Acorn 3D Software (AC-SEG-4009); Acorn 3DP Model (AC-101-XX)

Authorization

K260322 added anatomy alignment, measurements and correction simulation; it did not introduce a new ML segmentation model or modify the inherited PCCP.

Submission lineageK260322 → K252103 → K234009 PCCP.

Intended use

Visualize and segment medical images, align anatomy, measure spinopelvic geometry, plan pedicle screw placement, simulate correction and produce digital or physical three-dimensional models.

Input

CT, synthetic CT or CTA DICOM imaging; User-defined reference geometry and surgical correction inputs

Output

Editable vertebral and sacral segmentations; Aligned anatomy and spinopelvic measurements; Thoracic and lumbar pedicle-screw trajectories; Simulated correction plans and exportable three-dimensional models

Use boundary

Outputs require expert clinical judgment and user review; the software does not autonomously select or execute a surgical correction.

AI disclosure and function

AI/ML explicit

The suite retains a 3D U-Net machine-learning model for automatic vertebral and sacral segmentation; the newly added Correction module is algorithmic planning and simulation rather than a newly trained AI model.

No ML algorithm improvement or PCCP change occurred in K260322; the applicable segmentation PCCP was authorized in K234009 and carried through prior submissions.

No foundation-model, LLM, or generative-AI function was identified in the reviewed FDA materials.

Validation

Software verification, end-user validation and bench testing of the new Correction module, including simulated plans against surgeon-defined angular targets and intra- and inter-observer variability.

Comparator

The K252103 predicate suite and surgeon-defined correction inputs with predefined angular acceptance criteria.

Ground truth

Surgeon-defined target angular metrics and repeat measurements used as nonclinical reference values.

Representation

Not applicable to the reported bench and end-user testing for the new module; no new patient cohort was described for K260322.

Reported performance

Target angular accuracy

within 3 degreesBench accuracy of simulated correction against surgeon-defined inputs, not achieved postoperative alignment.

Intra- and inter-observer variability

less than 3 degrees by mean absolute difference and 2×SDRepeatability of simulated planning observations, not clinical-outcome consistency.

Limitations

• The new submission primarily validates correction simulation through bench and end-user testing.
• No patient-level deformity correction, reoperation or neurological-outcome evidence was reported.
• The ML segmentation evidence and PCCP originate from earlier submissions rather than a new model change in K260322.

It does not prove better correction, fewer complications, autonomous planning or a newly authorized adaptive model in K260322.

PCCP / Change control

Inherited from K234009; the indexed submission itself shows a “No” database flag. Controlled future updates to machine-learning segmentation of bony anatomy under the K234009 protocol; released algorithms remain trained, tuned and locked.

• Train, tune and lock each modified algorithm before release
• Assess changes against defined technical and performance specifications
• Mitigate degradation in segmentation performance within unchanged indications

A PCCP authorizes bounded changes under predefined controls. It is not permission for unrestricted or continuously self-learning behavior.

Postmarket check

No exact-name openFDA device-enforcement result was found in the targeted query.

No exact-name openFDA result found as of 2026-09-08; this is not proof that no recall exists.

Primary FDA sources

1. FDA AI-enabled device list — Submission K260322Confirms inclusion, list display name, company, lead panel, product code and decision date only.
2. FDA submission database — 510(k) K260322Confirms pathway, review panel, decision, primary and subsequent product codes, and No PCCP database flag.
3. FDA decision summary — 510(k) summary pp. 1-8; performance p. 4; AI and PCCP lineage pp. 7-8Supports intended use, predicates, current module changes, bench results, ML segmentation description and inherited PCCP lineage.

Sources checked Sep 8, 2026 · Next scheduled review Oct 8, 2026. “No exact-name result” is a bounded search finding, not proof that no safety event or recall exists.