Our team has worked on more than 28 client projects across the United States since we were founded in 2012.
Project durations range from a month up to several years.
In addition to selected client projects, we also present our own applications and a few of our open source software projects.
2019 | Birmingham, AL | DICOM standard
We developed and deployed a custom made DICOM de-identification server for the Radiology department at the University of Alabama at Birmingham to speed up data acquisition for research activities.
2017 | Boston, MA | MATLAB port, performance optimization
For CorticoMetrics we ported a complex numerical algorithm from MATLAB to Python. The company is a startup budding from a lab in an academic institution. Their codebase was written in MATLAB, but they wanted it to be ported to Python so it could be more scalable. We used our expertise in MATLAB and Python, understanding of complex algorithms and experience working with data-intensive medical imaging to fulfill this need.
2015 | Houston, TX | Software development, image processing, DICOM
We were hired by Mobius Medical Systems to support their in-house team. The team was building quality assurance (QA) software for radiation oncology treatment planning. The company did not have the bandwidth to complete their projects by the time expected by their customers. We provided in-house support to expedite Mobius's product launch, keeping their customers satisfied with their services.
2017 | Champaign, IL | Embedded development, UI/UX, regulatory documentation
We helped PhotoniCare successfully submit their first 510(k) for their TOMi™ Scope, an “advanced optical coherence tomography imager in miniature.” Due to our relevant medical imaging expertise, they hired us instead of in-house developers. As their software partner, we worked alongside their hardware engineers to refine their signal processing algorithms, design and develop their user interface, and write compliant regulatory documentation.
2016 | Norfolk, VA | Machine learning, image processing, web development, UI/UX
We collaborated with CIRS's engineering team to design and develop software to characterize an MRI scanner's geometric distortion and track it over time. CIRS needed support to develop a medical imaging tool in a timely manner. We brought expertise in deep learning algorithms, user interface design, and DICOM files, giving CIRS well-rounded support while building their product.
2015 | San Francisco, CA | Software development, native UI, image processing, DICOM
We collaborated with the in-house hardware and software teams at Magnetic Insight to design and implement a new MPI (Magnetic Particle Imaging) system. We helped Magnetic Insight produce the world's highest sensitivity and finest resolution MPI images by providing expertise in medical imaging.
2017 | Boston, MA | MRI Processing, PACS Server Configuration
CorticoMetrics developed a prototype of a software-based system for an MRI scanner that reduces error in tumor measurement. They needed help integrating the software into an MRI machine and hospital PACS to get it ready for the market. Our engineers were able to get their software working with the hospital IT and bring their product to market.
2016 | New York, NY | Web development, 3D viewer, image processing
We extended the existing visualization, data analysis, and 3D image reconstruction features of the InVivo Analytics web application. InVivo Analytics is a small academic startup that was interested in our software development services because of our expertise in medical imaging. They were looking to build a new team for developing their software. We took over their existing product and delivered working software to their company in a shorter timeline than they expected with their previous general-purpose software team.
2015 | Baton Rouge, LA | Web development, EMR integration
The Mary Bird Perkins Cancer Center needed a custom patient workflow dashboard. We were hired because of our experience working with electronic medical record systems. Our client received a deployed product designed and implemented by our team.
2014 | Houston, TX | Web development, clinical workflows, EMR Integration, HIPAA
We helped Matrix MD design web interfaces, build a HIPAA-compliant backend, and fine-tune a decision making algorithm for their Clinical Decision-Making Tool. MatrixMD lacked the bandwidth to bring their clinical decision-making tool to market. We were hired to supplement their engineering team and accelerate the development of the product.
MAUDE-Alerts makes market research for medical devices easy and reliable. MAUDE-Alerts sends notifications for new Medical Device Reports and provides a search feature with an easy to use form and comprehensive results. Read more to learn about the full capacity of the site.
We created a free site that allows visitors to view DICOM files in a browser. Patients may receive a CD-ROM disc from their physician containing the files from imaging studies. These images usually are DICOM files which require a specialized viewer to read. The Online DICOM Image Viewer allows patients to view DICOM images in the browser without a specialized viewer.
The DICOM standard is challenging to navigate, and the relationships between entities are difficult to follow. In an effort to make the standard more navigable for software developers and users, we developed the DICOM standard browser, which is freely available online.
React Slicer Toolkit (RST) is a framework to render medical images in the browser. RST can render sets of 2D and 3D image data from a variety of file formats, including DICOM. The framework has helped our developers expediently build custom medical image viewers for many different applications.
We developed a testing methodology and an associated Python framework that automates algorithm testing. It is the High Dimensionality Algorithm Tester (HDAT). Without an HDAT, testing image processing algorithms was difficult and time consuming. Our HDAT speeds up our development process and ensures less errors in test.
We built 3D Natural Neighbor Interpolation software to create 3D grids without a GPU (graphics processing unit). Existing versions of natural neighbor interpolation in Python only created 2D grids. Our algorithm is used by our company to create 3D grids out of scattered data.
We need to access the pixel values of 3D DICOM files frequently during client work. We have a DICOM-Numpy Library that reconstructs the 3D scan data from DICOM files. We shared this knowledge by open-sourcing the solution that has worked well for us.