FEIT Research Project Database

Biopolymers 4.0: Machine-learning-accelerated multiscale computational biopolymer design and experimental development of an archetypical intraocular device

Project Leader: Ellie Hajizadeh
Collaborators: Molecular Simulations and Drug Delivery Research Group, Department of Chemical Engineering, IIT Roorkee, India. Biomaterials and Multiscale Mechanics Lab (BMML), Department of Metallurgical and Materials Engineering, IIT Roorkee, India.
Primary Contact: Ellie Hajizadeh (ellie.hajizadeh@unimelb.edu.au)
Keywords: 3D printing; artificial intelligence; biopolymers; computational materials science; fluid dynamics
Disciplines: Biomedical Engineering,Chemical & Biomolecular Engineering,Mechanical Engineering

Materials 4.0, combining machine learning with experimental (1.0), theoretical (2.0), and computational (3.0) materials science, brings a paradigm shift to the discovery and development of novel materials through: a) accelerating the pace of exploring materials design space and b) enabling an inverse design approach contrary to the conventional trial-and-error forward design methodologies, where desired material properties are explicitly targeted, and the parameters necessary to achieve them are found by a way of solving a constrained optimization problem. We aim to adopt Materials 4.0 approach for autonomous self-exploratory computational design and experimental development of novel biopolymers for drug-loaded intraocular lens (IoL).

Despite being one of the most common and safe surgical procedure, cataract surgery is not risk free. Ocular inflammation and posterior capsule opacification (PCO) are most common post-surgery complications, whereprophylaxis for post-cataract surgery complications suffers from bioavailability.

On the other hand, Glaucoma can cause irreversible blindness, where conditions associated with Glaucoma are usually managed with the use of eye drops (Beta blockers), laser treatment, surgery, or combinations of all three, which involve cost, inconvenience, and associated complications.

The proposed work will focus on loading anti-metabolites and anti-glaucoma drugs in the novel IoLs based on biopolymers. While the anti-metabolites would address PCO complications, anti-glaucoma drugs will be useful for patients with coincidence of cataract and glaucoma.

The advantage of drug loaded IoLs as a treatment of choice lies in the fact that after implantation, they do not require further compliance from the patient. Furthermore, minimum inhibitory concentrations of the drugs can be maintained for longer durations contrary to eye drops. Moreover, eye drops might interfere with IoLs and may not be as effective. Thus, after IoL implantation, eye drops for treating glaucoma may be impracticable and making the need of drug loaded IoLs imperative.

The challenging objective in developing such devices based on biopolymers is to achieve a sustained release of drug at the therapeutic levels for a prolonged time along with considerations for the expiry of the loaded drug. The solution to this daunting requiremnet lies in an optimised chemistry and molecular architecture of the biopolymers for IoLs fabrication.

Therefore, the aim is (a) to perform multiscale materials simulation to accurately predict the transport phenomena ie, controlled drug release in biopolymers and b) use machine learning (ML) and data informatics to establish a mathematical link between biopolymer chemical and architectural features and controlled and sustained drug release mechanism. (a) & (b) form the so-called forward prediction. In the so-called inverse design, physics-informed ML models will be coupled to a statistical optimization algorithm to find the optimal polymer chemistry and architecture corresponding to the desired controlled drug release over 5 years. c) The experimental characterisations of the 3D printed biopolymers will be used both for validation of the multiscale simulation models and in training and testing ML algorithms.