Brainlab Elements (7.0); Brainlab Elements Image Fusion (5.0); Brainlab Elements Image Fusion Angio (1.0); Brainlab Elements Contouring (5.0); Brainlab Elements Fibertracking (3.0); Brainlab Elements BOLD MRI Mapping (1.0)
K243633Brainlab AG · cleared 2025-06-13 · product code QIH · Radiology
Premarket evidence — what FDA accepted
source quote (p.6)
“The Brainlab Elements are applications and background services for processing of medical images including functionalities such as data transfer, image co-registration, image segmentation, contouring and other image processing.”
source quote (p.13)
“Cranial tumors are auto-segmented as 3D objects in image sets with supported modality (MR-t1 contrast enhanced) by means of a machine learning algorithm. Anomaly detection is either the result of the atlas-based automatic segmentation of the brain or, on platforms with appropriate GPUs, the result of a machine learning algorithm. Probabilistic tracking algorithm Constrained Spherical Deconvolution tracking”
Validation studies (3)
Retrospective clinical
sample size not stated
endpoints: This test was to validate that Elements Virtual iMRI Cranial can be applied to cranial MR and intraoperative MR/CT/US image data related to image guided surgery approaches to compensate for surgery-related brain shift during resection.
Bench
sample size not stated
endpoints: To validate if Fibertracking allows to visualize cranial white matter structures such as motoric, language and visual tracts based on state of the art approaches for Fibertracking as Constrained Spherical Deconvolution (CSD) and probabilistic tracking.
Retrospective clinical
n=412 patients
endpoints: To validate Elements AI Tumor Segmentation can be used to semi-automatically segment supported cranial tumors (metastases, meningiomas, cranial and paraspinal nerve tumors, gliomas, glioneuronal and neuronal tumors) on 3D medical Contrast Enhanced T1-weighted MR images powered by the Anatomical Patient Model. The approach used was quantitative validation - comparison to ground-truth annotations.
Reported performance (3 observations)
source quote (p.22)
“Mean Recall All 0.85. Acceptance criteria were Dice ≥ 0.7, Recall ≥ 0.8 and Precision ≥ 0.8 for the lower bound of the respective 95 % confidence intervals.”
source quote (p.22)
“Mean Dice All 0.75. Acceptance criteria were Dice ≥ 0.7, Recall ≥ 0.8 and Precision ≥ 0.8 for the lower bound of the respective 95 % confidence intervals.”
source quote (p.22)
“Mean Precision All 0.86. Acceptance criteria were Dice ≥ 0.7, Recall ≥ 0.8 and Precision ≥ 0.8 for the lower bound of the respective 95 % confidence intervals.”
Each value carries its own analysis unit and task — never compare or pool across devices. Source: 510(k) summary PDF.
Predicate network
Postmarket — what happened after clearance
Recall and MAUDE counts are product-code-level (reports aren't reliably attributable to one device). Signals are descriptive observables with sources — never a judgment that the device is unsafe or drifting. Snapshot 2026-07-08.
Reimbursement — how devices like this got paid
Not yet tracked — no payment pathway indexed for this clearance (the reimbursement corpus is a growing seed set).