UNiD Spine Analyzer

The UNiD™ Spine Analyzer v4.0 is an advanced SaMD platform designed to enhance orthopedic spine surgery planning by providing tools for image measurement, surgical simulation, and implant selection. It leverages machine learning models trained on retrospective patient data to predict post-operative radiographic parameters, aiming to provide surgeons with deeper insights for optimizing surgical strategies.

The UNiD™ Spine Analyzer is a web-based Software as a Medical Device (SaMD) developed by MEDICREA International, Inc., designed to assist healthcare professionals in viewing, measuring radiological images, and planning orthopedic, particularly spine, surgeries. The system enables surgeons and service providers to perform both generic and spine-specific measurements, simulate surgical strategies, measure anatomical components for implant placement, and either design patient-specific rods or select from a pre-selection of standard implants, including a database of Medtronic implants. Version 4.0 introduces significant updates, including a new Degenerative Predictive Model and updates to existing Adult and Pediatric Predictive Models, all trained using retrospective longitudinal patient datasets. These models allow for the display of predicted quantitative radiographic parameters derived from static machine-learning algorithms when users view simulated surgeries, aiding in preoperative planning. Clinical judgment and experience remain crucial for its proper utilization. The scientific foundation of the UNiD Spine Analyzer v4.0 relies on advanced image processing for anatomical measurements and surgical simulation, enhanced by machine learning. The newly integrated and updated predictive models (Degenerative, Adult, and Pediatric) utilize retrospective patient data to forecast radiographic outcomes of planned spinal surgeries, thereby offering a tool to account for predicted spinal compensation. These predictions are strictly radiographic and do not incorporate patient-reported outcomes. The software includes an osteotomy module, various measuring tools (e.g., angle, LLordo, SVA) and surgical tools (e.g., wedge, open, resect), alongside features for custom implant design (e.g., UNiD Rod, cages, screws). Software development, verification, and validation, including hazard analysis per ISO 14971 and IEC 62304, and a usability study, demonstrated the device's intended performance, ease of use, and appropriate human factors, supporting its safe operation.

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 retrospective longitudinal patient datasets were utilized to train the Degenerative, Adult, and Pediatric predictive models, and what were the key demographic and clinical characteristics of these patient cohorts?; How does the system's static machine-learning model account for individual patient variability beyond the parameters included in the training datasets when predicting quantitative radiographic outcomes from planned surgeries?; What is the quantified accuracy or margin of error for the predicted radiographic values generated by the new predictive models (Degenerative, Adult, Pediatric) when compared to actual post-operative outcomes in validation studies?; Beyond displaying normative data and predicted values, are there any integrated functionalities within the software to alert surgeons to potential significant discrepancies or high-risk scenarios identified by the predictive models based on the planned surgical approach?; How does the UNiD Spine Analyzer v4.0 integrate with existing hospital Picture Archiving and Communication Systems (PACS) for image input and how does it manage patient data privacy and security, especially being a web-based SaMD?