Oncospace

Oncospace is a software-only device that assists radiation oncologists and dosimetrists in configuring and reviewing radiotherapy treatment plans for patients with malignant or benign diseases in the head and neck, thoracic, abdominal, and pelvic regions. It uses machine learning algorithms to predict organ-at-risk dose objectives, automate plan optimization initiation, and provide a user interface for plan evaluation, while maintaining human oversight and not directly controlling treatment machines.

The subject device expands anatomical coverage to include thoracic, abdominal, and gynecological regions (new to this submission), while head and neck and prostate models were updated for improved performance. All other technological characteristics remain identical to the predicate: Windows/web-browser operating system, client-server architecture, DICOM-RT compliance, no direct connection to radiation delivery devices, and identical user interface features for dose visualization and plan comparison.

Not stated in this summary.

Substantial equivalence is based on identical intended use and functionality: both devices configure and review radiotherapy plans for the same user population (medical professionals) and do not control radiation delivery. The expansion to new anatomical sites (thoracic, abdominal, gynecological) used identical machine learning methods as the predicate and demonstrated clinical non-inferiority through prospective testing showing no statistical difference in target coverage and equivalent or superior organ-at-risk sparing. The updated models for existing sites (head and neck, prostate) met performance acceptance criteria. Since the same modeling approach and feature types apply across all sites, the device design is not site-specific and raises no new safety or efficacy questions.