This PDF file contains the front matter associated with SPIE Proceedings Volume 10505, including the Title Page, Copyright information, Table of Contents, and Conference Committee listing.

Fourier Ptychography has shown that we can computationally correct physical aberrations. Thereby, allow us to move beyond the traditional strategy of accomplishing high quality imaging through the exacting refinement of the physical microscope system. I will report on the use of Fourier Ptychography to implement high quality parallel imaging with plastic molded lenses for 96 well plate imaging.

Multiphoton microscopes are of paramount importance in capturing neural activity with cellular resolution. However, the imaging speed and field-of-view of traditional two-photon microscopes is limited by raster scanning technologies. Temporally-focused two-photon (TFTP) microscopy is a wide-field scan-free approach to increase the speed of two-photon microscopy. In conventional TFTP microscopy, wide-field depth sectioning is obtained by compressing a spatially pre-chirped pulse at the focal plane of the objective. Unfortunately, the greater imaging speed of TFTP microscopes comes at the expense of poor imaging depth in tissue due to scattering of the short-wavelength fluorescence photons en-route to the imaging camera. Here we demonstrate a compressive high-speed two-photon microscope based on wide-field temporally-focused structured illumination, which eliminates the loss of image contrast from scattering of the fluorescence signal by leveraging a single-pixel detector. Specifically, we illuminate the sample with a rapid sequence of randomly structured temporally-focused wide-field illumination pulses and integrate the net two-photon fluorescence response on a single photomultiplier tube (PMT). Notably, the longer wavelength structured illumination is significantly less susceptible to scattering and the use of integrated measurements on a single PMT provides immunity to fluorescence scattering since these measurements are solely concerned with the net fluorescence. Furthermore, our approach provides greater speed than point scanning two-photon microscopes through the use of wide-field illumination and compressive image acquisition. Experimentally we demonstrate this system operating over a 200×250-μm field-of-view and at a compression rate of 10%, which provides an order of magnitude increase in speed over a comparable point scanning architecture.

Two ultrahigh-speed CCD image sensors with different characteristics were fabricated for applications to advanced scientific measurement apparatuses. The sensors are BSI MCG (Backside-illuminated Multi-Collection-Gate) image sensors with multiple collection gates around the center of the front side of each pixel, placed like petals of a flower. One has five collection gates and one drain gate at the center, which can capture consecutive five frames at 100 Mfps with the pixel count of about 600 kpixels (512 x 576 x 2 pixels). In-pixel signal accumulation is possible for repetitive image capture of reproducible events. The target application is FLIM. The other is equipped with four collection gates each connected to an in-situ CCD memory with 305 elements, which enables capture of 1,220 (4 x 305) consecutive images at 50 Mfps. The CCD memory is folded and looped with the first element connected to the last element, which also makes possible the in-pixel signal accumulation. The sensor is a small test sensor with 32 x 32 pixels. The target applications are imaging TOF MS, pulse neutron tomography and dynamic PSP. The paper also briefly explains an expression of the temporal resolution of silicon image sensors theoretically derived by the authors in 2017. It is shown that the image sensor designed based on the theoretical analysis achieves imaging of consecutive frames at the frame interval of 50 ps.

Existing multispectral imagers mostly use 2D array sensors to separately measure 2D data slices in a 3D spatialspectral data cube. They suffer from low photon efficiency, limited spectral range, and high cost. To address these issues, we propose to conduct multispectral imaging using a photodiode, to take full advantage of its high sensitivity, wide spectral range, low cost, and small size. Specifically, utilizing the photodiode’s fast response, a scene’s 3D spatial-spectral information is sinusoidally multiplexed into a dense 1D measurement sequence, and then demultiplexed computationally under the single-pixel imaging scheme. A proof-of-concept setup is built to capture multispectral data of 256 pixels × 256 pixels × 10 wavelength bands ranging from 450 nm to 650 nm. The imaging scheme holds great potentials for various biological applications such as fluorescence microscopy and endoscopy.

