arXiv subjects

Saikat Basu

Publications and source records attributed to Saikat Basu.

31 records · Page 2Linked to original sources

arXiv2018

Comparative Study of Simulated Nebulized and Spray Particle Deposition in Chronic Rhinosinusitis Patients

Introduction: Topical intranasal drugs are widely prescribed for Chronic Rhinosinusitis (CRS), although delivery can vary with device type and droplet size. The study objective was to compare nebulized and sprayed droplet deposition in the paranasal sinuses and ostiomeatal complex (OMC) across multiple droplet sizes in CRS patients using computational fluid dynamics (CFD). Methods: Three-dimensional models of sinonasal cavities were constructed from computed tomography (CT) scans of three subjects with CRS refractory to medical therapy using imaging software. Assuming steady-state inspiratory airflow at resting rate, CFD was used to simulate 1-120 μm sprayed droplet deposition in the left and right sinuses and OMC with spray nozzle positioning as in current nasal spray use instructions. Zero-velocity nebulization simulations were performed for 1-30 μm droplet sizes, maximal sinus and OMC deposition fractions (MSDF) were obtained, and sizes that achieved at least 50% of MSDF were identified. Nebulized MSDF was compared to sprayed droplet deposition. We also validated CFD framework through in vitro experiments. Results: Among nebulized droplet sizes, 11-14 μm droplets achieved at least 50% of MSDF in all six sinonasal cavities. Five of six sinonasal cavities had greater sinus and OMC deposition with nebulized droplets than with sprayed droplets at optimal sizes. Conclusions: Nebulized droplets may target the sinuses and OMC more effectively than sprayed particles at sizes achieving best deposition. Further studies are needed to confirm our preliminary findings. Several commercial nasal nebulizers have average particle sizes outside the optimal nebulized droplet size range found here, suggesting potential for product enhancement.

physics.med-ph↗

arXiv2018

Comparative Analysis of the Main Nasal Cavity and the Paranasal Sinuses in Chronic Rhinosinusitis: An Anatomic Study of Maximal Medical Therapy

Objective: Minimal literature exists investigating changes in inflammation with respect to the main nasal cavity (MNC) and paranasal sinuses (PS) before and after maximal medical therapy (MMT) for chronic rhinosinusitis (CRS). We hypothesized that MMT produces a differential level of change in the volume of air space in the MNC and PS, and that resolution of mucosal disease associated with the osteomeatal complex (OMC) influences clinical response to MMT. Study Design: Retrospective study of 12 pre- and post-MMT sinus-CT scans from 6 subjects with CRS, of which three succeeded and three failed therapy. Methods: Mimics was used to create 3D-models of the MNC and PS, and then analysis of the models was performed. Results: Mean differences in the sinonasal volume were 7866.5+/-4339.9 mm3 and 17869.10+/-19472.70 mm3, amongst the failures and successes, respectively. There is wide variability in the contribution of PS and MNC airspace volume change to the overall change in the sinonasal volume. In two subjects, the direction of volume change in the MNC and PS diverged with respect to the overall change in volume. Line-of-sight analysis demonstrated that successful responders to MMT had more patent MNC with direct access to the OMC. Conclusions: There is a differential contribution to sinonasal, airspace volume change after MMT, when comparing the MNC and PS. Response to MMT may not be solely attributable to PS change and may include a function of MNC change. Line-of-sight models suggest that direct access to the OMC may impact response to MMT.

physics.med-ph↗

arXiv2018

Quantum error-correcting code for ternary logic

Ternary quantum systems are being studied because these provide more computational state space per unit of information, known as qutrit. A qutrit has three basis states, thus a qubit may be considered as a special case of a qutrit where the coefficient of one of the basis states is zero. Hence both $(2 \times 2)$-dimensional as well as $(3 \times 3)$-dimensional Pauli errors can occur on qutrits. In this paper, we (i) explore the possible $(2 \times 2)$-dimensional as well as $(3 \times 3)$-dimensional Pauli errors in qutrits and show that any pairwise bit swap error can be expressed as a linear combination of shift errors and phase errors, (ii) propose a new type of error called quantum superposition error and show its equivalence to arbitrary rotation, (iii) formulate a nine qutrit code which can correct a single error in a qutrit, and (iv) provide its stabilizer and circuit realization.

