Recently, the “internet of things” has become a hot topic for researchers worldwide. One result of this development is that wearable terminals for health monitoring in daily life have appeared. At present, however, only a few sensors can be installed in these terminals, especially when implemented in smartphones. The sensors that are installed in smartphones can only realize a few applications, e.g., pedometers and heart rate monitors. In addition, because spectroscopic imagers for smartphones have not yet been put into practical use, it is currently impossible to measure spectral components using smartphones. The objective of this work is the realization of a noninvasive blood glucose sensor using a proposed one-shot Fourier spectroscopic imager for installation in a smartphone.
Our proposed one-shot Fourier spectroscopic imager1 is configured with a near-common-path wavefront-division interferometer,2,3 is strongly robust against mechanical vibration, and will be realized as a low-cost and small-sized spectroscopic unit (the optical system dimensions are , which is compatible with the thickness of a typical smartphone) because the imager consists of only three lenses. In addition, the spectral distribution can be obtained with a single shot (using one frame of image data). Therefore, it is easy to perform measurements of active biological tissues because the proposed spectroscopic imager has a high time resolution. Additionally, our proposed method has high resolution and high accuracy4 in a similar way to blood glucose sensing by optical coherence tomography56.–7 (OCT). OCT can monitor the optical properties of tissue without any unwanted signal from other layers. Our proposed one-shot Fourier spectroscopic imager can also achieve this because of limiting the focal plane and scanning the depth. Furthermore, our proposed method can obtain the spectroscopic characteristics without a refractive index between wanted and unwanted components for confirming the high contrast of the image, and the imager is smaller for installation into smartphones.
To realize a noninvasive medical sensor within a smartphone, it will be necessary to miniaturize the high sensitivity spectroscopic imagers. Demand for compact-size spectrometers already exists and they have been researched,89.–10 but it will also be necessary for spectroscopic imagers to use the weak reflected light from biological tissues for blood glucose measurements. However, in the conventional microelectromechanical system Fourier transform infrared spectroscopy method, which is based on the use of a Michelson interferometer, the light efficiency is poor because the mirror processing accuracy is low. In addition, acousto-optic tunable filter-based and dispersive spectrometers cannot improve the sensitivity, at least in principle,1112.13.–14 because these spectroscopies receive narrowband light. It is therefore impossible for conventional methods to achieve the required wavelength resolution and sensitivity because smartphones cannot accommodate expensive and large-scale cameras with high sensitivity and we cannot use strong light sources that are unsafe for use on human skin.
Because a one-shot Fourier spectroscopic imager introduces a field stop, which is a diffraction grating, into the Fourier transform plane, the resulting spectroscopic method shows a lack of sensitivity.15 However, we have improved the sensitivity using the wide-field-stop and beam-expansion method while simultaneously maintaining the required wavelength resolution. In this paper, we explain the principle and availability of the method and demonstrate the feasibility of biomedical measurements for the realization of a noninvasive blood glucose sensor.
Principle of the Wide-Field-Stop and Beam-Expansion Method
In this section, we explain the principle of the wide-field-stop and beam-expansion method for the sensitivity enhancement of one-shot Fourier spectroscopy. One-shot Fourier spectroscopy uses a special phase-shift interferometer, in which the horizontal axis on the imaging device is assigned to the phase-shift value. The spectroscope is a simple optical system that consists of three lenses: an objective lens, a cylindrical lens, which is the imaging lens, and a transmission-type relative-inclined phase-shifter, which is configured using a wedge glass and a cuboid glass.
First, the objective beam is emitted from a single bright point on the object plane and is then collimated by the objective lens. The transmission-type relative-inclined phase-shifter provides a continuous spatial phase difference between the objective beams. Subsequently, two beams are formed by the cylindrical lens on the imaging plane and these beams interfere with each other. We then confirm the interferogram as a fringe pattern spatially. In addition, our proposed method can realize a one-dimensional spectroscopic measurement with one shot (i.e., using a single frame of image data) because interferograms can be obtained simultaneously at each line on the light-receiving device. However, the spectroscopy system has low sensitivity because a slit, which acts as a field stop, is introduced into the optical system.
