Extended light-source-based lensless microscopy using constrained and regularized reconstruction.

Lensless microscopy is a simple, portable, and cost-effective method of microscopy. It has been extensively investigated with coherent light sources such as laser setups or partially coherent light sources such as an LED filtered with a pinhole aperture. The coherence of light has a direct influence on the resolution of the reconstructed object. This paper presents lensless microscopy with a spatially extended light source (white LED flashlight without any subsequent engineering). A reconstruction method based on constrained and regularized optimization is presented. The resolution reduced due to decreased coherence is gained by the presented method of object estimation.

Generalized curved beam back-projection method for near-infrared imaging using banana function.

Near-infrared imaging (NIRI) is a sub-surface imaging that makes a trade-off in recovery accuracy with depth of penetration. On the other hand, diffuse optical tomography (DOT) images tissue up to several centimeters. However, DOT reconstruction has a stability issue due to the inverse problem. This paper proposes a generalized continuous-wave technique to image objects of dimensions 4-6 cm comparable to DOT. A nonlinear Rosenbrock's banana function is fitted to the approximate photon path, and the fit parameter thus obtained gives the penetration depth of each channel. The calculated values of absorption change are back-projected along these curved paths for reconstruction without solving the inverse problem. This function serves as an operator for image reconstruction. Here numerical simulations, experimental validation on wax phantom with inclusions, finger joint, and degraded apple have been performed to show potential of the proposed method in imaging. Thus this computationally efficient, reliable, and simple method is suitable for practical and real-time NIRI applications.

A computer-aided tool for automatic volume estimation of hematoma using non-contrast brain CT scans.

The computation of hematoma volume is the key parameter for treatment planning of Intracerebral hemorrhage (ICH). Non-contrast computed tomography (NCCT) imaging is routinely used for the diagnosis of ICH. Hence, the development of computer-aided tools for three-dimensional (3D) computed tomography (CT) image analysis is essential to estimate the gross volume of hematoma. We propose a methodology for automatic estimation of the hematoma volume from 3D CT volumes. Our approach integrates two different methods, multiple abstract splitting (MAS) and seeded region growing (SRG) to develop a unified hematoma detection pipeline from pre-processed CT volumes. The proposed methodology was tested on 80 cases. The volume was estimated from the delineated hematoma region, validated against the ground-truth volumes, and compared with those obtained from the conventional ABC/2 approach. We also compared our results with the U-Net model (supervised technique) to show the applicability of the proposed method. The volume calculated from manually segmented hematoma was considered the ground truth. TheR2correlation coefficient between the volume obtained from the proposed algorithm and the ground truth is 0.86, which is equivalent to theR2value resulting from the comparison between the volume calculated by ABC/2 and the ground truth. The experimental results of the proposed unsupervised approach are comparable to the deep neural architecture (U-Net models). The average computation time was 132.76 ± 14 seconds. The proposed methodology provides a fast and automatic estimation of hematoma volume, which is similar to the baseline user-guided ABC/2 approach. Implementation of our method does not demand a high-end computational setup. Thus, recommended in clinical practice for computer-assistive volume estimation of hematoma from 3D CT volumes and can be implemented in a simple computer system.

Radius-optimized efficient template matching for lesion detection from brain images.

Computer-aided detection of brain lesions from volumetric magnetic resonance imaging (MRI) is in demand for fast and automatic diagnosis of neural diseases. The template-matching technique can provide satisfactory outcome for automatic localization of brain lesions; however, finding the optimal template size that maximizes similarity of the template and the lesion remains challenging. This increases the complexity of the algorithm and the requirement for computational resources, while processing large MRI volumes with three-dimensional (3D) templates. Hence, reducing the computational complexity of template matching is needed. In this paper, we first propose a mathematical framework for computing the normalized cross-correlation coefficient (NCCC) as the similarity measure between the MRI volume and approximated 3D Gaussian template with linear time complexity, [Formula: see text], as opposed to the conventional fast Fourier transform (FFT) based approach with the complexity [Formula: see text], where [Formula: see text] is the number of voxels in the image and [Formula: see text] is the number of tried template radii. We then propose a mathematical formulation to analytically estimate the optimal template radius for each voxel in the image and compute the NCCC with the location-dependent optimal radius, reducing the complexity to [Formula: see text]. We test our methods on one synthetic and two real multiple-sclerosis databases, and compare their performances in lesion detection with FFT and a state-of-the-art lesion prediction algorithm. We demonstrate through our experiments the efficiency of the proposed methods for brain lesion detection and their comparable performance with existing techniques.

