Advancement: Enhanced Detection and Identification of Bioparticles by Smart Signal Processing Techniques

Speaker Name: 
Vahid Ganjalizadeh
Speaker Title: 
PhD Candidate (Advisor: Professor Holger Schmidt)
Speaker Organization: 
Electrical & Computer Engineering
Start Time: 
Monday, September 9, 2019 - 3:00pm
End Time: 
Monday, September 9, 2019 - 5:00pm
Engineering 2, Room 215
Professor Holger Schmidt

Detection of bioparticles such as proteins, viruses, nucleic acids and other biomolecules is important for both research and many biomedical applications. Optofluidics as a relatively new field, integrates the photonics into the fluidic media to take advantage of the high sensitivity of photonics while maintaining the flexibilities of microfluidics. Chip-based optofluidic devices are, therefore, a promising approach for integrated biophotonics and point-of-care devices. The idea behind point-of-care devices is to be compact and cheap and simple to use. However, these constraints can affect the signal quality and sensitivity of device indeed. There have been many studies and improvements in fabrication processes to make devices with higher SNR, but still usually there are many targets not detected or identified correctly. As one example, I will show how three-dimensional hydrodynamic focusing of the particle stream improves the uniformity of the optical signal.

In addition, I will describe new approaches that explore new signal processing techniques which can aid us to improve sensitivity and robustness of the optofluidic devices in single and multiplexed biomolecular detection. One path is to combine wavelet analysis with machine learning algorithms to dramatically improve the detection rate and accuracy. A second avenue is the use of time-division multiplexing to enable detection of molecular analytes over a wide concentration range, covering both digital and analog regimes. I will present preliminary results for both approaches and will discuss plans to optimize their implementation.