To increase the detected light intensity, use of a wider slit in the field stop would obviously produce a higher light intensity. However, because our method is based on wavefront-division interferometry, a problem arises where the use of a wider single slit also makes the diffraction angle narrower, as shown in Fig. 1(a). This means that the narrower diameters of the collimated objective beams would cause the visibility of the interferograms to deteriorate. By installing a relative-inclined phase-shifter on the optical Fourier transform plane of an infinity-corrected optical system, the collimated half flux of the objective beams that are derived from single bright points on the objective surface penetrate through the wedge prism and the cuboid glass. The two beams then interfere with each other and form an interferogram in the form of spatial fringe patterns. We thus installed a plane-concave cylindrical lens between the wider slit and the objective lens to act as a beam expander.
Two main configurations are available for beam expanders: Keplerian and Galilean. In the proposed method, we installed a Galilean beam expander so that the spectrometer may be introduced into a smartphone, as the Galilean beam expander requires a shorter optical path length than the Keplerian device. A Galilean beam expander requires a planoconcave lens, but when we used a planoconcave lens as the Galilean beam expander, the optical path lengths were extended not only on the horizontal axis but also on the vertical axis. We then used a planoconcave cylindrical lens rather than a planoconcave lens to extend the optical path length on the horizontal axis only, as shown in Fig. 1(b), because it avoids cancellation of the adjacent bright spots that prevent us from obtaining the interferogram.16 As a result, the continuous spatial-phase-difference objective beams coincide with each other on the imaging plane and can be interfered. Therefore, we have improved the sensitivity of the proposed one-shot Fourier spectroscopy method using the wide-field-stop and beam-expansion method.
We confirmed the results of the wide-field-stop and beam-expansion method through verification experiments, as shown in Fig. 2. In the experiments, the one-shot Fourier spectroscopic imager consisted of an objective lens (focal length of 50 mm), an imaging lens (cylindrical lens, focal length of 50 mm), and a transmission-type relative-inclined phase-shifter. We then structured the Koehler illumination to average the distribution of the light source. We also introduced a planoconcave cylindrical lens (focal length of ) into the optical system to act as a Galilean beam expander. We used a complementary metal–oxide–semiconductor camera (XCD-MV6, Sony Corporation) on the imaging plane, the slit (horizontal pattern length of ; vertical pattern length of ) on the object plane, and a white light-emitting diode (OSW54L5111P, OptoSupply). The objective beams that were collimated by the objective lens formed on the imaging plane but did not interfere because the beam diameter was short and the beams coincided with each other without the Galilean beam expander. We did not obtain the interferogram and the spectral characteristics of the light source, as shown in Fig. 3(a). However, the objective beams that were collimated by the objective lens formed on the imaging plane and interfered with each other when using the Galilean beam expander. We could therefore obtain the required interferogram and the spectral characteristics of the light source because of the expansion of the length of the beams by the beam expander, as shown in Fig. 3(b). Therefore, we demonstrated the sensitivity enhancement required for the one-shot Fourier spectroscopic imager when using the wide-field-stop and beam-expansion method by confirming the target spectral characteristics and acquiring the interferogram after the amount of light available was increased by expanding the slit width.
We measured the absorbance of rat blood using a dispersive spectrometer (SolidSpec-3700, Shimadzu) and the absorbance of a human hand using the proposed one-shot Fourier spectroscopic imager with enhanced sensitivity. In this experiment, a halogen lamp (JR12V50WLN/NZ/EZ-IR, Ushio) was used as the light source, as shown in Fig. 4. Then we obtained the image and the interferogram as shown in Figs. 5(a) and 5(b), respectively. The two samples show similar absorption characteristics for hemoglobin, with a decrease from 600 to 700 nm and increase from 700 to 900 nm, as shown in Figs. 5(c) and 5(d). We have therefore confirmed the diffuse reflection of light from human skin and have demonstrated the feasibility of performing noninvasive biomedical measurements using the one-shot Fourier spectroscopic imager after the proposed sensitivity enhancement.
