Research

My lab is interested in understanding the design and computing principles of biological sensory systems and translating this knowledge into bio-inspired intelligent systems and machine learning algorithms. Research in my lab involves two main themes:

  • Systems Neuroscience: an investigation that will combine computational and electrophysiological approaches to examine fundamental principles of olfactory coding and signal processing, learning, and memory
  • Neuromorphic Engineering: development of novel, bio-inspired devices (e.g. ‘electronic nose’) and algorithmic tools for non-invasive medical diagnosis and homeland security applications

Biological Olfaction

A schematic of the early olfactory circuits in the locust brain

How does the nervous system convert sensory stimuli into neural representations? Odorants are detected by a large population of olfactory receptor neurons (ORNs), which convert chemical stimuli into an electrical signal that is relayed downstream for further processing in the olfactory bulb (OB, vertebrates) or the antennal lobe (AL, insects). In the AL/OB, interactions between ensembles of excitatory principal neurons and inhibitory local neurons reshape the ORN input into complex, slow spatio-temporal patterns that are superimposed on a faster oscillatory field potential activity. The patterned AL/OB responses contain information about odor identity and intensity and form the only odor representation available to the organism. Hence they are considered to be the ‘odor code’. The generated odor code is then transmitted to learning and memory centers.

Related Publications:

Artificial Olfaction

An electronic nose
MEMS-microsensor array for sensing toxic industrial chemicals

An electronic nose is an instrument that combines an array of cross-selective chemical sensors and a pattern recognition engine to recognize chemical species. We employ statistical and bio-inspired signal approaches to design & operate MEMS-based chemiresistive microsensor arrays and tune them for specific chemical sensing application

Target applications for electronic nose include:

  • Medical Diagnostics (Breath Analysis)
  • Homeland Security

Related Publications:

Insect-based brain machine interface for bio-hybrid Olfaction

Insect-based biorobots

The ability of biological olfaction far outshines the capabilities attained by artificial analogues. The canine olfactory system still remains the state-of-art sensing system for many engineering applications, including, homeland security and medical diagnosis.

Given that even simple biological organisms possess superior sensing capabilities than the state-of-art engineering technologies, could the capabilities of biological organisms be exploited for the purpose of creating a novel class of bio-hybrid sensing systems? Recent work has shown that insects have chemical sensors for a range of applications including explosive chemical species, cancer, and other biomarkers. While these efforts provide proof-of-concept results further validation of biological system’s capabilities, and whether and how robust insect-machine interfaces can be made for the purpose of remote-controlled sensing remain open problems.

Related Publications: