This video illustrates some examples of applications.
This video illustrates some examples of artificial intelligence applications in the development of personalized medicine in oncology (imaging, digital twins, etc.).
Feel free to also watch our videos on the concept of dynamic consent and on artificial intelligence and ethics in oncology.
In the fight against cancer, advances in AI are fueling many hopes.
But beyond the media hype, how will this technology promote personalized medicine? Actually, the term Artificial Intelligence encompasses a vast number of techniques. Among these, let’s take the example of artificial neural networks, which have garnered much attention in recent years. The functioning of an artificial neural network is inspired by the way our brains work. Its purpose is to answer a question based on data provided to it. For this, the network must learn from a set of example cases. For each case, the AI knows the data input into the network and the answers it should provide. The algorithm then adjusts the weight of each synapse to determine a model that would fit the set of cases.
Once out of the learning phase, the neural network can provide answers for new cases presented to it. The neural network is a good example of what AI enables: it can process large amounts of data while identifying patterns that a human might not discern. Against cancer, this is invaluable at a time when the explosion of data volumes presents an ever-greater challenge. Imaging, clinical and omics data, mobile applications; the amount of information for a single patient now exceeds a terabyte. AI offers hope in processing all this data and providing answers that take into account every dimension of the patient. This is crucial to accelerate the revolution in personalized medicine, helping us to prevent, diagnose, treat, and discover better.
In prevention, AI can help identify at-risk individuals and better detect precancerous cells. Diagnosis is accelerated by AI’s ability to identify patterns in imaging. During treatment, AI assists in decision-making: some AI systems even promise to predict a patient’s therapeutic response to immunotherapy. Finally, several research teams are using AI for drug discovery, some even envisioning creating new molecules through AI.
While all these applications contribute to the development of personalized medicine, the emerging concept of the digital twin fully illustrates AI’s potential.
Indeed, all of the patient’s data makes possible to create a virtual alter ego. The physician can evaluate different therapies using simulations based on AI and data from other digital twins similar to the patient.
The strategy ultimately adopted in real life and its outcomes, in turn, feed back into the system, refining simulations for other digital twins.
By enabling a transdisciplinary approach that considers all of a patient’s specific traits across all dimensions, AI has the potential to become a decisive asset in advancing personalized medicine against cancer