We demonstrate the confocal phase imaging at a video rate by a combination of dual comb spectroscopy (DSC) with 2D spectral encoding (2D-SE). After the image pixels of the sample is encoded on an optical frequency comb (OFC) by 2DSE, DSC of the image-encoded OFC passing through the confocal pinhole gives the mode-resolved amplitude and phase spectra. Based on one-to-one correspondence between the image pixels and OFC modes, the confocal amplitude and phase images are decoded from the mode-resolved amplitude and phase spectra, respectively. The phase spectrum measurement without the need for mechanical scanning enables the video-rate confocal phase imaging.

Single photon avalanche diodes (SPADs) are used in a wide range of applications, from fluorescence lifetime imaging microscopy (FLIM) to time-of-flight (ToF) 3D imaging. SPAD arrays are becoming increasingly established, combining the unique properties of SPADs with widefield camera configurations. Traditionally, the photosensitive area (fill factor) of SPAD arrays has been limited by the in-pixel digital electronics. However, recent designs have demonstrated that by replacing the complex digital pixel logic with simple binary pixels and external frame summation, the fill factor can be increased considerably. A significant advantage of such binary SPAD arrays is the high frame rates offered by the sensors (>100kFPS), which opens up new possibilities for capturing ultra-fast temporal dynamics in, for example, life science cellular imaging. In this work we consider the use of novel binary SPAD arrays in high-speed particle tracking in microscopy. We demonstrate the tracking of fluorescent microspheres undergoing Brownian motion, and in intra-cellular vesicle dynamics, at high frame rates. We thereby show how binary SPAD arrays can offer an important advance in live cell imaging in such fields as intercellular communication, cell trafficking and cell signaling.

Phytoplankton is highly diversified in species, differing in size, geometries, morphology and biochemical composition. Such diversity plays a critical role in the atmospheric carbon cycle and marine ecosystem. Large-scale quantitation and classification of phytoplankton with taxonomic information is thus of significance in environmental monitoring and even biofuel production. To this end, we report a high-throughput, label-free imaging flow cytometer (>10,000 cells/sec) based on quantitative phase time stretch imaging flow cytometry, combined with a supervised learning strategy for multi-class classification of phytoplankton (13 classes). This is in contrast to the previous demonstrations on integrating machine learning with time-stretch imaging which achieve high-accuracy binary (two-class) image-based classification. We leverage interferometry-free quantitative phase time-stretch imaging which favors generation of high-resolution and high-contrast single-cell (phytoplankton) images with both quantitative phase and amplitude contrasts, we can extract a catalogue of 109 image-content-rich features (44 from the amplitude image and 65 from the phase image), not only limited to sizes, shapes, but also sub-cellular morphology, e.g. local dry mass density statistics. By using the random forest algorithm for feature ranking, we select 30 most significant features for a multi-class SVM model and achieve a high classification accuracy (> 95%) across 13 classes of phytoplankton. Almost 50% of these selected features are derived from the quantitative phase and play an important role in classifying morphologically similar species, e.g. Thalassiosira versus Prorocentrum; Chaetoceros gracilis versus Merismopedia – demonstrating the classification power of this quantitative phase time-stretch imaging flow cytometer required for large-scale high-content screening and analysis.

Light-sheet fluorescence microscopy enables high-speed, high-resolution, gentle imaging of live biological specimens (spanning length scales from single cells to large organs) over extended periods (from hours to days). Here we report a reflective light sheet imaging technique that further improves the spatiotemporal resolution and collection efficiency of our dual-view light sheet microscope (diSPIM). By imaging samples on reflective coverslips, we enable simultaneous collection of four complementary views in 250 ms, doubling speed and improving information content relative to previous diSPIM. We further enhance spatial resolution to less than 300 nm in all three dimensions by combining a high numerical aperture (NA) lens with the use of reflective coverslips. We also present a modified deconvolution algorithm that removes associated epifluorescence contamination and fuses all views for resolution recovery. We demonstrate the broad applicability of the methods in a variety of samples, investigating the dynamics of mitochondria, membranes, Golgi, and microtubules in single cells, and neurodevelopment in nematode embryos. Finally, we describe computational methods for untwisting worm embryos and analyzing calcium activity in freely moving embryos. Our imaging and data processing tools pave the way for establishing a 4D dynamic neurodevelopmental atlas in nematode embryos.

