�鶹��madou

Many clinicians crave a crystal ball to foresee how a treatment will go for a patient, but Dr Luca Modenese is developing something much more useful – a digital twin of a patient, which health practitioners can use to tailor physical interventions and optimise surgeries for that individual.

Working in computational biomechanics, Luca has devised tools that extract data from MRIs and CT scans to automatically create a personalised digital model of a patient in minutes. This digital model exactly reproduces an individual’s musculoskeletal system and the unique way they move. 

A model of how brain, muscle and bone interact

There are two levels to the research. “The first is investigating how our brain controls our muscles to move and generate a certain amount of force, not more, not less,” explains Luca, who initially trained in mechanical engineering and now lectures in biomechanics at �鶹��madou’s School of Biomedical Engineering. 

He can use data from an MRI or other scan to build an exact model of an individual’s musculoskeletal anatomy, then add information on the electrical activity measured in a patient’s muscles when they contract.

“We can simulate exactly what the brain and the muscles are doing,” he says. 

The other aspect of his work is using the digital twin to assist in surgical planning and physiotherapy programs. “We can use generative AI [artificial intelligence] techniques to generate lots of realistic motions,” he explains.

Musculoskeletal conditions are the , according to the World Health Organization. But humans come in a variety of shapes and move in different ways. An intervention or implant that works very well for one person may be less successful for another.   

With the personalised biomechanical model, clinicians can input and adjust different interventions and identify which produces the best outcome. That means they can be more confident of a positive result before any brace, physiotherapy or surgery is applied to the patient.

“If you can represent what the brain is doing, the anatomy and condition of the muscles and the bones, and how they interact with each other, it is really possible to tell the surgeon, ‘If you do that on that patient it is not going to work very well. If you do it on this other one, it is going to work much better’,” Luca says.

Researchers are also using Luca’s patient-specific digital modelling to help develop exoskeletons and bionics.

Using the digital twin as a diagnostic tool 

In 2024, Luca started a collaboration with Griffith University researchers to collect and model data on children with cerebral palsy. is diagnosed with the disorder which, in most cases, is caused by an injury to the developing brain. It affects the child’s muscle control and they may require surgical intervention for bone deformities and to restore muscle balance as they grow. 

Luca is modelling causes and possible therapeutic options for this condition. His focus is on refining his digital model to the point that he can isolate whether an individual’s issue with walking is due to the brain sending the wrong signal, an impairment of the muscles, or a bone deformity. That will then guide what therapy is used. 

Making personal digital models more accessible

Luca’s work is effectively at the interface between mechanical engineering, quantitative anatomy, muscle physiology, neuroscience and computer science. However, for clinicians to use the digital model, it has to be made as simple as possible.

“I developed these tools to create the models automatically – that is a step forward. But the technologies I’m developing won’t have clinical uptake until they can be used smoothly by non-technical people.”