The hybrid modelling (HM) of the neuronal stimulation is a revolutionizing tool in neuroprosthetics. It consists of merging the electrical potential distribution induced by stimulating electrodes in a biological structure (FEM) with the neuronal spiking dynamics to predict neural responses (using NEURON software). Previously, HM has been successfully implemented and exploited for DBS, VNS and PNS, but has never been implemented for Intracortical microstimulation (ICMS) in humans. ICMS has shown to be able to restore the sense of touch and vision in iBCI applications and it can be used for many other applications requiring to precisely activate neural structures (micro-activations). However, ICMS inherently encompasses a vast number of parameters like electrode geometry, implant location and stimulation parameters. This creates a multidimensional space to explore that, changing each of the parameters constituting this space, results in a drastically different outcome. Currently, most of these parameters are tested and optimised through animal studies in trial-and-error or brute-force approaches. To find the optimal parameters, an efficient and elegant way is to exploit the power of neuro-AI that can find the optimal solution in a faster and a more efficient way, while limiting and focalizing the animal testing. HM could be exploited to design novel electrodes (different active sites size and geometry), implant surgical locations (number of electrodes and channels), stimulation policies (spatiotemporal parameters modulation), as well as the impact of different brain structures and anatomies to accurately evaluate the effect of specific ICMS patterning. Indeed, HM can support in the process of surgery planning and create personalized implantation. Additionally, different stimulation strategies can be tested to optimize the effects. This novel highly-detailed HM framework could drastically change the impact of the ICMS. Yet another aspect of implanted electrodes is the capabilities in neural recording. HM can be used to simulate neural recording abilities and the extent which the neural activities can be recorded. The ultimate goal is to create an in-silico brain for neural interfacing testing and possibly using it in the future for validation and verifications of electrodes. Applying AI-based algorithms on neuronal recordings after stimulation and combining them with behavioural data would guarantee a complete validation of the HM framework. The aims of this project can be summarized as follows: 1. Modelling a biophysically plausible cortical structure using FEM 2. Modelling the electrical interaction of the implants with the brain tissue . Furthermore, the effect of fibrotic encapsulation around the electrode (8), mechanical strain and blood flow would be studied. 3. Modelling the neuronal activation followed by purposely-designed stimulations exploiting powerful clusters (available at Chalmers) and neuro-AI. 3. Optimizing the ICMS patterning and personalizing implantation location for each specific subject. 4. Study the effect of the geometry of the electrodes in creating the electrical currents and neuron activation in different cortical layers. 5. Using AI algorithm to find detection threshold from computed data to validate the models and fill the gap between computational and behavioural models.