quant-ph↗

arXiv2018

Pixel-level Reconstruction and Classification for Noisy Handwritten Bangla Characters

Classification techniques for images of handwritten characters are susceptible to noise. Quadtrees can be an efficient representation for learning from sparse features. In this paper, we improve the effectiveness of probabilistic quadtrees by using a pixel level classifier to extract the character pixels and remove noise from handwritten character images. The pixel level denoiser (a deep belief network) uses the map responses obtained from a pretrained CNN as features for reconstructing the characters eliminating noise. We experimentally demonstrate the effectiveness of our approach by reconstructing and classifying a noisy version of handwritten Bangla Numeral and Basic Character datasets.

cs.CV↗

arXiv2018

DeepGlobe 2018: A Challenge to Parse the Earth through Satellite Images

We present the DeepGlobe 2018 Satellite Image Understanding Challenge, which includes three public competitions for segmentation, detection, and classification tasks on satellite images. Similar to other challenges in computer vision domain such as DAVIS and COCO, DeepGlobe proposes three datasets and corresponding evaluation methodologies, coherently bundled in three competitions with a dedicated workshop co-located with CVPR 2018. We observed that satellite imagery is a rich and structured source of information, yet it is less investigated than everyday images by computer vision researchers. However, bridging modern computer vision with remote sensing data analysis could have critical impact to the way we understand our environment and lead to major breakthroughs in global urban planning or climate change research. Keeping such bridging objective in mind, DeepGlobe aims to bring together researchers from different domains to raise awareness of remote sensing in the computer vision community and vice-versa. We aim to improve and evaluate state-of-the-art satellite image understanding approaches, which can hopefully serve as reference benchmarks for future research in the same topic. In this paper, we analyze characteristics of each dataset, define the evaluation criteria of the competitions, and provide baselines for each task.

cs.CV↗

arXiv2017

On computational fluid dynamics models for sinonasal drug transport: relevance of nozzle subtraction and nasal vestibular dilation

Generating anatomically realistic three-dimensional (3D) models of the human sinonasal cavity for numerical investigations of sprayed drug transport presents a host of methodological ambiguities. For example, subject-specific radiographic images used for 3D reconstructions typically exclude spray bottles. Subtracting a bottle contour from the 3D airspace and dilating the anterior nasal vestibule for nozzle placement augment the complexity of model-building. So, we explored the question: how essential are these steps to adequately simulate nasal airflow and identify the optimal delivery conditions for intranasal sprays? In particular, we focused on particle deposition patterns in the maxillary sinus, a critical target site for chronic rhinosinusitis (CRS). The models were reconstructed from post-surgery computed tomography scans for a 39-year-old Caucasian male, with CRS history. Inspiratory airflow patterns during resting breathing are reliably tracked through CFD-based steady state laminar-viscous modeling and such regimes portray relative lack of sensitivity to inlet perturbations. Consequently, we hypothesized that the posterior airflow transport and the particle deposition trends should not be radically affected by the nozzle subtraction and vestibular dilation. The study involved 1 base model and 2 derived models; the latter two with nozzle contours (two different orientations) subtracted from the dilated anterior segment of the left vestibule. We analyzed spray transport in the left maxillary sinus for multiple release conditions. Similar release points, localized on an approximately 2mm-by-4.5mm contour, facilitated improved maxillary deposition in all three test cases. This suggests functional redundancy of nozzle insertion in a 3D numerical model for identifying the optimal spray release locations.