Statistical modeling of ultrastructural features of murine dermal collagen under chronic low-dose whole body X-irradiation.

The present study investigated the changes in ultrastructural features of dermal collagen fibrils of mice following exposure to different cumulative chronic low-dose X-irradiation through digital image analysis-based statistical modeling. Pubertal mice were X-irradiated and dorsal skin biopsies were collected and processed for transmission electron microscopic (TEM) analysis. TEM features of collagen fibrils showed alteration in the cross-sectional area, population density and in the axial periodic pattern of light and dark bands. The mathematical analysis of histogram data from TEM images revealed some adaptive behavior in collagen structures of the X-irradiated group. This finding indicated that exposure to chronic low-dose X-radiation induced an altered steady state with adaptive variation in dermal collagen fibrils in irradiated mice.

Simultaneous multi-wavelength oscillation of Nd laser around 1.3 mum: A potential source for coherent terahertz generation.

Simultaneous oscillations of 1318.8nm, 1320.0nm, 1332.6nm, 1335.0nm, 1338.2nm and 1339.0nm in a side, pulsed-diode-laser-array pumped Nd:YAG laser is realized for both free running and Q-switched operation. An average power of 1.1W is obtained for an absorbed pump power of 7.1W with an effective optical slope efficiency of 33%. The difference frequency interactions among these wavelengths may be used to generate radiation in the range 0.13-3.43THz. With the two most intense lines at 1318.8nm and 1338.2nm, it is possible to generate coherent radiation at 3.3THz with numerically estimated peak power of 0.21W in a 1.5mm thick GaSe crystal.

Performance analysis of different wavelet feature vectors in quantification of oral precancerous condition.

This paper presents an automatic method for classification of progressive stages of oral precancerous conditions like oral submucous fibrosis (OSF). The classifier used is a three-layered feed-forward neural network and the feature vector, is formed by calculating the wavelet coefficients. Four wavelet decomposition functions, namely GABOR, HAAR, DB2 and DB4 have been used to extract the feature vector set and their performance has been compared. The samples used are transmission electron microscopic (TEM) images of collagen fibers from oral subepithelial region of normal and OSF patients. The trained network could classify normal fibers from less advanced and advanced stages of OSF successfully.

An object-based coding scheme for frontal surface of defective fluted ingot.

A novel image-based defect identification and coding technique has been proposed for fluted ingots, which are used for the production of locomotive wheels. The edge density map has been used for defect identification and an object-based coding approach has been applied for the storage of defective ingots. The complete scheme has been implemented in one of the integrated steel plants of India.

Lumen Segmentation in Intravascular Optical Coherence Tomography Using Backscattering Tracked and Initialized Random Walks.

Intravascular imaging using ultrasound or optical coherence tomography (OCT) is predominantly used to adjunct clinical information in interventional cardiology. OCT provides high-resolution images for detailed investigation of atherosclerosis-induced thickening of the lumen wall resulting in arterial blockage and triggering acute coronary events. However, the stochastic uncertainty of speckles limits effective visual investigation over large volume of pullback data, and clinicians are challenged by their inability to investigate subtle variations in the lumen topology associated with plaque vulnerability and onset of necrosis. This paper presents a lumen segmentation method using OCT imaging physics-based graph representation of signals and random walks image segmentation approaches. The edge weights in the graph are assigned incorporating OCT signal attenuation physics models. Optical backscattering maxima is tracked along each A-scan of OCT and is subsequently refined using global graylevel statistics and used for initializing seeds for the random walks image segmentation. Accuracy of lumen versus tunica segmentation has been measured on 15 in vitro and 6 in vivo pullbacks, each with 150-200 frames using 1) Cohen's kappa coefficient (0.9786 ±0.0061) measured with respect to cardiologist's annotation and 2) divergence of histogram of the segments computed with Kullback-Leibler (5.17 ±2.39) and Bhattacharya measures (0.56 ±0.28). High segmentation accuracy and consistency substantiates the characteristics of this method to reliably segment lumen across pullbacks in the presence of vulnerability cues and necrotic pool and has a deterministic finite time-complexity. This paper in general also illustrates the development of methods and framework for tissue classification and segmentation incorporating cues of tissue-energy interaction physics in imaging.

A partial reconstruction scheme for continuous wave diffuse optical tomography with reflection geometry.

