Study Of Multi-mode Microscopic Hyperspectral Imager For The Bio-sensing And High-dimensional Detection

With the advantages of being non-invasive,having high efficiency and high throughput,hyperspectral imaging(HSI)technology can simultaneously obtain spatial and spectral information.As obtained spectra can reflect the absorption of a specific substance,and provide rich complex structural information that is related to the vibration behavior of molecular bonds,HSI is used for a wide range of applications,such as on-site environmental monitoring,food quality assessment and biomedicine diagnosis.Based on HSI,we designed and built a set of multi-mode microscopic hyperspectral imager(MMHI),and verified the capabilities for bio-sensing and high-dimensional detection.Under the 40× objective,MMHI system can achieve 5μm spatial resolution and 3nm spectral resolution,which meets the requirements of most microscopic scenes.By designing two coaxial illumination structures,the MMHI system can work in three modes,i.e.reflection imaging for opaque samples,transmission imaging for transparent/translucent samples and fluorescence detection for fluorescent samples.Three experiments including detections for zebrafish,fingerprint and microalgae verified the functionality.Then,we performed transmission HSI for three kinds of microalgae,and verified their characteristic absorption.Combining the transmittance with machine learning algorithms,the classification,identification and growth stage prediction were achieved with very high accuracy.Finally,based on laser triangulation,a set of quantitative light-sheet microscopy(LSM)was established for 3D topography measurements,and was integrated with previous MMHI system,realizing 4D(3D in spatial and 1D in spectral)hyperspectral detection of corn seed.In conclude,experimental results show prospects of the system in high-dimensional and multi-mode detection,which lays a solid foundation for more scientific researches and industrial applications in the future.