Fluorescence imaging flow cytometry offers highly accurate analysis of a large number of cells compared with conventional flow cytometry by virtue of its imaging capability. Unfortunately, the throughput of conventional fluorescence imaging flow cytometers is limited to ~1,000 cells/sec, which is one order of magnitude lower than that of conventional non-imaging flow cytometers. This is due to the low data transfer rate of a CCD image sensor with a time-delay integration technique employed to achieve sufficient sensitivity for fluorescence imaging of fast flowing cells. Replacing the CCD image sensor with a CMOS image sensor can potentially overcome the throughput limitation by virtue of its high data transfer rate, but critically sacrifice the imaging sensitivity because the time-delay integration cannot be employed to current CMOS image sensors. Here we present a fluorescence imaging flow cytometer that achieves comparable throughput and sensitivity with non-imaging flow cytometers. It is enabled by high-energy-density light-sheet excitation of flowing cells on a mirror-embedded PDMS-based microfluidic chip and by fluorescence image acquisition with a CMOS image sensor. The light-sheet excitation allows us obtain fluorescence images of flowing cells at a speed of >1 m/s, corresponding to a high throughput of >10,000 cells/sec. To show its biomedical utility, we use it combined with machine learning to demonstrate accurate screening of white blood cells and real-time identification of cancer cells in blood.

Multi-channel microscopy and multi-channel flow cytometry generate high bit data streams. Multiple channels (both spectral and spatial) are important in diagnosing diseased tissue and identifying individual cells. Omega Optical has developed techniques for mapping multiple channels into the time domain for detection by a single high gain, high bandwidth detector. This approach is based on pulsed laser excitation and a serial array of optical fibers coated with spectral reflectors such that up to 15 wavelength bins are sequentially detected by a single-element detector within 2.5 μs. Our multichannel microscopy system uses firmware running on dedicated DSP and FPGA chips to synchronize the laser, scanning mirrors, and sampling clock. The signals are digitized by an NI board into 14 bits at 60MHz – allowing for 232 by 174 pixel fields in up to 15 channels with 10x over sampling. Our multi-channel imaging cytometry design adds channels for forward scattering and back scattering to the fluorescence spectral channels. All channels are detected within the 2.5 μs – which is compatible with fast cytometry. Going forward, we plan to digitize at 16 bits with an A-toD chip attached to a custom board. Processing these digital signals in custom firmware would allow an on-board graphics processing unit to display imaging flow cytometry data over configurable scanning line lengths. The scatter channels can be used to trigger data buffering when a cell is present in the beam. This approach enables a low cost mechanically robust imaging cytometer.

The optical sectioning capability is fundamental for three-dimensional imaging. One of the very few microscopes, which claim this property is light sheet-based fluorescence microscopy (LSFM). In general, FM provides a high contrast, since only specifically labelled cellular components are observed while all other structures remain “dark”. However: 1) Excitation light degrades endogenous organic compounds and bleaches fluorophores. 2) Only a finite number of excitable fluorophores is available, which limits the quantity of collectable emitted photons. 3) Organisms are adapted to the solar flux of 1.4 kW/m2. LSFM decouples the excitation and emission light pathways. The optical axis of the illumination objective lens is aligned with the focal plane of the perpendicularly arranged detection objective lens. By design, only the fluorophores around the focal plane are excited for each acquired two-dimensional image. Hence, each fluorophore is exposed only once to the illumination light when a three-dimensional image is recorded. The significance of the illumination-based optical sectioning property is that the viability and the fluorescence signal of a living specimen are retained while millions of images are recorded for days or even weeks. Further benefits: (i) a good axial resolution, (ii) imaging along multiple directions, (iii) deeper tissue penetration due to the low numerical aperture of the illumination objective lens, (iv) a high signal-to-noise ratio, (v) an unrestricted compatibility with fluorescent dyes and proteins, (vi) reduced fluorophore bleaching and (vii) photo-toxicity at almost any scale, (viii) millions of pixels are recorded in parallel and (ix) a dramatically improved viability of the specimen.

