Session Index

A system calibration method is developed for single photon emission microscope. The method includes the geometric calibration, grid-scan experiment and an imaging model for building the system matrix essential in the tomographic reconstruction. A helical reconstruction algorithm is also developed. The results of phantom imaging validate the proposed calibration procedure.

 

The sampling completeness coefficient (SCC) and reconstruction images for single photon emission microscope are simulated for circular and helical trajectories. The results show that SCC values are consistent with the reconstructed image quality; thus the SCC analysis can be used to evaluate the system design before building an actual system.

 

A biosensor based on dual-frequency liquid crystal (DFLC) was developed. Using its dielectric properties, we exploited a way to detect and quantify bovine serum albumin (BSA). Analyzing dielectric spectra in presence of BSA at various concentration, we established that the difference in dielectric permittivity is corresponding to BSA concentration.

 

We developed a carboxyl-GO based SPR biosensor can quantitative analysis of human PAPP-A2 in spiked plasma. The carboxyl-GO was shown to have biocompatible surfaces beneficial for the adsorption of covalently-linked protein in biosensor. The LOD of the PAPP-A2 in spiked plasma was determined to be as low as 25 fg/mL.

 

We demonstrated a compact, high-resolution spectral-domain optical coherence tomography (SD-OCT) system using a USB 3.0 2048-pixel line scan camera. Volumetric OCT imaging of the cell spheroids before and after adding gold nanoparticles (GNPs) was performed where increased OCT signal intensity and decreased imaging depth due to GNPs can be observed.

 

We have demonstrated the parallel multi-line scanning mechanism could push AEC to 1.7 μm from 3.5 μm of conventional Temporal focusing multiphoton microscopy (TFMPM). Only 8 patterns are required can have a uniform wide-field excited multiphoton image without the dependence on the required image pixels as line-scanning based TFMPM.

 

The fluorescence microscopy detects the analog signal from electro-optical detector indicating the fluorescence intensity. With the FPGA-based single photon counting technique that has high SNR detection ability, the high-speed optoelectronic signal converter improves the image contrast and could be easily integrated to the standard multiphoton or confocal fluorescence microscope.

 

Asymmetric illumination-based differential phase contrast (AIDPC) is an imaging method provides better resolution. Here we take numerical method to simulate AIDPC microscope, analyzing effect of different amplitude patterns. Two types of amplitude patterns are tested and preliminary simulation shows promising results. This simulation would help researchers to amend AIDPC system.

 

Measuring the fluorescence lifetime reveals the dissipative velocity of green algae for nonphotochemical quenching of excess light energy and the different molecular mechanisms that dissipate different light energies under different colored light illumination. This study can help identify which growth light energy to use maximize plant growth.

 

The impact of dental disease on humans should not be underestimated. However, the current dental diagnostic tools still need to be improved. We proposed a method for automatic diagnosis. The test results had a sensitivity of 86%, specificity of 100%, and total accuracy of 94%.

 

Light field technique at a single shot can get the whole volume image of observed sample. By integrating a light field microscopy to a temporal focusing-based multiphoton excitation fluorescence microscope, the excitation volume can be controlled. Therefore, light field microscope based on temporal focusing obtains three-dimensional image with less noise.

 

We integrated the 2D PhCs with guided mode resonance for improving the performance of the sensor. We observe it has a dip in the specific wavelength, which is more sensitive to the changes of the refractive index. In the glucose measurement, it shows an increasing sensitivity and a lower LOD.

 

A 2D photonic crystal biosensor with cost-effective measurement system is developed. By extracting the spectra of the diffracted beam under white light illumination, the biosensor is able to detect 0.001 μg/mL of Epstein-Barr nuclear antigen-1 antibody in PBS, which is related to small refractive index changes on the sensor surface.

 

We have implemented a full-range Doppler spectral-domain optical coherence tomography without additional phase shifters, which can obtain cross-sectional image and fluid velocity simultaneously. We also propose and demonstrate a digital dispersion compensation method to remove the image distortion and phase deviation.

 

We developed a carboxyl-GO-modified biomaterial-based SPR biosensor for the rapid and sensitive detection of CK19, in spiked human plasma. The lowest detectable concentration was 1 fg/mL. A spiked 10% human plasma CK19 detection of 0.05 pg/mL was achieved, well below the normal physiological level of serum protein (3.3 ng/mL).

 

We use organic light-emitting diodes as a light source to treat neonatal jaundice patients. We can not only solve the problem of overheating of the light source in the treatment, but also use the flexible characteristics of OLED to do a partial treatment in the future.

 

As the global population ages, dementia has become a major problem. However, the clinical diagnosis of dementia is not objective. We hope to understand the pathological mechanism of dementia and provide objective diagnosis basis by using NIRS and MCST. The results of this study show the feasibility of the method.

 

We classify and segment optical coherence tomography (OCT) images of skin tissue using various convolutional neural networks. Specifically, we employ supervised learning to train deep learning model. In addition, we use pixel-based performance metrics to evaluate the segmentation accuracy. Our research goal is to analyze OCT images via deep learning.

 

Yeast fermentation is the most mature technology for producing alcohol. However, the effect of ambient light source on this fermentation has not been seen yet; our research found that the illuminance is fixed at 500lx, the 2,300K has a faster fermentation rate, which is 38% higher than that without illumination.

 

Image recognition for insect pests plays a most fundamental role in the field of Intelligent Agriculture robotic systems. Our experiment shows that the test images without complicated background has a better insect pest recognition rate when using a Convolutional Neural Network as an image recognizer.

 

In this study, we propose to use optical coherence tomography (OCT) for the study of protein-related degeneration of Drosophila eyes. With OCT, the bristles of Drosophila eye can be identified and different mutant Drosophila were scanned with OCT for investigation of the progress of protein-related degeneration.

 

We used two-photon microscopy to acquired images of normal human lung cell line Beas2B, human lung adenocarcinoma CL1-0 and CL1-5 cell lines. Based on our program, we found 2D N/C-ratio is more sensitively than 3D N/C-ratio. Overall, we remove ambiguity in determination of N/C ratios and may improve clinical diagnosis.

 

In this research, the dynamic process of aggregation that forms microflower morphology in solution of lysozyme amyloid fibrils doped with spherical or spindle-like magnetic nanoparticles during the process of drying as well as their final microstructures were investigated.

 

We investigated how the stiffness of 3D collagen matrix affects invasion capability of OECM-1 spheroid via traction force microscopy and time- lapse microscopy. We observed that stiffer 3D collagen matrices limit contraction deformation of collagen and cancer invasion.