Jérôme Vétiillard, VP Research and Development (R&D) at Qualees, explains how digital twins are transforming health methodology and practices to improve population care.
What is a digital twin?
Jérôme Vétiillard : A digital twin is a dynamic digital model representing a physical, biological system or a set of processes constructed from the data produced by this system. It allows operation to be simulated to optimize performance under constraints, whether for industrial, clinical or financial applications. For example, it allows the predictive maintenance of industrial equipment or the simulation of a patient population in a clinical trial.
In summary, the digital twin is a sophisticated simulation tool that anticipates the behavior of a real system, thus providing preventative/predictive and optimized solutions.
And in the health sector?
JV: In healthcare, digital twins require high-quality data to faithfully represent the system being studied. They can take several forms, such as 3D modeling of a joint to determine the best prosthesis to implant (including beyond 3D modeling, biocompatibility data, material resistance, biomechanical calculations, etc.), or representation molecular of a macromolecule to evaluate the effectiveness of monoclonal antibodies to block certain receptors on the surface of cancer cells for example (such as anti-DLP-1 antibodies). They also optimize patient care pathways and could improve real-time clinical supervision, particularly in intensive care units (Internet of Medical Things, digital twins, real-time health data).
Can digital twins help with diagnosis?
JV: Absolutely. In radiology, for example, artificial intelligence uses digital models of tumors to compare with patient imaging data and make accurate diagnoses. More sophisticated models integrate various clinical, biological, proteomic and genomic data to offer a multimodal and precise vision of tumors, allowing beyond diagnosis, to choose the best treatments and provide reliable prognoses.
What are the advantages of digital twins in health and in particular of your specific approach to the digital twin generator?
JV : Cost savings and faster benefits for patients. Digital twins significantly reduce the cost and time required to generate experimental data, allowing thousands of scenarios to be tested quickly and inexpensively.
At Qualees, we have opted for “ultra-narrow” AI, generating specific, reliable, explainable and economically affordable digital twins, with a low ecological footprint: TweenMe is in fact the first “universal” generator of digital twins. These digital techniques have a real impact: during the Covid-19 crisis, they made it possible to divide the development time of new vaccines by five while securing their development.
What are the prospects for digital twins?
JV: Digital twins have immense potential in many fields, from healthcare to manufacturing to finance. Their development depends on the quality and accessibility of data, as well as compliance with ethical and legal regulations, such as GDPR and the AI Act.
At Qualees, we are working to structure this data and optimize AI algorithms to take advantage of HDLSS (High Dimension Low Sample Size) data which is characteristic of health data. HDLSS data poses particular constraints on the training of artificial intelligence algorithms that the design of TweenMe allows us to overcome.
Digital twins must find their economic model and be accessible to healthcare professionals, even without expertise in data science. Our TweenMe platform aims to make these tools affordable, accurate and easy to use.
Finally, the certification of digital twins is crucial to enable their daily clinical use. TweenMe combines the best of generative artificial intelligence with strict statistical quality control to avoid any errors or “hallucinations”, guarantee the reliability of the digital twins produced and facilitate their certification by regulatory authorities.
Digital twins represent a major breakthrough in healthcare, combining cost savings and rapid benefits for patients, while opening new perspectives for various applications. They are the key to the medicine of tomorrow, more precise, preventive and personalized.
Comments collected by Alexandre Terrini
A webinar to learn more about digital twins
To find out more, meet on June 25 from 1 to 2 p.m. for a webinar on digital twins open to all health stakeholders, whether specialists or not on the subject.
Registrations
){kind=link}