Fluorescence imaging flow cytometry is an emerging technique for analyzing a large number of cells with high accuracy over conventional flow cytometry by virtue of its imaging capability. Unfortunately, the cell throughput of conventional fluorescence imaging flow cytometers (~1,000 cells/sec) is much lower than that of standard non-imaging flow cytometers due to the use of a CCD image sensor having a limited data transfer rate, making it difficult to analyze a large population of cells. Here we report our experimental demonstration of highly accurate classification of microalgae with a frequency-division-multiplexed confocal imaging flow cytometer (IFC) that enables imaging of every single microalgal cell with an unprecedentedly high throughput of 20,000 cells/sec. The high-speed imaging performance of the IFC is enabled by employing frequency-division-multiplexed confocal microscopy, which uses a sensitive single-pixel photodetector such as an avalanche photodetector or a photomultiplier tube to obtain images of flowing cells. We stained three species of microalgae (Chlamydomonas reinhardtii, Haematococcus lacustris, and Euglena gracilis) with SYTO16 and obtained three-color images of the cells (bright-field, fluorescence staining of nuclei, and autofluorescence of chlorophyll). We extracted 243-dimensional features from each three-color image and employed a support vector machine to classify the cells with the obtained multi-dimensional data. As a result, the cells were successfully classified with an accuracy of 99.7%. Due to the IFC’s multi-color imaging capability with an unprecedentedly high throughput, our technique has a wide variety of potential applications other than microalga classification, such as accurate blood screening and liquid biopsy.

Biophysical properties of cells could complement and correlate biochemical markers to characterize a multitude of cellular states. Changes in cell size, dry mass and subcellular morphology, for instance, are relevant to cell-cycle progression which is prevalently evaluated by DNA-targeted fluorescence measurements. Quantitative-phase microscopy (QPM) is among the effective biophysical phenotyping tools that can quantify cell sizes and sub-cellular dry mass density distribution of single cells at high spatial resolution. However, limited camera frame rate and thus imaging throughput makes QPM incompatible with high-throughput flow cytometry – a gold standard in multiparametric cell-based assay. Here we present a high-throughput approach for label-free analysis of cell cycle based on quantitative-phase time-stretch imaging flow cytometry at a throughput of > 10,000 cells/s. Our time-stretch QPM system enables sub-cellular resolution even at high speed, allowing us to extract a multitude (at least 24) of single-cell biophysical phenotypes (from both amplitude and phase images). Those phenotypes can be combined to track cell-cycle progression based on a t-distributed stochastic neighbor embedding (t-SNE) algorithm. Using multivariate analysis of variance (MANOVA) discriminant analysis, cell-cycle phases can also be predicted label-free with high accuracy at >90% in G₁ and G₂ phase, and >80% in S phase. We anticipate that high throughput label-free cell cycle characterization could open new approaches for large-scale single-cell analysis, bringing new mechanistic insights into complex biological processes including diseases pathogenesis.

Single-cell classification based on the cell’s visual images, i.e., their phenotypes, can greatly complement genomic-based techniques for anomaly detection, which in turn has the potential for assistance in early cancer diagnosis. A high-speed imaging system is often needed for capturing the individual cell images, and in addition, the process involves big data computation, as we often have a large amount of cells for analysis and classification. Here, we focus on the latter, where we devise a deep convolutional neural network (CNN) and show its efficacy for the task. Specifically, making use of an asymmetric-detection time-stretch optical microscopy (ATOM) for fast image capture, we obtain datasets of four cell types (MB231, MCF7, THP1, and PBMC) exceeding 900,000 cell images. After preprocessing the data, such as discarding the empty images and adjusting for different experimental conditions, we build an eight-layer network where the first six are, alternately, convolutional and pooling layers, and the last two are full connection layers. We make use of the rectified linear function (ReLU) as the nonlinear activation function and max-pooling in downsampling. We compute the neural network classification by training with 65% of the data, and using the remaining 15% for validation and 20% for test. By comparing with other experimental settings and classification schemes, we find that a high data volume even with lower resolution images can outperform the opposite setting, and CNN is a more reliable scheme than other machine learning algorithms such as k-nearest neighbor and support vector machine, especially with fewer input data.

