Kitware, Inc. is a technology company headquartered in Clifton Park, New York. The company specializes in the research and development of open-source software in the fields of computer vision, medical imaging, visualization, 3D data publishing and technical software development. In addition to software development, the company offers other products and services such as books, technical support, consulting and customized training courses. Wikipedia.
Leotta M.J.,Kitware |
Mundy J.L.,Brown University
IEEE Transactions on Pattern Analysis and Machine Intelligence | Year: 2011
In automated surveillance, one is often interested in tracking road vehicles, measuring their shape in 3D world space, and determining vehicle classification. To address these tasks simultaneously, an effective approach is the constrained alignment of a prior model of 3D vehicle shape to images. Previous 3D vehicle models are either generic but overly simple or rigid and overly complex. Rigid models represent exactly one vehicle design, so a large collection is needed. A single generic model can deform to a wide variety of shapes, but those shapes have been far too primitive. This paper uses a generic 3D vehicle model that deforms to match a wide variety of passenger vehicles. It is adjustable in complexity between the two extremes. The model is aligned to images by predicting and matching image intensity edges. Novel algorithms are presented for fitting models to multiple still images and simultaneous tracking while estimating shape in video. Experiments compare the proposed model to simple generic models in accuracy and reliability of 3D shape recovery from images and tracking in video. Standard techniques for classification are also used to compare the models. The proposed model outperforms the existing simple models at each task. © 2011 IEEE. Source
Agency: Department of Defense | Branch: Air Force | Program: SBIR | Phase: Phase I | Award Amount: 149.93K | Year: 2015
ABSTRACT:Current full motion video (FMV) toolchains fall short of supporting semantically meaningful archive search for specific objects or object types. We propose the Hierarchical Dynamic Vision Exploitation system (HiDyVE), an end-to-end system for content-based object location and retrieval in operational FMV. HiDyVE combines convolutional neural networks (CNNs) specifically adapted to FMV's quality and data volume characteristics with state-of-the-art object proposal mechanisms for efficient processing. By exploiting CNN layering and sophisticated domain transfer techniques, we minimize the amount of domain-specific labeled data required for training, instead leveraging existing large labeled datasets (e.g. ImageNet) to train the lower layers of the CNN. High-level scene categorizations are also generated. We also store and index intermediate descriptors, allowing the system to dynamically adapt to query concepts not present during training. A sophisticated interactive query refinement (IQR) system further incorporates user feedback to refine the search space, enabling the analyst to more quickly converge on relevant results. Real-world issues of video quality and metadata burn-in are addressed by our proven FMV front-end, which automatically detects on-screen static elements. User interaction is facilitated by our open-source FMV GUI toolkit.BENEFIT:This project will advance the state-of-the-art in computer vision, particularly content-based image recognition and scene understanding in video, but also more broadly by transitioning state-of-the-art techniques for processing high-resolution static images to lower-resolution video data. These improvements would positively impact many application areas, including aerial and ground-based video intelligence, surveillance, and reconnaissance (ISR); autonomous navigation; and social media understanding. Contributions from this effort will drive significant interest in the computer vision research community to leverage convolutional neural networks and harness contextual information for dramatically improved object recognition and matching in full motion video (FMV). Further, there is significant commercial potential in addition to military and defense applications. FMV and other video data are growing at unprecedented rates, and companies are looking at unique ways to capitalize on commercial opportunities. Commercial industries such as automotive, semiconductor, consumer electronics, food & packaging, healthcare, and logistics are using vision tracking systems for applications including quality assurance & inspection, tracking, measurement, and identification. The proposed technology has the potential to significantly increase the performance of Kitwares existing technologies, such as video tracking and activity recognition, where scene understanding will help algorithms adapt well to new content.
