KEYWORDS: Education and training, Data modeling, Optical properties, Finite element methods, Scattering, 3D modeling, Reconstruction algorithms, Tissues, Absorption, Voxels
SignificanceFrequency-domain diffuse optical tomography (FD-DOT) could enhance clinical breast tumor characterization. However, conventional diffuse optical tomography (DOT) image reconstruction algorithms require case-by-case expert tuning and are too computationally intensive to provide feedback during a scan. Deep learning (DL) algorithms front-load computational and tuning costs, enabling high-speed, high-fidelity FD-DOT.AimWe aim to demonstrate a simultaneous reconstruction of three-dimensional absorption and reduced scattering coefficients using DL-FD-DOT, with a view toward real-time imaging with a handheld probe.ApproachA DL model was trained to solve the DOT inverse problem using a realistically simulated FD-DOT dataset emulating a handheld probe for human breast imaging and tested using both synthetic and experimental data.ResultsOver a test set of 300 simulated tissue phantoms for absorption and scattering reconstructions, the DL-DOT model reduced the root mean square error by 12%±40% and 23%±40%, increased the spatial similarity by 17%±17% and 9%±15%, increased the anomaly contrast accuracy by 9%±9% (μa), and reduced the crosstalk by 5%±18% and 7%±11%, respectively, compared with model-based tomography. The average reconstruction time was reduced from 3.8 min to 0.02 s for a single reconstruction. The model was successfully verified using two tumor-emulating optical phantoms.ConclusionsThere is clinical potential for real-time functional imaging of human breast tissue using DL and FD-DOT.
Molecular imaging tools that can image plant metabolism and effects of external agricultural treatments in the micro-environment of plant tissues are significant for further understanding plant biology and optimizing the formulation of new agricultural products. Mass spectrometry, a common tool used by plant biologists, is unable to resolve nano-crystalline active ingredients (AIs) on the leaf surface nor achieve 3D molecular imaging of living plants. To address that, multiphoton microscopy (MPM) and fluorescence lifetime imaging microscopy (FLIM) are combined to achieve sub-cellular, depth-resolved fluorescence lifetime of both AIs and intrinsic proteins/pigments (e.g., chlorophyll and/or cytosolic NADH) after the herbicide treatment application. Here we present a method using a custom-designed, high-speed MPM-FLIM system, “Instant FLIM”, to achieve real-time, unlabeled 3D functional molecular imaging of intrinsic proteins and pigments in optically thick and highly scattering plant samples with the application of external treatments. To validate the capability of MPM-FLIM to measure intrinsic proteins and pigments within plant tissues, we present the results of unlabeled bluegrass blades samples. To demonstrate simultaneous imaging of 3D molecular plant tissue and the agricultural AI nano-crystals deposition and formation, we evaluate the performance of the MPM-FLIM by applying commercial herbicide product to gamagrass blade sample. Additionally, to measure the herbicide-induced cellular-level functional responses within living plant tissues, 3D time-resolved molecular MPM-FLIM imaging of hemp dogbane leaf with herbicide is performed. Results demonstrate MPM-FLIM is capable of 3D simultaneous functional imaging of label-free living plant tissues and the quantitative measurements of the location and formation of AI nanocrystals within the plant tissues.
