Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of Graph Search.
Diabetes is a global epidemic with millions of people dying from it every year. Progress in microelectronics and machine learning has led to a huge number of FDA approvals of devices for continuous monitoring and treatment of diabetes with one product using light in its sensing principle. Invitro demonstrators were designed, build and tested successfully for monitoring the genetically engineered INS-1E F10 cell line that works on the FRET principle based on a ratio-metric principle. The advantage of the ratio-metric probe is that we directly see the relative changes in the glucose levels of the stimulated INS-1E F10 cells, the common effects are canceled out in the ratio to a great extent. This first part of the work reports the design and testing of a bio-electronics system that measures cytosolic [Ca2+] in-vitro in the INS-1E pancreatic b cells using a 430 nm excitation light with the excitation wavelength of the probe being at 435 nm. We inserted a plasmid in the cell that is responsible for the FRET based [Ca2+] protein sensor complex, that emit light at both 485 nm and 535 nm. The system has a linear response upon characterization with flurosphere for various intensities of light. The system is sensitive for 2D cell cultures on normal petri dishes but suffers from poor SNR. Cell culturing to form 3D cell aggregates using a shaker in the incubator increases the number of cells and thus enhances the signal quality 35 times and 21 time for green and blue light counts respectively. The first system has been able to continuously monitor the fluorescence of a clone of INS-1E b-cells upon glucose stimulation, diazoxide injection and KCl injection, thus revealing insights in the cell response mechanisms. The second in-vitro system has been developed, tested and deployed. The angle independent absorption-based optical long pass filter MIDOPT LP470 improves the signal-to-noise ratio (SNR) by rejecting the reflections from the excitation source and auto-fluorescence components. A significant 480% improvement in witnessed in the SNR of the F10 cells after using the LP470 filter. The system has successfully monitored the genetically engineered INS-1E pancreatic b-cell line upon stimulation with 11 mM glucose in the cell culture medium and exchanging with KRBS solution with 0 mM glucose, which corresponds to a pulsating binary glucose stimulation. The system has been used to characterize the total number of cells in millions, cell growth and death, effect of temperature and step-wise, continuous and binary glucose stimulation. Insights about the noise floor and saturation with respect to the LED bias current and integration times were obtained from both the electronics and biological perspectives. While too faint light intensity and short photo-detector integration times drives us down in the noise floor, the opposite on the contrary has a deleterious effect on the cells as it leads to a loss of fluorescence. An analog front end readout amplifier for monitoring glucose was designed and simulated with UMC CMOS 180 nm technology. Detailed small signal analysis of the circuits were done and the design was tested with simulations using the cadence tool. The results were highly correlated and showed a promising step towards a low noise and low power analog front end circuit aimed at bio-medical applications.
Loïc Dayon, Andreas Wiederkehr
Vassily Hatzimanikatis, Ljubisa Miskovic, Michaël Roger Germain Moret