Image quality and photon measurement with good SNR (signal to noise ratio) in continuous wave diffuse optical tomography depend on the source detector density and sensitivity of photo detector. For large volume objects, it is difficult to obtain detectable light intensity with good SNR over the whole boundary. As an alternative, instead of the full boundary, the measurements are taken over a semi circle as in reflection geometry and a partial reconstruction scheme for the same is proposed in this paper. The cross-sectional optical parameters are reconstructed for different half of the sample with modified boundary conditions and finally the average of all the reconstructions are considered as the final reconstructed image. Simulation and experimental results have been illustrated to validate the proposed method. The main advantage of this scheme is to improve signal to noise ratio which controls the quality of reconstruction in actual phantoms. The use of continuous wave measurement makes the system cost effective as well.

Contrast improvement of continuous wave diffuse optical tomography reconstruction by hybrid approach using least square and genetic algorithm.

Reconstruction of the absorption coefficient of tissue with good contrast is of key importance in functional diffuse optical imaging. A hybrid approach using model-based iterative image reconstruction and a genetic algorithm is proposed to enhance the contrast of the reconstructed image. The proposed method yields an observed contrast of 98.4%, mean square error of 0.638×10⁻³, and object centroid error of (0.001 to 0.22) mm. Experimental validation of the proposed method has also been provided with tissue-like phantoms which shows a significant improvement in image quality and thus establishes the potential of the method for functional diffuse optical tomography reconstruction with continuous wave setup. A case study of finger joint imaging is illustrated as well to show the prospect of the proposed method in clinical diagnosis. The method can also be applied to the concentration measurement of a region of interest in a turbid medium.

Probabilistic graphical modeling of speckle statistics in laser speckle contrast imaging for noninvasive and label-free retinal angiography.

This paper introduces a noninvasive and label-free approach for retinal angiography using Laser speckle contrast imaging (LSCI). Retinal vessel structure is segmented using a Hidden Markov Random Field (HMRF) based model. Prior to that, k-means clustering is used to obtain initial parameter set and labels for HMRF. Final parameter set for HMRF is estimated using expectation-maximization (EM) algorithm and final labeling is achieved using maximum aposteriori (MAP) algorithm. Clique energy for HMRF is computed from eigenvalue analysis of structure tensor for each pixel. This helps to get connectivity in the direction of strongest tangents in its neighborhood, facilitating the tracking of fine vessels in retinal vascular network. Quantitative evaluation shows an average vessel segmentation accuracy of 96.41% in normal condition with substantial improvement in tracking capability of fine vessels. Changes in blood flow can be tracked and observed at segmented output; particularly applicable for the study of different pathological conditions.

A novel shape-based coding-decoding technique for an industrial visual inspection system.

This paper describes a unique single camera-based dimension storage method for image-based measurement. The system has been designed and implemented in one of the integrated steel plants of India. The purpose of the system is to encode the frontal cross-sectional area of an ingot. The encoded data will be stored in a database to facilitate the future manufacturing diagnostic process. The compression efficiency and reconstruction error of the lossy encoding technique have been reported and found to be quite encouraging.

A distortion corrected single camera-based weight estimation technique for industrial objects.

This paper describes a unique single camera-based dimensional measurement with a self-calibration method of image-based measurement. The system has been designed and implemented in one of the integrated steel plants in India. The purpose of the system is to obtain the frontal cross-sectional area of an ingot irrespective of its distance from the camera head. Automatic calibration is achieved by attaching a magnetic template of known area. This self-calibrating system is further refined to correct for the various distortions arising out of lens characteristics. The results obtained through field trials have been reported and found to be quite encouraging.

Robustness of TCA cycle at steady-state: an LMI-based analysis and synthesis framework.

A novel analysis and synthesis framework is devised for synergism and saturation system, commonly known as S-system, for improving the robustness of the TCA cycle. In order to minimize the perturbation sensitivity, a measure of robustness of the network, a new design framework is proposed. The design constraints are formulated in computationally attractive convex optimization framework. The proposed multi-objective optimization problem, framed as Linear Matrix Inequality (LMI), makes a trade-off between the robustness and the control effort of the synthesized TCA cycle.

An optimization-based design framework for steering steady states and improving robustness of glycolysis-glycogenolysis pathway.