SignificanceMachine learning (ML) models based on deep convolutional neural networks have been used to significantly increase microscopy resolution, speed [signal-to-noise ratio (SNR)], and data interpretation. The bottleneck in developing effective ML systems is often the need to acquire large datasets to train the neural network. We demonstrate how adding a “dense encoder-decoder” (DenseED) block can be used to effectively train a neural network that produces super-resolution (SR) images from conventional microscopy diffraction-limited (DL) images trained using a small dataset [15 fields of view (FOVs)].AimThe ML helps to retrieve SR information from a DL image when trained with a massive training dataset. The aim of this work is to demonstrate a neural network that estimates SR images from DL images using modifications that enable training with a small dataset.ApproachWe employ “DenseED” blocks in existing SR ML network architectures. DenseED blocks use a dense layer that concatenates features from the previous convolutional layer to the next convolutional layer. DenseED blocks in fully convolutional networks (FCNs) estimate the SR images when trained with a small training dataset (15 FOVs) of human cells from the Widefield2SIM dataset and in fluorescent-labeled fixed bovine pulmonary artery endothelial cells samples.ResultsConventional ML models without DenseED blocks trained on small datasets fail to accurately estimate SR images while models including the DenseED blocks can. The average peak SNR (PSNR) and resolution improvements achieved by networks containing DenseED blocks are ≈3.2 dB and 2 × , respectively. We evaluated various configurations of target image generation methods (e.g., experimentally captured a target and computationally generated target) that are used to train FCNs with and without DenseED blocks and showed that including DenseED blocks in simple FCNs outperforms compared to simple FCNs without DenseED blocks.ConclusionsDenseED blocks in neural networks show accurate extraction of SR images even if the ML model is trained with a small training dataset of 15 FOVs. This approach shows that microscopy applications can use DenseED blocks to train on smaller datasets that are application-specific imaging platforms and there is promise for applying this to other imaging modalities, such as MRI/x-ray, etc.
Frequency domain (FD) diffuse optical spectroscopy (DOS) can be used to recover absolute optical properties of biological tissue, providing valuable clinical feedback, including in diagnosis and monitoring of breast tumours. In this study, tomographic (3D) and topographic (2D) techniques for spatially-varying optical parameter recovery are presented, based on a multi-distance, handheld DOS probe. Processing pipelines and reconstruction quality are discussed and quantitatively compared, demonstrating the trade-offs between depth sensitivity, optical contrast, and computational speed. Together, the two techniques provide both depth sensitive real-time feedback, and high-resolution 3D reconstruction from a single set of measurements, enabling faster and more accurate clinical feedback.
Structured interrogation (SI) is a frequency-domain NIRS technique that adjusts the relative phase between intensity-modulated light sources and generates a various spatial pattern of depth sensitivity. It can be used to extract quantitative information from multi-layered tissues. We apply Cramér-Rao lower bound (CRLB) analysis for phase selection. Then, the selected SI phases are used to resolve the optical properties of a fat/muscle tissue model. We found that CRLB is effective for selecting SI phase shift, and a combination of SI measurements and multi-distance measurements can result in an accurate estimation of optical properties and allow for more compact fd-NIRS probes.
We present a new frequency-domain near-infrared spectroscopy method for extracting deep tissue information in heterogeneous tissues by interfering multiple intensity-modulated light sources. This method of structured interrogation (SI) can alter the spatial distribution of depth sensitivity, and therefore target different depths in the tissue. We found through analytical theory and a simulation study that SI has enhanced sensitivity to dynamic changes in deeper tissues and reduced sensitivity to superficial layers compared to multi-distance (MD) measurements. We also demonstrate that SI phase-only measurements are sufficient to accurately estimate optical properties.
Fluorescence lifetime imaging microscopy (FLIM) is an important technique to understand the chemical microenvironment in cells and tissues since it provides additional contrast compared to conventional fluorescence imaging. When two fluorophores within a diffraction limit are excited, the resulting emission leads to nonlinear spatial distortion and localization effects in intensity (magnitude) and lifetime (phase) components. To address this issue, in this work, we provide a theoretical model for convolution in FLIM to describe how the resulting behavior differs from conventional fluorescence microscopy. We then present a Richardson-Lucy (RL) based deconvolution including total variation (TV) regularization method to correct for the distortions in FLIM measurements due to optical convolution, and experimentally demonstrate this FLIM deconvolution method on a multi-photon microscopy (MPM)-FLIM images of fluorescent-labeled fixed bovine pulmonary arterial endothelial (BPAE) cells.
