Kitware has extensive experience creating and maintaining open-source software tools in an environment of large, distributed development teams. The popularity of this software can be demonstrated by a simple statistic: Kitware's public sewer http://public.kitware.com routinely receives between two and three million web hits per month. Kitware also develops proprietary tools with open architectures enabling the delivery of both closed and open technology modules. For example, the VolView (VolView) volume visualization system supports run-time plug-ins for image processing and segmentation. Most of the current plug-ins are open-source; however, Kitware is actively developing commercial plug-ins for targeted applications. This experience has led us to understand key features of the development process and how to leverage open software environments into supportive commercial products and services. Some of the lessons learned include the value of open development environments, particularly those that are cross-platform (i.e., hardware, operating system, and compiler independent). Crossplatforms tools are inherently more stable than a single-platform tool because the use of software in different environments routinely reveals problems that any single platform might not. For example, while C++ code might compile and run fine on one system, a different compiler might not compile the same code, or the executable might fail on startup due to different initialization processes. Another important lesson is insuring the availability of integration tools. Large software efforts such as that envisioned by NAMIC inherently involve extended communities of experts and affiliated organizations. These researchers typically bring their own software and methodologies that are often designed without a-priori knowledge of other systems. To a large extent the success of tools like VTK and ITK has been their fundamental design philosophy: that they are players in a larger suite of software. This has led to integration features such as support for multiple language bindings, extensive IO capabilities, event/callback mechanisms, and object factories for adaptive, run-time instantiation of objects. Finally, the effective creation of technology requires a mechanism to deliver it to the end customers. While we view the primary customer as researchers addressing the driving biological projects, it should be noted that most of the cores in NAMIC will require effective technology delivery platforms. For example, researchers may want to focus on a key algorithm, such as a segmentation module, and they do not want to have to build the infrastructure to read, process, and view data in order to test an algorithm. Instead, plugging the algorithm into a framework that provides these facilities accelerates the creation of technology and its delivery to other team members and the community at large. Technology delivery platforms also benefit those training and disseminating information because the platform provides a natural packaging mechanism with which to teach and communicate about a particular technology module.
| Lyu, Ilwoo; Perdomo, Jonathan; Yapuncich, Gabriel S et al. (2018) Group-wise Shape Correspondence of Variable and Complex Objects. Proc SPIE Int Soc Opt Eng 10574: |
| Hong, Sungmin; Fishbaugh, James; Gerig, Guido (2018) 4D CONTINUOUS MEDIAL REPRESENTATION BY GEODESIC SHAPE REGRESSION. Proc IEEE Int Symp Biomed Imaging 2018:1014-1017 |
| Swanson, Meghan R; Wolff, Jason J; Shen, Mark D et al. (2018) Development of White Matter Circuitry in Infants With Fragile X Syndrome. JAMA Psychiatry 75:505-513 |
| Swanson, Meghan R; Shen, Mark D; Wolff, Jason J et al. (2018) Naturalistic Language Recordings Reveal ""Hypervocal"" Infants at High Familial Risk for Autism. Child Dev 89:e60-e73 |
| Ghayoor, Ali; Vaidya, Jatin G; Johnson, Hans J (2018) Robust automated constellation-based landmark detection in human brain imaging. Neuroimage 170:471-481 |
| Wachinger, Christian; Toews, Matthew; Langs, Georg et al. (2018) Keypoint Transfer for Fast Whole-Body Segmentation. IEEE Trans Med Imaging : |
| Swanson, Meghan R; Shen, Mark D; Wolff, Jason J et al. (2017) Subcortical Brain and Behavior Phenotypes Differentiate Infants With Autism Versus Language Delay. Biol Psychiatry Cogn Neurosci Neuroimaging 2:664-672 |
| Veni, Gopalkrishna; Elhabian, Shireen Y; Whitaker, Ross T (2017) ShapeCut: Bayesian surface estimation using shape-driven graph. Med Image Anal 40:11-29 |
| Irimia, Andrei; Goh, Sheng-Yang Matthew; Wade, Adam C et al. (2017) Traumatic Brain Injury Severity, Neuropathophysiology, and Clinical Outcome: Insights from Multimodal Neuroimaging. Front Neurol 8:530 |
| Rutherford, Helena J V; Maupin, Angela N; Landi, Nicole et al. (2017) Parental reflective functioning and the neural correlates of processing infant affective cues. Soc Neurosci 12:519-529 |
Showing the most recent 10 out of 668 publications