We propose the use of the wide-field-stop and beam-expansion method for sensitivity enhancement of the one-shot Fourier spectroscopy method. We discuss the principles of the sensitivity improvement method, which supports both the sensitivity and wavelength resolution requirements of this spectroscopic method. Additionally, we confirm the suitability of the method through verification experiments and biomedical experiments. We have thus demonstrated the feasibility of realizing a spectroscopic imager that can be installed in smartphones for health monitoring applications.
S. Sato et al., “Proposal of one-shot-type spectroscopic-tomography for non-invasive medical-measurement,” Proc. SPIE 8798, 87980I (2013).PSISDG0277-786Xhttp://dx.doi.org/10.1117/12.2032494Google Scholar
J. Szlachetko et al., “Wavelength-dispersive spectrometer for x-ray microfluorescence analysis at the x-ray microscopy beamline ID21 (ESRF),” J. Synchrotron Radiat. 17, 400–408 (2010).JSYRES0909-0495http://dx.doi.org/10.1107/S0909049510010691Google Scholar
Y. Yasuno et al., “One-shot-phase-shifting Fourier domain optical coherence tomography by reference wavefront tilting,” Opt. Express 12, 6184 (2004).OPEXFF1094-4087http://dx.doi.org/10.1364/OPEX.12.006184Google Scholar
A. Ishida et al., “Quantitative measurement of biological substances in daily-life environment with the little-finger-size one-shot spectroscopic tomography,” Proc. SPIE 8951, 89510Y (2014).PSISDG0277-786Xhttp://dx.doi.org/10.1117/12.2038836Google Scholar
H. Ullah et al., “Autocorrelation optical coherence tomography for glucose quantification in blood,” Laser Phys. Lett. 12(12), 125602 (2015).1612-2011http://dx.doi.org/10.1088/1612-2011/12/12/125602Google Scholar
K. Larin et al., “Specificity of noninvasive blood glucose sensing using optical coherence tomography technique: a pilot study,” Phys. Med. Biol. 48(10), 1371–1390 (2003).PHMBA70031-9155http://dx.doi.org/10.1088/0031-9155/48/10/310Google Scholar
R. Esenaliev et al., “Noninvasive monitoring of glucose concentration with optical coherence tomography,” Opt. Lett. 26(13), 992 (2001).OPLEDP0146-9592http://dx.doi.org/10.1364/OL.26.000992Google Scholar
S. Nakadate et al., “Real-time vibration measurement by a spatial phase-shifting technique with a tilted holographic interferogram,” Appl. Opt. 36, 281 (1997).APOPAI0003-6935http://dx.doi.org/10.1364/AO.36.000281Google Scholar
S. Murata et al., “Spatial phase-shifting digital holography for three-dimensional particle tracking velocimetry,” Jpn. J. Appl. Phys. 48, 09LB01 (2009).http://dx.doi.org/10.1143/JJAP.48.09LB01Google Scholar
Y. Inoue et al., “Variable phase-contrast fluorescence spectrometry for fluorescently stained cells,” Appl. Phys. Lett. 89, 121103 (2006).APPLAB0003-6951http://dx.doi.org/10.1063/1.2356312Google Scholar
H. Kobayashi et al., “Spectroscopic tomography by imaging-type two-dimensional Fourier spectroscopy,” Optics 41, 136–144 (2012).Google Scholar
J. Katrašnik et al., “Spectral characterization and calibration of AOTF spectrometers and hyper-spectral imaging system,” Chemometr. Intell. Lab. 101, 23–29 (2010).http://dx.doi.org/10.1016/j.chemolab.2009.11.012Google Scholar
N. Kawashima, S. Hosono and I. Ishimaru, “Built-in hyperspectral camera for smartphone in visible, near-infrared and middle-infrared lights region (second report): sensitivity improvement of Fourier-spectroscopic imaging to detect diffuse reflection lights from internal human tissues for healthcare sensors,” Proc. SPIE 9855, 985505 (2016).PSISDG0277-786Xhttp://dx.doi.org/10.1117/12.2223392Google Scholar