Snehashis Roy | Johns Hopkins University (original) (raw)

Papers by Snehashis Roy

Medical Imaging 2015: Image Processing, 2015

Quantification of hemorrhages in head computed tomography (CT) images from patients with traumati... more Quantification of hemorrhages in head computed tomography (CT) images from patients with traumatic brain injury (TBI) has potential applications in monitoring disease progression and better understanding of the pathophysiology of TBI. Although manual segmentations can provide accurate measures of hemorrhages, the processing time and inter-rater variability make it infeasible for large studies. In this paper, we propose a fully automatic novel pipeline for segmenting intraparenchymal hemorrhages (IPH) from clinical head CT images. Unlike previous methods of model based segmentation or active contour techniques, we rely on relevant and matching examples from already segmented images by trained raters. The CT images are first skull-stripped. Then example patches from an "atlas" CT and its manual segmentation are used to learn a two-class sparse dictionary for hemorrhage and normal tissue. Next, for a given "subject" CT, a subject patch is modeled as a sparse convex combination of a few atlas patches from the dictionary. The same convex combination is applied to the atlas segmentation patches to generate a membership for the hemorrhages at each voxel. Hemorrhages are segmented from 25 subjects with various degrees of TBI. Results are compared with segmentations obtained from an expert rater. A median Dice coefficient of 0.85 between automated and manual segmentations is achieved. A linear fit between automated and manual volumes show a slope of 1.0047, indicating a negligible bias in volume estimation.

2013 IEEE 10th International Symposium on Biomedical Imaging, 2013

Magnetic resonance imaging (MRI) is widely used for analyzing human brain structure and function.... more Magnetic resonance imaging (MRI) is widely used for analyzing human brain structure and function. MRI is extremely versatile and can produce different tissue contrasts as required by the study design. For reasons such as patient comfort, cost, and improving technology, certain tissue contrasts for a cohort analysis may not have been acquired during the imaging session. This missing pulse sequence hampers consistent neuroanatomy research. One possible solution is to synthesize the missing sequence. This paper proposes a data-driven approach to image synthesis, which provides equal, if not superior synthesis compared to the state-of-the-art, in addition to being an order of magnitude faster. The synthesis transformation is done on image patches by a trained bagged ensemble of regression trees. Validation was done by synthesizing T 2-weighted contrasts from T 1-weighted scans, for phantoms and real data. We also synthesized 3 Tesla T 1-weighted magnetization prepared rapid gradient echo (MPRAGE) images from 1.5 Tesla MPRAGEs to demonstrate the generality of this approach.

SPIE Medical Imaging, 2014

Computed tomography (CT) is the standard imaging modality for patient dose calculation for radiat... more Computed tomography (CT) is the standard imaging modality for patient dose calculation for radiation therapy. Magnetic resonance (MR) imaging (MRI) is used along with CT to identify brain structures due to its superior soft tissue contrast. Registration of MR and CT is necessary for accurate delineation of the tumor and other structures, and is critical in radiotherapy planning. Mutual information (MI) or its variants are typically used as a similarity metric to register MRI to CT. However, unlike CT, MRI intensity does not have an accepted calibrated intensity scale. Therefore, MI-based MR-CT registration may vary from scan to scan as MI depends on the joint histogram of the images. In this paper, we propose a fully automatic framework for MR-CT registration by synthesizing a synthetic CT image from MRI using a co-registered pair of MR and CT images as an atlas. Patches of the subject MRI are matched to the atlas and the synthetic CT patches are estimated in a probabilistic framework. The synthetic CT is registered to the original CT using a deformable registration and the computed deformation is applied to the MRI. In contrast to most existing methods, we do not need any manual intervention such as picking landmarks or regions of interests. The proposed method was validated on ten brain cancer patient cases, showing 25% improvement in MI and correlation between MR and CT images after registration compared to state-of-the-art registration methods.

