MathWorks and Nvidia Collaborate on Software-Defined Workflows in Med Tech

MathWorks has recently unveiled an exciting integration that brings Matlab functionality to the Nvidia Holoscan platform. This collaboration opens up new possibilities for medical device engineers, allowing them to leverage their existing Matlab algorithm codebase to create and deploy AI-enabled applications using Holoscan.

This integration significantly speeds up the development and deployment process for streaming data analysis and visualization applications. By encapsulating algorithms and functions into GPU-accelerated Holoscan operators, engineers can achieve real-time data processing and inference, enhancing the efficiency and performance of their applications.

Medical device engineers face the challenge of navigating complex global regulations while keeping up with rapid technological advancements. This dynamic environment often leads to devices becoming outdated shortly after launch, prompting the rise of Software as a Medical Device (SaMD).

Holoscan serves as a powerful sensor processing platform that simplifies the creation and implementation of AI and high-performance computing applications for real-time insights. It offers a comprehensive infrastructure for scalable, software-defined streaming data processing at the edge, enabling the integration of cutting-edge AI applications in clinical settings.

The integration of Matlab into Holoscan equips medical device engineers with a wide range of built-in matrix operations and toolbox functions for tasks such as image and signal processing, filtering, transforms, and deep learning algorithms. This seamless integration enhances the capabilities of engineers in developing innovative medical solutions.

The International Medical Device Regulators Forum defines SaMD as software designed for medical purposes that operate independently of hardware medical devices. To remain competitive post-deployment, SaMD products must have the flexibility to incorporate additional software features over time, necessitating the establishment of a software-defined workflow by engineers.

David Niewolny, Director of Business Development for MedTech at NVIDIA, emphasized the impact of artificial intelligence on the medical technology industry. He highlighted the collaboration between NVIDIA and MathWorks in accelerating AI-powered innovation within the med tech sector through the integration of MATLAB into the Holoscan platform.

Implementing a Holoscan pipeline with MATLAB involves several key steps, including creating a MATLAB function, generating accelerated CUDA code using GPU Coder, developing a Holoscan operator wrapper, and integrating the new MATLAB operator into the Holoscan application. This streamlined process enables the creation of a functional, software-defined workflow tailored for medical devices.

MathWorks' Director of Product Marketing, David Rich, emphasized the significant performance boost that engineers can experience by utilizing MATLAB functions within the NVIDIA Holoscan platform. The collaboration between MathWorks and industry leaders like NVIDIA is poised to drive innovation in medical device development, empowering customers to design, test, and validate their products while adhering to industry regulations and standards.

For more information on the integration of MATLAB into the Holoscan platform, visit Nvidia Holoscan GitHub.