KEYWORDS: 3D image processing, Microscopy, Compressed sensing, Luminescence, In vivo imaging, Signal to noise ratio, Confocal microscopy, Reconstruction algorithms, Microscopes, 3D modeling
Fluorescence microscopy has been a significant tool to observe long-term imaging of embryos (in vivo) growth over time. However, cumulative exposure is phototoxic to such sensitive live samples. While techniques like light-sheet fluorescence microscopy (LSFM) allows for reduced exposure, it is not well suited for deep imaging models. Other computational techniques are computationally expensive and often lack restoration quality. To address this challenge, one can use various low-dosage imaging techniques that are developed to achieve the 3D volume reconstruction using a few slices in the axial direction (z-axis); however, they often lack restoration quality. Also, acquiring dense images (with small steps) in the axial direction is computationally expensive. To address this challenge, we present a compressive sensing (CS) based approach to fully reconstruct 3D volumes with the same signal-to-noise ratio (SNR) with less than half of the excitation dosage. We present the theory and experimentally validate the approach. To demonstrate our technique, we capture a 3D volume of the RFP labeled neurons in the zebrafish embryo spinal cord (30 μm thickness) with the axial sampling of 0.1 μm using a confocal microscope. From the results, we observe the CS-based approach achieves accurate 3D volume reconstruction from less than 20% of the entire stack optical sections. The developed CS-based methodology in this work can be easily applied to other deep imaging modalities such as two-photon and light-sheet microscopy, where reducing sample photo-toxicity is a critical challenge.
KEYWORDS: Fluorescence lifetime imaging, Microscopy, 3D acquisition, 3D image processing, Multiphoton microscopy, Luminescence, In vivo imaging, 3D modeling, Tissues, Super resolution
Fluorescence lifetime imaging microscopy (FLIM) adds an additional dimension to fluorescence microscopy by measuring the fluorophore interactions with the microenvironment in addition to all the benefits and power of fluorescence microscopy. Real-time FLIM, however, requires overcoming unique technical challenges to achieve similar imaging speeds as can be achieved through standard in vivo microscopy techniques. This talk will present an “InstantFLIM” system that achieves real-time (acquisition and processing), super-resolution, 3D in vivo multiphoton FLIM by overcoming these limitations. The system is demonstrated in intact-skull mouse and zebrafish brain imaging models, and 3D autofluorescence FLIM of highly scattering plant tissue.
We propose and demonstrate the first analytical model of the spatial resolution of frequency-domain (FD) fluorescence lifetime imaging microscopy (FLIM) that explains how it is fundamentally different with the common resolution limit of the conventional fluorescence microscopy. Frequency modulation (FM) capture effect is also observed by the model, which results in distorted FLIM measurements. A super-resolution FLIM approach based on a localization-based technique, super-resolution radial fluctuations (SRRF), is presented. In this approach, we separately process the intensity and lifetime to generate a super-resolution FLIM composite image. The capability of the approach is validated both numerically and experimentally in fixed cells sample.
KEYWORDS: Fluorescence lifetime imaging, Image segmentation, Microscopy, Denoising, In vivo imaging, Luminescence, Convolutional neural networks, Signal to noise ratio, Imaging systems, Signal processing
Fluorescence lifetime imaging microscopy (FLIM) systems are limited by their slow processing speed, low signal- to-noise ratio (SNR), and expensive and challenging hardware setups. In this work, we demonstrate applying a denoising convolutional network to improve FLIM SNR. The network will integrated with an instant FLIM system with fast data acquisition based on analog signal processing, high SNR using high-efficiency pulse-modulation, and cost-effective implementation utilizing off-the-shelf radio-frequency components. Our instant FLIM system simultaneously provides the intensity, lifetime, and phasor plots in vivo and ex vivo. By integrating image de- noising using the trained deep learning model on the FLIM data, provide accurate FLIM phasor measurements are obtained. The enhanced phasor is then passed through the K-means clustering segmentation method, an unbiased and unsupervised machine learning technique to separate different fluorophores accurately. Our experimental in vivo mouse kidney results indicate that introducing the deep learning image denoising model before the segmentation effectively removes the noise in the phasor compared to existing methods and provides clearer segments. Hence, the proposed deep learning-based workflow provides fast and accurate automatic segmentation of fluorescence images using instant FLIM. The denoising operation is effective for the segmentation if the FLIM measurements are noisy. The clustering can effectively enhance the detection of biological structures of interest in biomedical imaging applications.
