Multi-modal electro-optical systems for healthcare applications

Abstract

In this dissertation, two types of multi-modal electro-optical systems based on three ICs are proposed.

The first IC is designed for hybrid integration of the CMOS and organic optical sensors for PPG signal monitoring as a sticker-type system. The P-FCB technology is adopted for connecting the CMOS IC with organic optical sensors. The proposed IC is implemented with 0.18 μm CMOS technology and five features are provided in the hybrid system for continuous PPG signal monitoring. 1) Power-efficient low-molecular weight based OLEDs are adopted as the light emitting elements. 2) In the CMOS IC, an analog front-end processor and a digital processor are integrated to achieve a small form-factor, and low-noise system. 3) The natural degradation of the organic optical sensors is compensated with the CMOS IC. 4) The large parasitic elements of the organic sensors are considered in the analog front-end design. 5) Motion artifact is rejected with the contact impedance sensor and cascaded adaptive filters.

The second IC is implemented as clock synchronized multi-node WBAN transceiver for biomedical system. The 240 MHz clock signal is broadcasted from the hub IC and synchronized in every node IC. Multi-node ICs construct the channel environment-adapted ad-hoc network with the SR-RSSI scheme. So that, the 72 % power is reduced even in the harsh channel condition (<-50 dB attenuation). The proposed IC is implemented with 0.18 μm CMOS technology with consumes 274 μW in node IC and 2.8 mW in hub IC, respectively.

By combining first and second ICs, multi-node sticker-type $SpO_2$/ExG monitoring system is hybrid integrated on the convenient/flexible PET substrate demonstrates small area (5.5cm x 2.5cm), light weight (2g including batteries) and the lowest power consumption (141 μW) in the comparison table. In the proposed system, the OLED driver and OPD sensor readout with optical calibration loop measures $SpO_2$ values and ExG signal from the various body parts while dissipating only 141 μW. The measured $SpO_2$/ExG data from the SN-SoCs is transferred to the hub-SoC through the BCC transceiver. In addition, the clock transmitter in the hub-SoC broadcasts the clock of 240 MHz to the SN-SoCs to remove the bulky off-chip crystal. Accurate $SpO_2$/ExG data were successfully acquired from the 20 human subjects and the extracted $SpO_2$ results’ maximum standard error with 1000 samples is 1.13 compared with commercial reference device.

The third IC is designed for patch-type optical and electrical tomographic imaging system for skin disease diagnosis. The proposed IC is fabricated in 0.18 μm high voltage CMOS process with 7.54 mW average power consumption. From 2.8 ~ 4.2 V input voltage converts to 1.8 V for VDD and to 7 V for $V_{LED}$. CDMA DOT current driver operates with 16 MHz clock frequency and 3.9 kHz chip rate to achieve 7.0 fps hemodynamics imaging speed. Current range is wide from 0.05 mA to 6.4 mA with a 0.05 mA step. DOT receiver achieves fast settling time of DC rejection (50 μs) and wide dynamic range (100.4 dB). Weaver demodulator based EIT receiver operates with I/Q mismatch calibrator within current range of 0.1 to 1 mA. As a result,0.5 mm sensitivity and 2 mm selectivity dual tomographic imaging can be obtained for skin disease diagnosis.