A Smartphone-Based Biosensor for Non-Invasive Monitoring of Total Hemoglobin Concentration in Humans with High Accuracy
Zhipeng Fan,
Yong Zhou,
Haoyu Zhai,
Qi Wang,
Honghui He
Affiliations
Zhipeng Fan
Guangdong Research Center of Polarization Imaging and Measurement Engineering Technology, Shenzhen Key Laboratory for Minimal Invasive Medical Technologies, Institute of Biopharmaceutical and Health Engineering, Tsinghua Shenzhen International Graduate School, Tsinghua University, Shenzhen 518055, China
Yong Zhou
Shenzhen Maidu Technology Co., Ltd., Shenzhen 518000, China
Haoyu Zhai
Guangdong Research Center of Polarization Imaging and Measurement Engineering Technology, Shenzhen Key Laboratory for Minimal Invasive Medical Technologies, Institute of Biopharmaceutical and Health Engineering, Tsinghua Shenzhen International Graduate School, Tsinghua University, Shenzhen 518055, China
Qi Wang
Guangdong Research Center of Polarization Imaging and Measurement Engineering Technology, Shenzhen Key Laboratory for Minimal Invasive Medical Technologies, Institute of Biopharmaceutical and Health Engineering, Tsinghua Shenzhen International Graduate School, Tsinghua University, Shenzhen 518055, China
Honghui He
Guangdong Research Center of Polarization Imaging and Measurement Engineering Technology, Shenzhen Key Laboratory for Minimal Invasive Medical Technologies, Institute of Biopharmaceutical and Health Engineering, Tsinghua Shenzhen International Graduate School, Tsinghua University, Shenzhen 518055, China
In this paper, we propose a smartphone-based biosensor for detecting human total hemoglobin concentration in vivo with high accuracy. Compared to the existing biosensors used to measure hemoglobin concentration, the smartphone-based sensor utilizes the camera, memory, and computing power of the phone. Thus, the cost is largely reduced. Compared to existing smartphone-based sensors, we developed a highly integrated multi-wavelength LED module and a specially designed phone fixture to reduce spatial errors and motion artifacts, respectively. In addition, we embedded a new algorithm into our smartphone-based sensor to improve the measurement accuracy; an L*a*b* color space transformation and the “a” parameter were used to perform the final quantification. We collected 24 blood samples from normal and anemic populations. The adjusted R2 of the prediction results obtained from the multiple linear regression method reached 0.880, and the RMSE reached 9.04, which met the accuracy requirements of non-invasive detection of hemoglobin concentration.