KEYWORDS: Super resolution, Microscopy, Luminescence, Organisms, Data modeling, X-rays, X-ray imaging, Visualization, Super resolution microscopy, Magnetic resonance imaging
Fluorescence microscopy has enabled a dramatic development in modern biology by visualizing biological organ- isms with micrometer scale resolution. However, due to the diffraction limit, sub-micron/nanometer features are difficult to resolve. While various super-resolution techniques are developed to achieve nanometer-scale resolu- tion, they often either require expensive optical setup or specialized fluorophores. In recent years, deep learning has shown potentials to reduce the technical barrier and obtain super-resolution from diffraction-limited images. For accurate results, conventional deep learning techniques require thousands of images as a training dataset. Obtaining large datasets from biological samples is not often feasible due to photobleaching of fluorophores, phototoxicity, and dynamic processes occurring within the organism. Therefore, achieving deep learning-based super-resolution using small datasets is challenging. We address this limitation with a new convolutional neural network based approach that is successfully trained with small datasets and achieves super-resolution images. We captured 750 images in total from 15 different field-of-views as the training dataset to demonstrate the technique. In each FOV, a single target image is generated using the super-resolution radial fluctuation method. As expected, this small dataset failed to produce a usable model using traditional super-resolution architecture. However, using the new approach, a network can be trained to achieve super-resolution images from this small dataset. This deep learning model can be applied to other biomedical imaging modalities such as MRI and X-ray imaging, where obtaining large training datasets is challenging.
In this paper, we show that interfering multiple photon density waves created by intensity-modulated sources in frequency domain diffuse optical spectroscopy (fd-DOS) can be used to recover the optical properties of homogenous and heterogeneous tissues. While fd-DOS can recover the optical properties of homogenous tissue using a single source-detector pair, heterogeneous or layered tissues such as breast, brain, and skin require additional source-detector pairs with multiple separations. Through modelling, we show that the varying illumination patterns created by the interference of two intensity modulated sources can be used to recover the optical properties of two-layer tissue using only a single detector and two phased sources. Two-dimensional fd-DOS models of the conventional multi-distance and proposed multi-phase approaches were compared for homogenous and two-layered tissues. In a homogenous tissue with absorption and reduced scattering coefficients representative of human breast, the simulation results showed that both multi-distance and multi-phase approaches are capable of recovering the absorption and reduced scattering coefficients of the tissue. However, the multi-phase approach has less precision than the conventional multidistance approach. In the two-layer model, the multi-phase approach was capable of recovering the optical properties of both layers, while the multi-distance approach could not.
We propose and demonstrate a novel multiphoton frequency-domain fluorescence lifetime imaging microscopy (MPM-FD-FLIM) system that is able to generate 3D lifetime images in deep scattering tissues. The imaging speed of FD-FLIM is improved using phase multiplexing, where the fluorescence signal is split and mixed with the reference signal from the laser in a multiplexing manner. The system allows for easy generation of phasor plots, which not only address multi-exponential decay problems but also clearly represent the dynamics of the fluorophores being investigated. Lastly, a sensorless adaptive optics setup is used for FLIM imaging in deep scattering tissues. The capability of the system is demonstrated in fixed and living animal models, including mice and zebrafish.