Multimodal Brain Image Analysis, 2013

Magnetic resonance imaging (MRI) is a noninvasive modality that has been widely used to image the... more Magnetic resonance imaging (MRI) is a noninvasive modality that has been widely used to image the structure of the human brain. Unlike reconstructed x-ray computed tomography images, MRI intensities do not possess a calibrated scale, and the images suffer from wide variability in intensity contrasts due to scanner calibration and pulse sequence variations. Most MR image processing tasks use intensities as the principal feature and therefore the results can vary widely according to the actual tissue intensity contrast. Since it is difficult to control the MR scanner acquisition protocols in multi-scanner cross-sectional studies, results achieved using image processing tools are often difficult to compare in such studies. Similar issues can happen in longitudinal studies, as scanners undergo upgrades or improvements in pulse sequences, leading to new imaging sequences. We propose a novel probabilistic model to transform image contrasts by matching patches of a subject image to a set of patches from a multi-contrast atlas. Although the transformed images are not for diagnostic purpose, the use of such contrast transforms is shown for two applications, (a) to improve segmentation consistency across scanners and pulse sequences, (b) to improve registration accuracy between multi-contrast image pairs by transforming the subject image to the contrast of the reference image and then registering the transformed subject image to the reference image. Contrary to previous intensity transformation methods, our technique does not need any information about landmarks, pulse sequence parameters or imaging equations. It is shown to provide more consistent segmentation across scanners compared to state-of-the-art methods.

Information Processing in Medical Imaging, 2011

Neuroinformatics, Jan 2013

… Imaging: From Nano …, 2008

Medical Image Analysis, Feb 2012

Biomedical Imaging: From …, 2009

Tissue classification algorithms developed for magnetic resonance images commonly assume a Gaussi... more Tissue classification algorithms developed for magnetic resonance images commonly assume a Gaussian model on the statistics of noise in the image. While this is approximately true for voxels having large intensities, it is less true as the underlying intensity becomes smaller. In this paper, the Gaussian model is replaced with a Rician model, which is a better approximation to the observed signal. A new classification algorithm based on a finite mixture model of Rician signals is presented wherein the expectation maximization algorithm is used to find the joint maximum likelihood estimates of the unknown mixture parameters. Improved accuracy of tissue classification is demonstrated on several sample data sets. It is also shown that classification repeatability for the same subject under different MR acquisitions is improved using the new method.

ISBI, 2011 , 2011

We present a compressed sensing based approach to remove gain field from magnetic resonance (MR) ... more We present a compressed sensing based approach to remove gain field from magnetic resonance (MR) images of the human brain. During image acquisition, the inhomogeneity present in the radio-frequency (RF) coil appears as shading artifact in the intensity image. The inhomogeneity poses problem in any automatic algorithm that uses intensity as a feature. It has been shown that at low field strength, the shading can be assumed to be a smooth field that is composed of low frequency components. Thus most inhomogeneity correction algorithms assume some kind of explicit smoothness criteria on the field. This sometimes limits the performance of the algorithms if the actual inhomogeneity is not smooth, which is the case at higher field strength. We describe a model-free, nonparametric patch-based approach that uses compressed sensing for the correction. We show that these features enable our algorithm to perform comparably with a current state of the art method N3 on images acquired at low field, while outperforming N3 when the image has non-smooth inhomogeneity, such as 7T images.

VLSI Design, 2006. , 2006

Backend planning for SoCs needs to account for power pads and pins for different power domains. I... more Backend planning for SoCs needs to account for power pads and pins for different power domains. IO power pad requirements for high speed interfaces, are directly dependent on the worst case switching of output buffers. This work proposes an algorithm that takes switching activity patterns of a set of output buffers for an interface and generates an optimized IO power and ground pad locations. Optimisation is achieved by splitting the spatial locations of the drivers into smaller groups and solving pad requirement problem for each of the groups. Ground bounce is the main component based on which the pad count is estimated. Special requirements like multiple power domains, different packages (TQFP, BGA) etc, have also been addressed. Its been shown by simulations that up to 20% reduction in pad count can be achieved if switching patterns are available.

Journal of The Acoustical Society of America, 2011

SPIE Medical Imaging, 2011

It gives me immense pleasure and satisfaction to express my heartfelt gratitude to Prof. Somnath ... more It gives me immense pleasure and satisfaction to express my heartfelt gratitude to Prof. Somnath Sengupta for accepting me as his project student and providing me with excellent guidance throughout my project duration. His engaging discussions and introspective suggestions were the constant motivating factors that helped me to proceed in my work. Had it not been for him, I doubt whether the work would have reached the shape it has reached today. I am particularly grateful for the time he devoted to me and my work in spite of his many pressing engagements. He offered viewpoints and insights which went far beyond the narrow work and helped me embark on new ideas. It is impossible to put in words the way in which my association with him has not only enriched my work but also my life.

… Imaging: From Nano …, 2010