Isothermic determination of aqueous solution glucose concentration in low mg/dL range by CW-photoacoustic-based protocol
详细信息 查看全文
- 作者：S. Camou ; Y. Ueno ; E. Tamechika
- 关键词：Noninvasive blood glucose monitoring ; Continuous-wave photoacoustic ; Frequency shift protocol ; High sensitivity ; Temperature dependence
- 刊名：Sensors and Actuators B: Chemical
- 出版年：2013
- 出版时间：August, 2013
- 年：2013
- 卷：185
- 期：Complete
- 页码：568-574
- 全文大小：1684 K
文摘
In order to decrease the impact of diabetes mellitus on patient quality of life, noninvasive and continuous monitoring of blood glucose level represents the ideal detection protocol for an appropriate decision-making process. The limit of detection of such a sensor should be lower than 70 mg/dL for efficient detection of hypoglycemia episodes. However, a reasonable anticipation of loss of accuracy when moving from in vitro to in vivo requires demonstration of sensitivity several times lower than this value in the ideal in vitro conditions for any viable alternative to the standard finger-prick-based method. This paper then investigates the detection of aqueous glucose solutions at low concentration levels based on a measurement protocol called ¡°frequency shift¡± (FS). This protocol, based on the continuous-wave photoacoustic technique, is equivalent to an acoustic velocity measurement and therefore depends on the temperature as well. A thermo-regulated bath system was then used to control the temperature with an accuracy of ¡À0.035 ¡ãC. The sensor showed a linear glucose dependence characterized by a slope of 0.19 % /g/dL, which is consistent with previously reported results, and accuracy of about ¡À26 mg/dL was obtained. Using a differential experimental protocol, we were able to improve the temperature control accuracy down to ¡À0.015 ¡ãC. Measurements down to a 9.6 mg/dL concentration level were demonstrated with an accuracy of ¡À3.5 mg/dL despite a bias from temperature fluctuations equivalent to a ¡À11 mg/dL glucose concentration variation. The FS protocol then enables low milligram per deciliter glucose concentration measurements at levels several times lower than the requirements. However, temperature variations strongly limit the measurement accuracy due to the lack of FS selectivity to glucose. To take full advantage of the proposed approach, temperature should then be monitored simultaneously and its effect compensated to allow comparable sensitivity in in vivo conditions.