We present the first experimental demonstration of super-resolution multiphoton frequency-domain (FD) fluorescence lifetime imaging microscopy (FLIM). This is obtained through a novel microscopy technique called generalized stepwise optical saturation (GSOS). GSOS√utilizes the linear combination of M steps of raw images to improve the imaging resolution by a factor of √M . Here, a super-resolution multiphoton FD-FLIM is demonstrated on various samples, including fixed cells and biological tissues, with a custom-built two-photon FD-FLIM microscope. We demonstrate simultaneous super-resolution intensity and fluorescence lifetime images of a variety of cell cultures and ex vivo tissues. Combined with multiphoton excitation, the proposed GSOS microscopy is able to generate super-resolution FLIM images deep in scattering samples.
In this report the development of a low SWAP-C mid-infrared imaging system is outlined. This system is designed with a commercially available Vanadium-oxide (VOx) microbolometer camera costing $250 controlled via a Raspberry Pi (RPi) and Python. The camera used was previously characterized to have a NEDT of 25 mK for an integration time of 3.43s at 7 Hz framerate, but from software modifications discussed in this paper it was found that the cameras framerate can be pushed to 32 Hz lowering this integration time to 0.75 s. Due to the low SWAP-C characteristics of the design and the plug-and-play nature of the camera paired with Python code, this system can enable MIR imaging applications that are currently limited by the SWAPC characteristics of currently available detection systems. After outlining the development of interfacing the camera with a computer/microcontroller, the cameras code is extended to a client-server operation that allows for wireless control of the imaging system. This further enables remote operation for applications such as dronebased monitoring/surveillance or trace explosives detection. The report concludes with the discussion of two potential applications of distributed imaging and spectroscopic organics detection.
Many long-wave infrared spectroscopic imaging applications are limited by the portability and cost of detector arrays. We present a characterization of a newly available, low-cost, uncooled vanadium oxide microbolometer array, the Seek Compact, in accordance with common infrared detector specifications: noise-equivalent differential temperature (NEDT), optical responsivity spectra, and Allan variance. The Compact’s imaging array consists of 156×206 pixels with a 12-μm pixel pitch, 93% of the pixels yield useful temperature readings. Characterization results show optical response between λ=7.4 and 12 μm with an NEDT of 148 mK (at ≈7 fps). Comparing these results to a research-grade camera, the Seek Compact exhibits a 4× and 48× reduction in weight (2.0/0.5 lbs) and cost ($12,000/$250) but takes 93× longer to achieve the same NEDT (1.55 s/16.6 ms for 45 mK). Additionally, a proof-of-concept spectral imaging experiment of SiN thin films is conducted. Leveraging this price reduction and spectroscopic imaging capability, the Seek Compact has potential in enabling field-deployable and distributed active midinfrared spectroscopic imaging, where cost and portability are the dominate inhibitors and high frame rates are not required.
Precise detection of trace amount of molecules, such as the disease biomarkers present in biofluids or explosive residues, requires high sensitivity detection. electrospray ionization–mass spectrometry (ESI-MS) is a common and effective technique for sensitive trace molecular detection in small-volume liquid samples. In ESI-MS, nano-liter volume samples are ionized and aerosolized by ESI, and fed into MS for mass analysis. ESI-MS has proven to be a reliable ionization technique for coupling liquid phase separations like liquid chromatography (LC) and capillary zone electrophoresis (CE) with the highly specific resolving power of MS. While CE and ESI can be performed on a microfluidic chip having a footprint of a few cm2, MS is typically at least 100 times bigger in size than a micro-chip. A reduced size, weight, and power profile would enable semi-portable applications in forensics, environmental monitoring, defense, and biological/pharmaceutical applications. To achieve this goal, we present an initial study evaluating the use of mid-infrared absorption spectroscopy (MIRAS) in place of MS to create a ESI-MIRAS system. To establish feasibility, we perform ESI-MIRAS on phospholipid samples, which have been previously demonstrated to be separable by CE. Phospholipids are biomarkers of degenerative neurological, kidney, and bone diseases and can be found in biofluids such as blood, urine and cerebrospinal fluid. To establish sensitivity limits, calibration samples of 100 μM concentration are electrospray deposited on to a grounded Si wafer for different times (1 minutes to 4 minutes with a 1 minute step). The minimum detectable concentration-time product, where a FTIR globar is used as the MIR source, is found ~200 μM·s.
