We quickly put a small team together and developed a web application that enables surgeons and users to upload DICOM image data to a cloud server for design of anatomical models. The web application included a user friendly front end developed in Javascript, that communicates with an AWS hosted API and back end database and data storage platform.

In conjunction with the web application, we also developed a 3D visualization and modeling application that connected to the cloud platform and enabled downloading the DICOM image data, and perform sophisticated segmentation, visualization, model generation, and 3D design operations. The 3D models designed using the software could then be exported as STL files for 3D printing. The software also enabled communication of the STL files to the web platform so they can be viewed and approved by the surgeons and other users requesting the model before the model is manufactured.

The comany's flagship product utilizes machine learning algorithms that constantly evaluate a patient's vital signs searching for any changes that could signal a clinical event that might warrant further investigation by physicians and caregivers. We helped design and develop the clinical trial plan to demonstrate the safety and effectiveness of the device, including identifying inclusion/exclusion criteria, safety and performance endpoints and acceptance criteria and sample size calculations, and were heavily involved in designing the statistical analysis plan and final clinical report.

We worked very closely with the organization's leadership team to draft, submit and successfully prosecute a 510k application for the company's flagship software product which included extensive clinical data and software documentation. We managed the entire submission process and engaged with the FDA on behalf of the company to ensure a successful outcome.

We worked closely with the company's internal development team and in conjunction with other consultants to develop the device control and data management software. Specifically, our software represented the primary user interface of the device and also controlled the hardware components that initiated the stimulus and captured the images of the response allowing the user to perform scans with the device. We integrated computer vision software that was developed externally, and also implemented functionality to communicate patient and scan data to a cloud platform for later review and patient workflow management.

We also provided documentation in accordance with FDA's design controls regulation. The software and documentation we provided was later used in a 510k submission to commercialzie the device.

Working closely with the Executive team, we owned the regulatory process on multiple 510k applications with the FDA, and engaged with the FDA reviewers in the Neurology division on behalf of the company via pre-submission meetings, 510k discussions and clarifications, and breakthrough device designation meetings, which led to the device being granted breakthrough device designation status by the FDA. We additionally assisted in drafting the product's technical file that was submitted to attain CE Mark.

As part of these efforts, we assisted the company in many additional capacities, including providing input on clinical trial protocols and statistical analysis plans, human factors testing, software and hardware V&V testing, biocompatibility testing, electrical safety and EMC testing, etc.

We quickly assembled a team of software experts and developed the core elements of the software platform that controlled the imaging hardware sub-systems and also presented the primary user interface of the device. We worked closely with various hardware sub-system vendors and quicky developed a state machine based software platform to control multiple, complex hardware sub-systems, enabling them to work in unison to acquire a scan. We developed data communication pipelines to rapidly transfer data to be visualized on the screen, and integrated algorithms running on a GPU (developed externally) to perform complex segmentation steps.

Our team also developed modules to integrate the client's device with other devices commonly used in a cardiac catheterization lab, including the real-time X-ray systems used for catheter tracking, to enable multi-modality fusion of data. We also enabled communication of image data to external devices such as PACS, in DICOM and other standard formats.

We quickly led the effort to evaluate the changes to software requirements, hazard assessments documentation, and performed a detailed cybersecurity assessment of their updated software, in order to comply with FDA guidance documents related to cybersecurity documentation. The cybersecurity assessment included a threat model assessment of all systems exposed to identified threat vectors, and a set of mitigation recommendations to improve the robustness of the system. We also provided updated software V&V test protocols to ensure that any new software mitigations were tested for robustness and effectiveness. We additionally performed a detailed assessment of all software changes relative to FDA's guidance to determine whether the modifictaions were significant enough to warrant a new marketing application with the FDA.