D2P

D2P is a software system designed for processing DICOM medical images into output files for surgical planning and the creation of 3D-printed physical anatomical replicas. It employs advanced visualization, automated segmentation incorporating deep learning, and integration with 3D printing and virtual reality to support diagnostic and treatment decisions across multiple medical specialties.

The D2P software is a stand-alone, modular software package designed for the advanced visualization of DICOM imaging data. Its primary function is to serve as a software interface and image segmentation system, facilitating the transfer of imaging information from a medical scanner to an output file suitable for pre-operative surgical planning. This output file can be used to create physical anatomical replicas, which are intended for adjunctive use alongside other diagnostic tools and expert clinical judgment. These replicas assist in diagnosis, patient management, and treatment selection for a variety of applications, including cardiovascular, craniofacial, gastrointestinal, genitourinary, neurological, and musculoskeletal cases. The software incorporates several key functionalities, such as a DICOM viewer, automated segmentation (which utilizes a deep learning neural network for prediction), editing tools, and pre-printing capabilities. It offers seamless integration with 3D Systems' printers and software packages, as well as Virtual Reality visualization for non-diagnostic purposes, including a measurement tool. Technologically, it enhances standard medical image viewing by incorporating features like the ability to remove images from a DICOM dataset and the use of deep learning for improved segmentation. Non-clinical performance data, derived from measurement accuracy studies, usability assessments, and decimation studies, confirmed the software meets its specifications. Validation also demonstrated that accurate anatomical models for various clinical applications can be printed using compatible 3D printers and materials.

AI Regulatory Concerns

Powered by

Jev reads the full FDA letter and returns calibrated concern probabilities per regulatory category. These are AI estimates, not regulatory advice.

• Q1. What specific deep learning algorithms are implemented for the automated segmentation feature, and what datasets were used for their training and validation across the various indicated anatomical regions?; How does the accuracy of the physical replicas produced from the D2P software's output compare to traditional imaging or other 3D modeling techniques when used for diagnostic purposes in cardiovascular or neurological applications?; What are the typical processing times for segmenting complex anatomical structures, and how does the 'decimation study' impact the fidelity versus processing speed for the output file?; Are there any limitations on the types of DICOM images (e.g., CT, MRI, specific sequences) that the D2P software can effectively process for segmentation and 3D model generation?; How does the integration with Virtual Reality for non-diagnostic use, including the measurement tool, enhance pre-operative planning or anatomical understanding compared to 2D/3D screen visualization alone?