Multiphoton microscopy (MPM) imaging of intrinsic two-photon excited fluorescence (TPEF) is performed on humanized sickle cell disease (SCD) mouse model splenic tissue. Distinct morphological and spectral features associated with SCD are identified and discussed in terms of diagnostic relevance. Specifically, spectrally unique splenic iron-complex deposits are identified by MPM; this finding is supported by TPEF spectroscopy and object size to standard histopathological methods. Further, iron deposits are found at higher concentrations in diseased tissue than in healthy tissue by all imaging methods employed here including MPM, and therefore, may provide a useful biomarker related to the disease state. These newly characterized biomarkers allow for further investigations of SCD in live animals as a means to gain insight into the mechanisms impacting immune dysregulation and organ malfunction, which are currently not well understood.
We present our study on compact, label-free dissolved lipid sensing by combining capillary electrophoresis
separation in a PDMS microfluidic chip online with mid-infrared (MIR) absorption spectroscopy for biomarker
detection. On-chip capillary electrophoresis is used to separate the biomarkers without introducing any extrinsic
contrast agent, which reduces both cost and complexity. The label free biomarker detection could be done by
interrogating separated biomarkers in the channel by MIR absorption spectroscopy. Phospholipids biomarkers of
degenerative neurological, kidney, and bone diseases are detectable using this label free technique. These
phospholipids exhibit strong absorption resonances in the MIR and are present in biofluids including urine, blood
plasma, and cerebrospinal fluid. MIR spectroscopy of a 12-carbon chain phosphatidic acid (PA) (1,2-dilauroyl-snglycero-
3-phosphate (sodium salt)) dissolved in N-methylformamide, exhibits a strong amide peak near
wavenumber 1660 cm-1 (wavelength 6 μm), arising from the phosphate headgroup vibrations within a low-loss
window of the solvent. PA has a similar structure to many important phospholipids molecules like
phosphatidylcholine (PC), phosphatidylinositol (PI), phosphatidylethanolamine (PE), phosphatidylglycerol (PG),
and phosphatidylserine (PS), making it an ideal molecule for initial proof-of-concept studies. This newly proposed
detection technique can lead us to minimal sample preparation and is capable of identifying several biomarkers from
the same sample simultaneously.
KEYWORDS: Luminescence, Two photon imaging, Hydrogen, Absorption, Oxygen, Sapphire lasers, Confocal microscopy, Medical research, Molecules, In vivo imaging
We present the application of two-photon fluorescence (TPF) imaging to monitor intracellular hydrogen peroxide (H 2 O 2 ) production in brain cells. For selective imaging of H 2 O 2 over other reactive oxygen species, we employed small-molecule fluorescent probes that utilize a chemoselective boronate deprotection mechanism. Peroxyfluor-6 acetoxymethyl ester detects global cellular H 2 O 2 and mitochondria peroxy yellow 1 detects mitochondrial H 2 O 2 . Two-photon absorption cross sections for these H 2 O 2 probes are measured with a mode-locked Ti:sapphire laser in the wavelength range of 720 to 1040 nm. TPF imaging is demonstrated in the HT22 cell line to monitor both cytoplasmic H 2 O 2 and localized H 2 O 2 production in mitochondria. Endogenous cytoplasmic H 2 O 2 production is detected with TPF imaging in rat astrocytes modified with d-amino acid oxidase. The TPF H 2 O 2 imaging demonstrated that these chemoselective probes are powerful tools for the detection of intracellular H 2 O 2 .
