can be used for diagnosis and cancer treatment planning to ensure safe and successful image guided radiotherapy. To assess an MRI scanners ability to produce accurate images, a test object, or phantom, with known magnetic properties; such as T1 and T2 can be used. Most current phantom approaches consist of liquids and gels contained in vessels with T1 and T2 relaxation times that correspond to the relaxation times of a tissue or tissues being simulated. While these phantoms are radiologically relevant, they fail to simulate the heterogeneous and dynamic nature of human tissues. Additive manufacturing has been used to fabricate anthropomorphic medical imaging phantoms and despite improved anatomical accuracy, there remains a lack of flexible, soft tissue mimicking materials for simulating clinical scenarios where anatomical motion is important. Recently, extrusion additive manufacturing technologies, such as direct ink writing (DIW), have been explored to produce flexible anatomical models for soft robotic applications and therefore has potential to be applied in producing elastomeric MRI phantoms. The challenge is to develop material compositions that: (1) achieve the desired T1 and T2 relaxation values, (2) possess optimum pre-cure viscoelasticity for printability and (3) maintain post-cure mechanical properties comparable to those human tissues. This PhD thesis aims to contribute towards overcoming these challenges by presenting formulation development work and characterisation methodologies to produce printable MR-relevant elastomeric silicone materials and assess their performance as MRI phantoms. A printable silicone formulation, made up of high viscosity and low viscosity polydimethylsiloxanes (PDMS) was blended with silicone oil and characterised for T1 and T2 on 298 MHz NMR system. The amount of silicone oil added ranged between 5-50 % of the total phantom material. As silicone oil was increased, the T1 relaxation range shifted from 1158 (± 3) - 1359 (± 1) ms to 1323 (± 2) – 1742 (± 12) ms and the T2 relaxation range from 13 (±1) – 22 (± 3) ms to 31 (± 4)- 345 (± 11) ms. Adding silicone oil to silicone elastomer produced formulations which covered a broad range of clinically relevant relaxation values for tissues including grey and white matter in the brain. The effect of viscosity and degree of crosslinking on T1 and T2 was also studied by altering the ratio of high viscosity PDMS to low viscosity PDMS and altering the ratio of catalyst to PDMS. While these approaches produced variations in T1 and T2, the variations seen in formulations blended in with silicone oil were much greater. Oscillatory and rotational rheological tests assessed the printability of the silicone phantom material as the weight percentage of silicone oil was increased. My uncured materials exhibited yield stress ranges of 24 -131 Pa where yield strength decreased as the content of silicone oil increased. The same materials produced a linear viscoelastic region (LVER) range of 0.0001-0.0015 % indicating printability through a range of relaxation times. The PDMS-silicone oil materials were used to produce two-phase cubic brain phantom models via direct ink writing (DIW) extrusion 3D printing and casting. Both phantoms were imaged and the portion with silicone oil appeared brighter under T2 weighted imaging and displayed T1 and T2 relaxation values of 1195 (± 7) and 60 (± 4) respectively. To date and to the author’s knowledge, this is the first attempt at manufacturing elastomeric MRI phantoms and phantom formulations with tunable T1 and T2 relaxation times. The present NMR and rheological studies indicate that through a range of T1 and T2 times the material remained within a printable range; offering potential to 3D print realistic human mimics. The availability of soft anthropomorphic phantoms would be a significant advancement in developing magnetic resonance imaging sequences for guided cancer radiotherapy treatments in areas where motion is inevitable.

Date of Award	25 Oct 2024
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Benjamin Rowland (Co Supervisor), Brian Derby (Co Supervisor), James O'Connor (Co Supervisor), Ben Dickie (Co Supervisor) & Stephen Edmondson (Main Supervisor)