Stimulated Raman scattering (SRS) microscopy that is capable of both high speed imaging and rapid spectroscopy will be advantageous for detailed chemical analysis of heterogeneous biological specimens. We have developed a system based on spectral focusing SRS technology, with the integration of a rapid scanning optical delay line (RSODL), which allows continuous tuning of SRS spectra by scanning a galvo mirror. We demonstrated SRS spectral measurements of dimethyl sulfoxide solution at low concentrations, and multi-color imaging of rice pollens and HeLa cells with line-by-line delay tuning to reduce motion artifacts, as well as fast acquisition of SRS spectra at specific regions of interest.

High-speed label-free imaging with chemical contrast is effective for non-invasive analysis of the metabolic heterogeneity of single cells. Stimulated Raman scattering (SRS) microscopy enables high-speed label-free image acquisition with molecular vibrational specificity. While single-color SRS microscopy only acquires images at a certain vibrational frequency, multicolor SRS microscopy successively acquires SRS images at different vibrational frequencies, which then can be used to investigate the distributions of different intracellular molecules. However, its imaging speed remains an order of magnitude slower than that of single-color video-rate SRS microscopy. Previous approaches to circumvent this issue used either only two colors with limited chemical specificity or multiplex detection of SRS spectra using a photodiode array at the expense of imaging speed. Here we demonstrate high-speed four-color SRS imaging using a single photodiode by introducing fast wavelength-switched laser pulses. The fast wavelength switching is realized by the use of an optical intensity modulator as a time gate, a diffraction grating, and fiber delay lines. Using the developed system, we demonstrate motion-artifact-free multicolor SRS imaging of polymer beads and living cells. The results firmly support that our method is a powerful tool for the label-free analysis of living cells in microbiology, oncology, plant science, and medicine.

We present a novel approach to balanced-detection stimulated Raman scattering (SRS) spectroscopy and microscopy, called In-line Balanced Detection (IBD). IBD-SRS not only completely removes high-frequency laser fluctuations, a crucial ingredient to improve signal-to-noise ratio in modulation-transfer techniques, but also passively and automatically balances the low-frequency signal variation due to spatially varying sample transmission. It takes advantage of polarization multiplexing and an inherently stable common-path geometry. A birefringent crystal before the sample creates two orthogonally polarized and collinear replicas of the Stokes pulse, with ~10-ps relative delay. The first serves as a reference pulse; the second, temporally overlapped with the pump pulse, probes the Raman response of the sample. As reference and probe pulses cross the sample at the same position, they maintain their balance during image acquisition. IBD can be implemented in any conventional SRS setup, by simply adding a few passive optical elements in the beam. We have tested its performances on a home-built multimodal laser-scanning microscope, coupled with a compact fiber-format laser source. We obtain common-mode noise rejection up to 30 dB with respect to the unbalanced case, thus reaching shot-noise-limited detection, without the need of any electronic (active) auto-balancing. We have employed IBD-SRS to distinguish different polymer beads, to locate lignin and cellulose in the walls of plant cells and to visualize the three-dimensional distribution of lipids in HuH7 and HepaRG hepatic cells. To demonstrate the suitability of IBD-SRS in scattering environments, we show a significant image-quality improvement also when measuring lipids in thick bovine liver tissues.

