Loading...

Improving the Accuracy of a Microparticle Biosensor by Artificial Intelligence

Ghassab, Hamid Reza | 2023

34 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 56350 (08)
  4. University: Sharif University of Technology
  5. Department: Mechanical Engineering
  6. Advisor(s): Taghipoor, Mojtaba
  7. Abstract:
  8. In this research, the ability of artificial intelligence to improve the biosensor performance of a microfluidic system has been investigated. Coulter counter is a microfluidic system that measures the concentration of particles in a fluid using signals obtained from a biosensor. The method of this system is called " Resistive Pulse Sensing (RPS)" method. The pulses in the Coulter counter signals are affected by the number, shape, size and speed of particles passing through the orifice. The pulse of two particles of the same size and different shape in these systems are very similar and this makes it difficult for an operator to distinguish the type of particle. Another disadvantage of microfluidic coulter counters is the noise of the obtained signals. Artificial intelligence systems have different capabilities, including representing and extracting signal features. Both of these abilities can be used to eliminate the disadvantages of the RPS method. In this research, using artificial neural networks, systems have been created to reduce the noise of a microfluidic coulter counter. On average, the fluctuation amplitude and fluctuation intensity index of the noised signal by the artificial intelligence method were 0.45 and 0.33 compared to the wavelet method. By using the features extraction ability of artificial intelligence, systems have been created that have the ability to detect the type of particles from shape of signals. The particle identification system introduced in this report has distinguished yeast and polystyrene particles of the same size with 98.7% accuracy
  9. Keywords:
  10. Lab-on-a-Chip ; Microfluidic System ; Artificial Intelligence ; Noise Reduction ; Signal Processing ; Biosensor

 Digital Object List

 Bookmark

No TOC