physics.flu-dyn↗

arXiv2017

On angled bounce-off impact of a drop impinging on a flowing soap film

Small drops impinging angularly on thin flowing soap films frequently demonstrate the rare emergence of bulk elastic effects working in-tandem with the more common-place hydrodynamic interactions. Three collision regimes are observable: (a) drop piercing through the film, (b) it coalescing with the flow, and (c) it bouncing off the film surface. During impact, the drop deforms along with a bulk elastic deformation of the film. For impacts that are close-to-tangential, the bounce-off regime predominates. We outline a reduced order analytical framework assuming a deformable drop and a deformable three-dimensional film, and the idealization invokes a phase-based parametric study. Angular inclination of the film and the ratio of post and pre impact drop sizes entail the phase parameters. We also perform experiments with vertically descending droplets impacting against an inclined soap film, flowing under constant pressure head. Model predicted phase domain for bounce-off compares well to our experimental findings. Additionally, the experiments exhibit momentum transfer to the film in the form of shed vortex dipole, along with propagation of free surface waves. On consulting prior published work, we note that for locomotion of water-walking insects using an impulsive action, the momentum distribution to the shed vortices and waves are both significant, taking up respectively 2/3-rd and 1/3-rd of the imparted streamwise momentum. In view of the potentially similar impulse actions, this theory is applied to the bounce-off examples in our experiments, and the resultant shed vortex dipole momenta are compared to the momenta computed from particle imaging velocimetry data. The magnitudes reveal identical order ($10^{-7}$ N$\cdot$s), suggesting that the bounce-off regime can be tapped as a simple analogue for interfacial bio-locomotion relying on impulse reactions.

physics.flu-dyn↗

arXiv2016

Error tracing in linear and concatenated quantum circuits

Descriptions of quantum algorithms, communication etc. protocols assume the existence of closed quantum system. However, real life quantum systems are open and are highly sensitive to errors. Hence error correction is of utmost importance if quantum computation is to be carried out in reality. Ideally, an error correction block should be placed after every gate operation in a quantum circuit. This increases the overhead and reduced the speedup of the quantum circuit. Moreover, the error correction blocks themselves may induce errors as the gates used for error correction may be noisy. In this paper, we have proposed a procedure to trace error probability due to noisy gates and decoherence in quantum circuit and place an error correcting block only when the error probability exceeds a certain threshold. This procedure shows a drastic reduction in the required number of error correcting blocks. Furthermore, we have considered concatenated codes with tile structure layout lattice architecture[25][21],[24] and SWAP gate based qubit transport mechanism. Tracing errors in higher levels of concatenation shows that, in most cases, after 1 or 2 levels of concatenation, the number of QECC blocks required become static. However, since the gate count increases with increasing concatenation, the percentage saving in gate count is considerably high.

quant-ph↗

arXiv2016

Core Sampling Framework for Pixel Classification

The intermediate map responses of a Convolutional Neural Network (CNN) contain information about an image that can be used to extract contextual knowledge about it. In this paper, we present a core sampling framework that is able to use these activation maps from several layers as features to another neural network using transfer learning to provide an understanding of an input image. Our framework creates a representation that combines features from the test data and the contextual knowledge gained from the responses of a pretrained network, processes it and feeds it to a separate Deep Belief Network. We use this representation to extract more information from an image at the pixel level, hence gaining understanding of the whole image. We experimentally demonstrate the usefulness of our framework using a pretrained VGG-16 model to perform segmentation on the BAERI dataset of Synthetic Aperture Radar(SAR) imagery and the CAMVID dataset.