A robust synthesis technique is devised for synergism and saturation systems, commonly known as S-systems, for controlling the steady states of the glycolysis-glycogenolysis pathway. The development of the robust biochemical network is essential owing to the fragile response to the perturbation of intrinsic and extrinsic parameters of the nominal S-system. The synthesis problem is formulated in a computationally attractive convex optimization framework. The linear matrix inequalities are framed to aim at the minimization of steady-state error, improvement of robustness, and utilization of minimum control input to the biochemical network.

A closed-loop control scheme for steering steady states of glycolysis and glycogenolysis pathway.

Biochemical networks normally operate in the neighborhood of one of its multiple steady states. It may reach from one steady state to other within a finite time span. In this paper, a closed-loop control scheme is proposed to steer states of the glycolysis and glycogenolysis (GG) pathway from one of its steady states to other. The GG pathway is modeled in the synergism and saturation system formalism, known as S-system. This S-system model is linearized into the controllable Brunovsky canonical form using a feedback linearization technique. For closed-loop control, the linear-quadratic regulator (LQR) and the linear-quadratic gaussian (LQG) regulator are invoked to design a controller for tracking prespecified steady states. In the feedback linearization technique, a global diffeomorphism function is proposed that facilitates in achieving the regulation requirement. The robustness of the regulated GG pathway is studied considering input perturbation and with measurement noise.

Cancer classification from gene expression data by NPPC ensemble.

The most important application of microarray in gene expression analysis is to classify the unknown tissue samples according to their gene expression levels with the help of known sample expression levels. In this paper, we present a nonparallel plane proximal classifier (NPPC) ensemble that ensures high classification accuracy of test samples in a computer-aided diagnosis (CAD) framework than that of a single NPPC model. For each data set only, a few genes are selected by using a mutual information criterion. Then a genetic algorithm-based simultaneous feature and model selection scheme is used to train a number of NPPC expert models in multiple subspaces by maximizing cross-validation accuracy. The members of the ensemble are selected by the performance of the trained models on a validation set. Besides the usual majority voting method, we have introduced minimum average proximity-based decision combiner for NPPC ensemble. The effectiveness of the NPPC ensemble and the proposed new approach of combining decisions for cancer diagnosis are studied and compared with support vector machine (SVM) classifier in a similar framework. Experimental results on cancer data sets show that the NPPC ensemble offers comparable testing accuracy to that of SVM ensemble with reduced training time on average.

Swept-source optical coherence tomography of lower limb wound healing with histopathological correlation.

Direct noninvasive visualization of wound bed with depth information is important to understand the tissue repair. We correlate skin swept-source-optical coherence tomography (OCT) with histopathological and immunohistochemical evaluation on traumatic lower limb wounds under honey dressing to compare and assess the tissue repair features acquired noninvasively and invasively. Analysis of optical biopsy identifies an uppermost brighter band for stratum corneum with region specific thickness (p < 0.0001) and gray-level intensity (p < 0.0001) variation. Below the stratum corneum, variation in optical intensities is remarkable in different regions of the wound bed. Correlation between OCT and microscopic observations are explored especially in respect to progressive growth and maturation of the epithelial and subepithelial components. Characteristic transition of uniform hypolucid band in OCT image for depigmented zone to wavy highly lucid band in the pigmented zone could be directly correlated with the microscopic findings. The transformation of prematured epithelium of depigmented area, with low expression of E-cadherin, to matured epithelium with higher E-cadherin expression in pigmented zone, implicated plausible change in their optical properties as depicted in OCT. This correlated evaluation of multimodal images demonstrates applicability of swept-source-OCT in wound research and importance of integrated approach in validation of new technology.

Discriminant analysis for fast multiclass data classification through regularized kernel function approximation.

In this brief we have proposed the multiclass data classification by computationally inexpensive discriminant analysis through vector-valued regularized kernel function approximation (VVRKFA). VVRKFA being an extension of fast regularized kernel function approximation (FRKFA), provides the vector-valued response at single step. The VVRKFA finds a linear operator and a bias vector by using a reduced kernel that maps a pattern from feature space into the low dimensional label space. The classification of patterns is carried out in this low dimensional label subspace. A test pattern is classified depending on its proximity to class centroids. The effectiveness of the proposed method is experimentally verified and compared with multiclass support vector machine (SVM) on several benchmark data sets as well as on gene microarray data for multi-category cancer classification. The results indicate the significant improvement in both training and testing time compared to that of multiclass SVM with comparable testing accuracy principally in large data sets. Experiments in this brief also serve as comparison of performance of VVRKFA with stratified random sampling and sub-sampling.