Up/down regulation of microRNA panels has been correlated to cardiovascular diseases and cancer. Frequent
miRNA profiling at home can hence allow early cancer diagnosis and home-use chronic disease monitoring, thus
reducing both mortality rate and healthcare cost. However, lifetime of miRNAs is less than 1 hour without preservation
and their concentrations range from pM to mM. Despite rapid progress in the last decade, modern nucleic acid analysis
methods still do not allow personalized miRNA profiling---Real-time PCR and DNA micro-array both require elaborate
miRNA preservation steps and expensive equipment and nano pore sensors cannot selectively quantify a large panel
with a large dynamic range.
We report a novel and low-cost optical fiber sensing platform, which has the potential to profile a panel of
miRNA with simple LED light sources and detectors. The individual tips of an optical imaging fiber bundle (mm in
diameter with 7000 fiber cores) were etched into cones with 10 nm radius of curvature and coated with Au. FRET
(Forster Resonant Energy Transfer) hairpin oligo probes, with the loop complementary to a specific miRNA that can
release the hairpin, were functionalized onto the conic tips. Exciting light in the optical fiber waveguide is optimally
coupled to surface plasmonics on the gold surface, which then converges to the conic tips with two orders of magnitude enhancement in intensity. Unlike nanoparticle plasmonics, tip plasmonics can be excited over a large band width and hence the plasmonic enhanced fluorescence signal of the FRET reporter is also focused towards the tip--- and is further enhanced with the periodic resonant grid of the fiber array which gives rise to pronounced standing wave interference patterns. Multiplexing is realized by functionalizing different probes onto one fiber bundle using a photoactivation process.
We present Finite-Difference Time-Domain (FDTD) simulations to explore feasibility of chip-to-chip waveguide
coupling via Optical Quilt Packaging (OQP). OQP is a newly proposed scheme for wide-bandwidth, highly-efficient
waveguide coupling and is suitable for direct optical interconnect between semiconductor optical sources, optical
waveguides, and detectors via waveguides. This approach leverages advances in quilt packaging (QP), an electronic
packaging technique wherein contacts formed along the vertical faces are joined to form electrically-conductive and
mechanically-stable chip-to-chip contacts. In OQP, waveguides of separate substrates are aligned with sub-micron
accuracy by protruding lithographically-defined copper nodules on the side of a chip. With OQP, high efficiency chip-to-chip
optical coupling can be achieved by aligning waveguides of separate chips with sub-micron accuracy and reducing
chip-to-chip distance. We used MEEP (MIT Electromagnetic Equation Propagation) to investigate the feasibility of OQP
by calculating the optical coupling loss between butt coupled waveguides. Transmission between a typical QCL ridge
waveguide and a single-mode Ge-on-Si waveguide was calculated to exceed 65% when an interchip gap of 0.5 μm and
to be no worse than 20% for a gap of less than 4 μm. These results compare favorably to conventional off-chip coupling.
To further increase the coupling efficiency and reduce sensitivity to alignment, we used a horn-shaped Ge-on-Si
waveguide and found a 13% increase in coupling efficiency when the horn is 1.5 times wider than the wavelength and 2
times longer than the wavelength. Also when the horizontal misalignment increases, coupling loss of the horn-shaped
waveguide increases at a slower rate than a ridge waveguide.
MOCVD grown quantum cascade lasers (QCLs) have demonstrated about the same level performance as MBE
grown QCLs. With the regrowth capability to fabricate buried heterostructure (BH) waveguides, the QCL output
power has been dramatically increased and that opens the door to many mid-IR (and THz) applications. With the
stable and high growth rate to produce high performance and reliable BH lasers, commercialization of QCLs with
reasonable qualification and affordable price becomes possible. Furthermore with a good gain material and the
etching and regrowth capability, optoelectronic integration can be realized using MOCVD growth techniques. We
compare the MBE and MOCVD growth techniques and discuss important issues on growth rate stabilization and the
control of growth quality at the hetero-interface. We also go over a few growth and integration examples we are
working on that are preferentially done by MOCVD. Finally we describe a detailed QCL BH regrowth study and
discussed how that can be done right.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.