cs.CV↗

arXiv2016

A Theoretical Analysis of Deep Neural Networks for Texture Classification

We investigate the use of Deep Neural Networks for the classification of image datasets where texture features are important for generating class-conditional discriminative representations. To this end, we first derive the size of the feature space for some standard textural features extracted from the input dataset and then use the theory of Vapnik-Chervonenkis dimension to show that hand-crafted feature extraction creates low-dimensional representations which help in reducing the overall excess error rate. As a corollary to this analysis, we derive for the first time upper bounds on the VC dimension of Convolutional Neural Network as well as Dropout and Dropconnect networks and the relation between excess error rate of Dropout and Dropconnect networks. The concept of intrinsic dimension is used to validate the intuition that texture-based datasets are inherently higher dimensional as compared to handwritten digits or other object recognition datasets and hence more difficult to be shattered by neural networks. We then derive the mean distance from the centroid to the nearest and farthest sampling points in an n-dimensional manifold and show that the Relative Contrast of the sample data vanishes as dimensionality of the underlying vector space tends to infinity.

cs.CV↗

arXiv2016

OptCon: An Adaptable SLA-Aware Consistency Tuning Framework for Quorum-based Stores

Users of distributed datastores that employ quorum-based replication are burdened with the choice of a suitable client-centric consistency setting for each storage operation. The above matching choice is difficult to reason about as it requires deliberating about the tradeoff between the latency and staleness, i.e., how stale (old) the result is. The latency and staleness for a given operation depend on the client-centric consistency setting applied, as well as dynamic parameters such as the current workload and network condition.We present OptCon, a novel machine learning-based predictive framework, that can automate the choice of client-centric consistency setting under user-specified latency and staleness thresholds given in the service level agreement (SLA). Under a given SLA, OptCon predicts a client-centric consistency setting that is matching, i.e., it is weak enough to satisfy the latency threshold, while being strong enough to satisfy the staleness threshold. While manually tuned consistency settings remain fixed unless explicitly reconfigured, OptCon tunes consistency settings on a per-operation basis with respect to changing workload and network state. Using decision tree learning, OptCon yields 0.14 cross validation error in predicting matching consistency settings under latency and staleness thresholds given in the SLA. We demonstrate experimentally that OptCon is at least as effective as any manually chosen consistency settings in adapting to the SLA thresholds for different use cases. We also demonstrate that OptCon adapts to variations in workload, whereas a given manually chosen fixed consistency setting satisfies the SLA only for a characteristic workload.

cs.DC↗

arXiv2015

Learning Sparse Feature Representations using Probabilistic Quadtrees and Deep Belief Nets

Learning sparse feature representations is a useful instrument for solving an unsupervised learning problem. In this paper, we present three labeled handwritten digit datasets, collectively called n-MNIST. Then, we propose a novel framework for the classification of handwritten digits that learns sparse representations using probabilistic quadtrees and Deep Belief Nets. On the MNIST and n-MNIST datasets, our framework shows promising results and significantly outperforms traditional Deep Belief Networks.

cs.CV↗

arXiv2015

DeepSat - A Learning framework for Satellite Imagery

Satellite image classification is a challenging problem that lies at the crossroads of remote sensing, computer vision, and machine learning. Due to the high variability inherent in satellite data, most of the current object classification approaches are not suitable for handling satellite datasets. The progress of satellite image analytics has also been inhibited by the lack of a single labeled high-resolution dataset with multiple class labels. The contributions of this paper are twofold - (1) first, we present two new satellite datasets called SAT-4 and SAT-6, and (2) then, we propose a classification framework that extracts features from an input image, normalizes them and feeds the normalized feature vectors to a Deep Belief Network for classification. On the SAT-4 dataset, our best network produces a classification accuracy of 97.95% and outperforms three state-of-the-art object recognition algorithms, namely - Deep Belief Networks, Convolutional Neural Networks and Stacked Denoising Autoencoders by ~11%. On SAT-6, it produces a classification accuracy of 93.9% and outperforms the other algorithms by ~15%. Comparative studies with a Random Forest classifier show the advantage of an unsupervised learning approach over traditional supervised learning techniques. A statistical analysis based on Distribution Separability Criterion and Intrinsic Dimensionality Estimation substantiates the effectiveness of our approach in learning better representations for satellite imagery.

cs.CV↗