Research Group web site

My work at Imperial, at the Institute of Biomedical Engineering, started on the Retinal Prosthesis (RP) project. RP is a device which should restore vision to blind people. Efforts to build a RP tried to mimic success of the cochlear prosthesis which is able to restore a functional level of hearing to profoundly deaf people. However a similar success with RP was and still seems elusive. A number of expert groups around the world have been trying with a lot of investment over 25 years and still a limited progress. Why? Why can’t we just stimulate parts of the vision system and recreate visual perception in the brain of a blind person. Answer to this question put me on this path of research in a quest for understanding operational principles of biological systems which they use for sensing environment and performing biological computation.

The efficient information encoding and computation in the presence of noise is the basis of the sensory transduction and neural architecture of all living organisms. In order to better understand these processes I started my work at the junction of biology and physics and engineering. Starting with the retina I tried to understand the processes of phototransduction (or signal transduction in general), information encoding in neural activity and neural processing of this information. One of the primary tools in this area is optogenetics, where I made some pioneering contributions in modeling opsins. This should help us not only in designing RP but also in the general pursuit of building bio-inspired and neuromorphic devices and systems which capture properties of biological neural computation into analogue and mixed-signal VLSI. They emulate the efficient computational capabilities of wetware. 

Living cells, like small analogue computers, receive information from their surroundings, process that information and communicate between themselves. They perform the information processing tasks using noisy components in noisy environment and with limited energy resources, but still they achieve very complex computation. This is achieved by using very sophisticated molecular machinery in the form of various membrane proteins, phospholipids, neurotransmitters etc.

I have been working on creating mathematical models and simulating some important but complex biochemical processes in living organisms, such as light induced neurostimulation, the cell signal transduction mechanisms (G-protein coupled cascade – GPCC), etc. The insights provided by the modelling represent the fundamental basis in understanding how to design new generations of biomedical devices, such as advanced neuroprostheses, or transistors based devices to emulate neural computation.

My main research interest in the last few years was in the areas of mathematical biology and bio-inspired technologies. 

Optogenetics

• Photo-cycle models of channelrhodopsin2 (ChR2)
• Modelling of neurons expressing ChR2 mutants/other ion pumps
• Modulation of neural activity using light
• Optogenetic retinal implant (developing camera/processing/LED driver/LED array system)

Retinal circuitry

• Information Theory used to decipher the Receptive Field Vectors of various RGCs
•  Linear-Nonlinear-Poisson models for retina ganglion cell response
• Bio-inspired circuits and systems based on the retina circuitry

Stochastic Models of Phototransduction and G-protein coupled cascades

• Modelling of the TRP ion-channels activity
• System and Circuit models of phototransduction and enzymatic cascades in general
• For example: Sensory transduction channels (for smell, vision, hearing, touch and taste) are based on G-protein coupled pathways (GPCP). GPCP represent a set of enzymatic amplifiers which have an end effect of modulating ion channels permeability and hence modifying the electric (i.e. ion) current in and out of the cell. The whole mechanism can be represented with a cascade of analogue amplifiers (characterised by gain, noise, bandwidth, power, etc) and appropriately analysed.

Thermal effects in neural activity

• stimulated by external electric fields (COMSOL & NEURON)
• thermo-electric modulation using MIR laser illumination

Silicon Photonics: Optical modulator in unmodified CMOS

Nanoelectronics

• devices, circuits and architectures 
• fault tolerance
• computational models (digital, analogue, biological, quantum)
• self assembly of nanoparticles