OptimMRI (v2)

OptimMRI (v2) is a software application designed to enhance neurosurgical functional planning by processing pre-operative MR images to create 3D annotated brain models. It employs advanced image processing and machine learning algorithms to accurately localize critical structures like the STN and VIM, aiding clinicians in pre-operative decision-making.

OptimMRI (v2) is a web-based software application designed to assist qualified medical professionals in processing, visualizing, and interpreting anatomical structures from medical images, specifically for neurosurgical functional planning. It utilizes pre-operative DICOM compatible MR images of the brain to generate 3D annotated models. The software employs advanced image processing techniques and machine learning models, trained on a proprietary clinical database, to identify and mark specific brain regions. Its workflow supports the creation of pre-operative plans by allowing users to display, review, analyze, annotate, interpret, and export these 3D models, with outputs compatible with neurosurgical planning software supporting the 3D DICOM format. The core science of OptimMRI (v2) lies in its ability to aid in the precise localization of critical brain structures, such as the Subthalamic Nuclei (STN) and the Ventral Intermediate Nucleus (VIM), including a newly added model for the inferior-lateral region of the VIM. The software implements three distinct machine learning models for segmenting these regions of interest, converting them into 3D cross markers. Non-clinical performance testing demonstrated the accuracy of these annotations, with validation studies showing that at least 90% of surface distances for the newly targeted inferior-lateral VIM structure were within 2.0mm when compared to established reference devices, confirming the software's capability to provide accurate anatomical localization for neurosurgical planning.

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 machine learning algorithms are employed for the segmentation of STN and VIM, and how was the proprietary clinical database curated and validated for training these models?; How does the semi-automatic segmentation process in OptimMRI (v2) compare to fully manual or fully automatic segmentation methods in terms of user workflow efficiency and accuracy for STN/VIM localization?; What are the quantifiable benefits or improvements in neurosurgical functional planning outcomes when using OptimMRI (v2) compared to traditional planning methods or its predicate device without the new VIM model?; Beyond STN and VIM, are there plans to expand the software's capability to segment other critical brain structures relevant to different neurosurgical procedures?; How does the web-based architecture of OptimMRI (v2) address data security, patient privacy, and integration with existing hospital IT infrastructure and DICOM PACS systems?