We developed 4/8K CMOS sensors, and spinning-disk based multi-color imaging system for living animals. 4/8K pixels, 60fps, and 4ch imaging covered from micro- to macro-scale in time and space, which enabled us to track single cell for long time including diseases progressions. Optical path was enlarged for large CMOS sensors, and pixel numbers. Amount of information per time was increased up to 6TB/1hr, and these system enabled us to track single cell for long time with sub-micro resolutions. We integrated optics, mechanics and electronics into one system, and sample manipulations were supported by robotics and VR (virtual reality). To understand dynamically developing thrombus and vascular responses, which is one of the main components of cardiovascular diseases, we performed intravital vessel visualization. We visualized, acute thrombus formation and chronic inflammatory responses in one views. We revealed platelet behaviors and platelet morphological changes at micro level, and malfunctioning at macro organ levels in real-time manners. Inflammatory reactions in hours are simultaneously evaluated by single imaging modality.

A biological sample consists of a variety of complex biomolecules, and fluorescence microscopy enables visualization of specific molecules at the sub-cellular level. However, these fluorescence techniques require certain fluorescence dyes to label the sample, and the fluorophores raise serious problems such as photo toxicity and photobleaching which could affect biological functionality in living systems. Advanced label-free optical imaging techniques based on nonlinear optical phenomena overcome these limitations of fluorescence microscopy. We have developed a novel label-free multimodal multiphoton nonlinear optical imaging system based on a near-IR femtosecond laser with photonic crystal fiber and pulse shaper. This highly integrated system offers numerous label-free techniques including third harmonic generation, three-photon excited fluorescence, second harmonic generation, two-photon excited fluorescence, fluorescence lifetime imaging, and broadband coherent anti-Stokes Raman scattering microspectroscopy in one platform. All of the nonlinear signals are spectrally separated by dichroic filters and simultaneously measured by photomultiplier tubes. Moreover, this system includes phase-variance optical coherence tomography as well to enable vascular imaging. We have applied our system to investigate processes in numerous biological samples. Our imaging technique is highly integrated and time efficient to generate big data, offering an array of biomolecular information at one time without staining, three-dimensional sub-micron resolution with deeper penetration, and less photodamage. The big data output from this system is analyzed by multivariate analysis such as principal component analysis and hierarchical cluster analysis. Therefore, this novel technology and methodology will have a great impact on fast in vivo label-free biomedical imaging as a big data generator.

Fluorescence hyperspectral imaging is a powerful tool for in vivo biological studies. The ability to recover the full spectra of the fluorophores allows accurate classification of different structures and study of the dynamic behaviors during various biological processes. However, most existing methods require significant instrument modifications and/or suffer from image acquisition rates too low for compatibility with in vivo imaging. In the present work, a fast (up to 18 frames per second) hyperspectral two-photon fluorescence microscopy approach was demonstrated. Utilizing the beamscanning hardware inherent in conventional multi-photon microscopy, the angle dependence of the generated fluorescence signal as a function beam’s position allowed the system to probe of a different potion of the spectrum at every single scanning line. An iterative algorithm to classify the fluorophores recovered spectra with up to 2,400 channels using a custom high-speed 16-channel photon multiplier tube array. Several dynamic samples including live fluorescent labeled C. elegans were imaged at video rate. Fluorescence spectra recovered using no a priori spectral information agreed well with those obtained by fluorimetry. This system required minimal changes to most existing beam-scanning multi-photon fluorescence microscopes, already accessible in many research facilities.

We present a spatial frequency domain imaging (SFDI) study of local hemodynamics in the human finger cuticle of healthy volunteers performing paced breathing and the forearm of healthy young adults performing normal breathing with our recently developed Real Time Single Snapshot Multiple Frequency Demodulation – Spatial Frequency Domain Imaging (SSMD-SFDI) system. A two-layer model was used to map the concentrations of deoxy-, oxy-hemoglobin, melanin, epidermal thickness and scattering properties at the subsurface of the forearm and the finger cuticle. The oscillations of the concentrations of deoxy- and oxy-hemoglobin at the subsurface of the finger cuticle and forearm induced by paced breathing and normal breathing, respectively, were found to be close to out-of-phase, attributed to the dominance of the blood flow modulation by paced breathing or heartbeat. Our results suggest that the real time SFDI platform may serve as one effective imaging modality for microcirculation monitoring.

