|
Figure 1. Schematic diagram of the 4D NIR spectral-microtomographic system. As the source light enters the optical system, the sample and reference beam paths are split and combined before images are acquired. SCL: supercontinuum laser; TBPF: tunable bandpass filter; PCF: photonic crystal fiber; L1 and L2: lenses; M1: mirror; BS1, BS2 and BS3: beam splitters; DGM: dual-axis galvanometer mirror; CL: condenser lens; OL: objective lens; TL: tube lens; VOPL: variable optical path length.
|
|
Figure 2. Example images acquired with the developed system. (a,b) The amplitude and phase images of a hair, respectively, imaged at 1100 nm. (c,d) The amplitude and phase images of the same hair, respectively, imaged at 1500 nm. The amplitude images in (a,c) are normalized with the background images, i.e., the amplitude images acquired without the sample at the corresponding wavelengths. The phase images in (b,d) are shown after subtracting the background images, i.e., the phase images acquired without the sample at the corresponding wavelengths (unit: radian). The amplitude and phase images for the two different wavelengths (1100 and 1500 nm) clearly show the wavelength-dependent absorption and refractive index properties of the hair. (e) A bright-field image of the same hair. Scale bars: 50 μm.
|
|
Figure 3. Digital holographic tomography data processing. For each wavelength, the amplitude and phase images recorded for an illumination angle are used to synthesize a scattered light field from the sample. The arrow in (a) represents the wave vector k0→=(uo,v0,w0) of the illumination beam, whose magnitude and direction correspond to nm/λ and the illumination direction onto the sample plane, respectively. As shown in (b), the Fourier transform of the scattered light field is mapped onto the Ewald’s sphere, which is shifted in the opposite direction to the wave vector, in the 3D spatial frequency space. (X,Y,Z) are the Cartesian coordinates with Z being the optical axis direction. (U,V,W) are the spatial frequency components corresponding to (X,Y,Z), respectively.
|
|
Figure 4. Examples of the absorption coefficient and refractive index maps reconstructed from a series of scattered field measurements. (a,b) The absorption coefficient and refractive index maps, respectively, for the centre cross-section at a wavelength of 1100 nm. (c,d) The absorption coefficient and refractive index maps, respectively, for the centre cross-section at a wavelength of 1500 nm. The unit for the absorption coefficient is cm−1. Scale bars: 50 μm.
|
|
Figure 5. (a) a 3D-rendered image of hair calculated from the 3D refractive index map at a wavelength of 1100 nm. Although the optical system is capable of a 5 nm spectral resolution, the step size of 50 nm was found to provide adequate spectral resolution to capture the relevant characteristics of the absorbance of the hair sample in the SWIR wavelength range. (b) An example of the 3D absorption coefficient map of hair at a wavelength of 1500 nm. Five vertical cross-sections at 30 μm intervals are shown as an example. (c,d) The mean absorption coefficient and mean refractive index value, respectively, of the hair sample as a function of the wavelength in the 1100–1650 nm range. Scale bars in (a,b): 50 μm.
|
|
Figure 6. Refractive index cross-sections of the sea urchin embryos in different developmental stages: (a) egg, (b) two cells, (c) four cells, (d) morula, (e) blastula, (f) gastrula. The refractive index was measured at a wavelength of 1100 nm. The scale bar (50 μm) in (a) applies to all the images.
|
|
Figure 7. Bright-field images of the sea urchin embryos in different developmental stages: (a) egg, (b) two cells, (c) four cells, (d) morula, (e) blastula, (f) gastrula. The scale bar (50 μm) in (a) applies to all the images.
|
|
Figure 8. Refractive index of the sea urchin embryos in different developmental stages measured at a wavelength of 1100 nm. In each box, the central mark (red) indicates the median, the bottom and top edges (blue) indicate the 25th and 75th percentiles, respectively, and the whiskers (black) mark the range of data points not considered outliers (red crosses).
|
|
Figure 9. 3D point spread function (PSF) of the developed instrument measured with gold nanoparticles (GNPs). (a,b) The horizontal and vertical cross-sections, respectively, of the 3D PSF before applying the regularization. (c,d) The horizontal and vertical cross-sections, respectively, of the 3D PSF after applying 30 iterations of the regularization. The 3D PSF measured with GNPs was averaged along the polar coordinate in each horizontal cross-section. The transverse and axial resolutions determined with the full-width at half maximum (FWHM) are 1.43 μm and 3.53 μm before the regularization, and 1.51 μm and 1.57 μm after the regularization. The scale bar (2 μm) in (a) applies to all the images.
|