Agency: Department of Health and Human Services | Branch: National Institutes of Health | Program: SBIR | Phase: Phase II | Award Amount: 635.67K | Year: 2015
DESCRIPTION provided by applicant Nearly million Americans suffer traumatic brain injury TBI annually which constitutes a significant US medical health concern Although neuroimaging plays an important role in pathology localization and surgical planning TBI clinical care does not currently take full advantage of neuroimaging computational technology We propose to develop validate and commercialize computational algorithms based on our methods for image segmentation and registration These methods can accommodate the presence of large pathologies in TBI cases can yield quantitative measures from chronic and acute TBI data for research into characterizing injury monitoring pathology evolution informing patient prognosis and can aid clinicians in optimizing TBI patient care workflows We will accomplish our goal during the proposed Phase II effort by building upon our Phase I successes Featured in conference and journal publications during Phase I we devised a novel andquot low rank sparseandquot method for registering brain MRI scans from TBI patients with large pathologies to healthy brain atlases enabling more accurate identification and quantification of anatomic changes In conjunction with our foundational andquot geometric metamorphosisandquot work into quantifying lesion infiltration and recession over time our set of methods now address the major hurdles associated with TBI patient understanding Under this Phase II STTR proposal we will specifically focus on extending our computational methods for multimodal neuroimaging of TBI data processing We will provide finite element models created over a range of clinical cases of mild to severe TBI determine refined measures of patient change from longitudinal registrations integrate those methods into local and cloud based environments that support academic and commercial use and validate the complete commercial system using extensive TBI data collections including neuropsychological motor cognitive and behavioral outcome measures in a customer oriented study PUBLIC HEALTH RELEVANCE In the US approximately million individuals are victims of traumatic brain injury TBI annually with many requiring surgical intervention or long term care Initial assessment and treatment of TBI have appropriately become major US healthcare initiatives yet the effects of TBI can be particularly challenging for the patient and for healthcre systems Neuroimaging data analysis methods however are presently not properly employed to address this challenge Herein we propose to refine apply and test tools initiated under our Phase I STTR to perform the combined efficient analysis of multimodal neuroimage data sets for use in assessing the extent of brain injury its change over time and its effective treatment
Agency: Department of Health and Human Services | Branch: National Institutes of Health | Program: SBIR | Phase: Phase II | Award Amount: 1.26M | Year: 2015
DESCRIPTION provided by applicant There are several features of ultrasound imaging that make it attractive to clinicians and preclinical researchers including its relatively low cost rel time imaging capability safety and portability For example ultrasound imaging is widely used for anatomical imaging and blood flow measurements in the heart and large vessels Ultrasound however is typically not used in oncology because it has relatively poor quantitative capability with respect to tumor morphology or malignancy In our Phase I work we demonstrated that our new contrast enhanced dual frequency ultrasound technologies andquot Acoustic Angiographyandquot can capture detailed in vivo images of tissue vasculature in animal models of breast cancer and we have shown that vascular morphology is an indicator of tumor malignancy in those animal models That work built upon our prior vessel analysis research and algorithms that showed that quantifiable vascular morphology metrics from Magnetic Resonance Imaging data are reliable predictors of tumor malignancy and response to therapy in humans In the proposed Phase II work we will conduct the research necessary for the commercialization of our acoustic angiography system for preclinical research The team has been expanded to include SonoVol the manufacturer of ultrasound systems for preclinical research We will research and evaluate methods to ensure that our commercial system will be easy to use and consistently produce effective measures of tumor malignancy and response to therapy We will validate the product in a blinded study of breast cancer treatment efficacy in support of a preclinical trial i e the targeted commercial use for the proposed system PUBLIC HEALTH RELEVANCE Ultrasound is a relatively safe low cost portable real time imaging device however its images are relatively poor for detecting and diagnosing tumors We propose that ultrasound can be extended to tumor assessment via a commercial system that combines new micro bubble contrast agents that enhance the appearance of vessels within ultrasound images with an ultrasound imaging probe that we developed for capturing contrast enhanced ultrasound images and with novel vascular image analysis algorithms that we have also developed In Phase I we showed that our system is viable In the proposed Phase II work we will conduct the research and development needed to commercialize that system for assessing tumor malignancy and response to treatment in support of preclinical trials of experimental cancer therapies
Gessner R.C.,University of North Carolina at Chapel Hill |
Aylward S.R.,Kitware |
Dayton P.A.,University of North Carolina at Chapel Hill
Radiology | Year: 2012
Purpose: To determine if the morphologies of microvessels could be extracted from contrast material-enhanced acoustic angiographic ultrasonographic (US) images and used as a quantitative basis for distinguishing healthy from diseased tissue. Materials and Methods: All studies were institutional animal care and use committee approved. Three-dimensional contrast-enhanced acoustic angiographic images were acquired in both healthy (n = 7) and tumor-bearing (n = 10) rats. High-spatial-resolution and high signal-to-noise acquisition was enabled by using a prototype dual-frequency US transducer (transmit at 4 MHz, receive at 30 MHz). A segmentation algorithm was utilized to extract microvessel structure from image data, and the distance metric (DM) and the sum of angles metric (SOAM), designed to distinguish different types of tortuosity, were applied to image data. The vessel populations extracted from tumor-bearing tissue volumes were compared against vessels extracted from tissue volumes in the same anatomic location within healthy control animals by using the two-sided Student t test. Results: Metrics of microvascular tortuosity were significantly higher in the tumor population. The average DM of the tumor population (1.34 ± 0.40 [standard deviation]) was 23.76% higher than that of the control population (1.08 ± 0.08) (P < .0001), while the average SOAM (22.53 ± 7.82) was 50.73% higher than that of the control population (14.95 ± 4.83) (P < .0001). The DM and SOAM metrics for the control and tumor populations were significantly different when all vessels were pooled between the two animal populations. In addition, each animal in the tumor population had significantly different DM and SOAM metrics relative to the control population (P < .05 for all; P value ranges for DM, 3.89 × 10-7 to 5.63 × 10-3; and those for SOAM, 2.42 × 10-12 to 1.57 × 10-3). Conclusion: Vascular network quantification by using high-spatialresolution acoustic angiographic images is feasible. Data suggest that the angiogenic processes associated with tumor development in the models studied result in higher instances of vessel tortuosity near the tumor site. © RSNA, 2012. Source