Dual comb spectroscopy (DCS) is based on the combination of Fourier transform spectroscopy with an optical frequency comb (OFC), and has a spectral resolution of MHz order over a spectral range of several tens THz. Furthermore, nonmechanical time-delay scanning enables the rapid data acquisition. While DCS imaging is required for hyperspectral imaging, an imaging sensor cannot be used for DCS imaging because of a slow response time compared to the temporal evolution of interferogram signal. Therefore, it is essential to acquire the interferogram signal by use of a single-channel detector while scanning the sample position or the focal point. If DCS imaging can be achieved without the need for such scanning, the application field of the DCS imaging will be largely expanded. One promising method to achieve the scanless imaging is a single-pixel imaging (SPI). SPI enables scan-less imaging by use of optical spatial coding on the sample with a single-channel detector. Also, the spatial averaging effect improves an image quality.

In this paper, we present combination of DCS with SPI, namely a DCS-SPI. DCS-SPI provides 12,000 mode-resolved hyperspectral images in both amplitude and phase at a spatial resolution of 46 μm without the need for mechanical scanning. Furthermore, we determined thickness of a chromium layer from a phase image in the near-infrared wavelength region.

Digital holography is a well-known method to perform three-dimensional imaging by recording the light wavefront information originating from the object. Not only the intensity, but also the phase distribution of the wavefront can then be computed from the recorded hologram in the numerical reconstruction process. However, the reconstructions via the traditional methods suffer from various artifacts caused by twin-image, zero-order term, and noise from image sensors. Here we demonstrate that an end-to-end deep neural network (DNN) can learn to perform both intensity and phase recovery directly from an intensity-only hologram. We experimentally show that the artifacts can be effectively suppressed. Meanwhile, our network doesn’t need any preprocessing for initialization, and is comparably fast to train and test, in comparison with the recently published learning-based method. In addition, we validate that the performance improvement can be achieved by introducing a prior on sparsity.

In this paper a novel two-stage adaptive framework for denoising of differential interference contrast (DIC) images followed by Gabor based gray-level co-occurrence matrix (GLCM) feature extraction methodology is proposed. The first stage consists of a hybrid cascade of anisotropic diffusion denoising (Perona–Malik diffusion) and unsharp masking (USM) based detail enhancement filter to remove noise from DIC images without losing significant morphological features of healthy and precancerous tissues while enlarging the image sharpness. The hybrid filter parameters are obtained by joint stochastic optimization of the image quality metrics. The estimated denoised image with the highest signal to noise ratio (SNR) from Stage I, is used for subsequent textural feature extraction. GLCM window considers neighborhood blocks with similar local statistics with well-preserved local structures between a pixel texture and its nearest neighbors. The efficacy of our denoised DIC imaging with Gabor based GLCM feature descriptors in analysis of healthy and precancerous tissues is experimentally validated for its competitive denoising performance and detail structure preservation of DIC images. The relative change in magnitude and phase information as manifested from Gabor filter coupled with GLCM based spatial statistical measures of tissues as cancer progress validates the adequacy of the proposed scheme for its early stage cancer detection ability in cervical tissues.

In minimal invasive laser osteotomy precise information about the ablation process can be obtained with LIBS in order to avoid carbonization, or cutting of wrong types of tissue. Therefore, the collecting fiber for LIBS needs to be optimally placed in narrow cavities in the endoscope. To determine this optimal placement, the plasma plume expansion dynamics in ablation of bone tissue by the second harmonic of a nanosecond Nd:YAG laser at 532 nm has been studied. The laserinduced plasma plume was monitored in different time delays, from one nanosecond up to one hundred microseconds. Measurements were performed using high-speed gated illumination imaging. The expansion features were studied using illumination of the overall visible emission by using a gated intensified charged coupled device (ICCD). The camera was capable of having a minimum gate width (Optical FWHM) of 3 ns and the timing resolution (minimum temporal shift of the gate) of 10 ps. The imaging data were used to generate position–time data of the luminous plasma-front. Moreover, the velocity of the plasma plume expansion was studied based on the time-resolved intensity data. By knowing the plasma plume profile over time, the optimum position (axial distance from the laser spot) of the collecting fiber and optimal time delay (to have the best signal to noise ratio) in spatial-resolved and time-resolved laser-induced breakdown spectroscopy (LIBS) can be determined. Additionally, the function of plasma plume expansion could be used to study the shock wave of the plasma plume.

Skyline computation is a method of extracting interesting entries from a large population with multiple attributes. These entries, called skyline or Pareto optimal entries, are known to have extreme characteristics that cannot be found by outlier detection methods. Skyline computation is an important task for characterizing large amounts of data and selecting interesting entries with extreme features. When the population changes dynamically, the task of calculating a sequence of skyline sets is called continuous skyline computation. This task is known to be difficult to perform for the following reasons: (1) information of non-skyline entries must be stored since they may join the skyline in the future; (2) the appearance or disappearance of even a single entry can change the skyline drastically; (3) it is difficult to adopt a geometric acceleration algorithm for skyline computation tasks with high-dimensional datasets. Our new algorithm called jointed rooted-tree (JR-tree) manages entries using a rooted tree structure. JR-tree delays extend the tree to deep levels to accelerate tree construction and traversal. In this study, we presented the difficulties in extracting entries tagged with a rare label in high-dimensional space and the potential of fast skyline computation in low-latency cell identification technology.

Diffuse optical tomography (DOT) as a new functional imaging has important clinical applications in many aspects such as benign and malignant breast tumor detection, tumor staging and so on. For quantitative detection of breast tumor, a three-wavelength continuous-wave DOT prototype system combined the ultra-high sensitivity of the photon-counting detection and the measurement parallelism of the lock-in technique was developed to provide high temporal resolution, high sensitivity, large dynamic detection range and signal-to-noise ratio. Additionally, a CT-analogous scanning mode was proposed to cost-effectively increase the detection data. To evaluate the feasibility of the system, a series of assessments were conducted. The results demonstrate that the system can obtain high linearity, stability and negligible inter-wavelength crosstalk. The preliminary phantom experiments show the absorption coefficient is able to be successfully reconstructed, indicating that the system is one of the ideal platforms for optical breast tumor detection.

This contribution proposes a novel extension of the existing ‘Hyper Kurtosis based Modified Duo-Histogram Equalization’ (HKMDHE) algorithm, for multi-objective contrast enhancement of biomedical images. A novel modified objective function has been formulated by joint optimization of the individual histogram equalization objectives. The optimal adequacy of the proposed methodology with respect to image quality metrics such as brightness preserving abilities, peak signal-to-noise ratio (PSNR), Structural Similarity Index (SSIM) and universal image quality metric has been experimentally validated. The performance analysis of the proposed Stochastic HKMDHE with existing histogram equalization methodologies like Global Histogram Equalization (GHE) and Contrast Limited Adaptive Histogram Equalization (CLAHE) has been given for comparative evaluation.

We present a first evaluation of a new multi-exposure laser speckle contrast imaging (MELSCI) system for assessing spatial variations in the microcirculatory perfusion. The MELSCI system is based on a 1000 frames per second 1-megapixel camera connected to a field programmable gate arrays (FPGA) capable of producing MELSCI data in realtime. The imaging system is evaluated against a single point laser Doppler flowmetry (LDF) system during occlusionrelease provocations of the arm in five subjects. Perfusion is calculated from MELSCI data using current state-of-the-art inverse models. The analysis displayed a good agreement between measured and modeled data, with an average error below 6%. This strongly indicates that the applied model is capable of accurately describing the MELSCI data and that the acquired data is of high quality. Comparing readings from the occlusion-release provocation showed that the MELSCI perfusion was significantly correlated (R=0.83) to the single point LDF perfusion, clearly outperforming perfusion estimations based on a single exposure time. We conclude that the MELSCI system provides blood flow images of enhanced quality, taking us one step closer to a system that accurately can monitor dynamic changes in skin perfusion over a large area in real-time.