Precision AI Surgical Planning System (PAI-SPS)

Performance Metrics

Clinical Evidence

The evidence base is one paper. Wigmore and colleagues (JSES International, August 2025) enrolled 46 adults undergoing anatomic or reverse total shoulder arthroplasty planned in the ASTRA software, which the paper names as previously Precision AI. One patient was excluded after the surgeon switched from a stemless to a stemmed humeral component during surgery, leaving 45: mean age 69, 24 men and 21 women, 31 with osteoarthritis and 14 with cuff-tear arthropathy, 20 with severe posterior glenoid wear (Walch B2, B3 or C). Eight fellowship-trained surgeons operated through a delto-pectoral approach. Ethics approval was granted prospectively by Monash University (study 37215). The endpoint was technical: the difference between the planned and the achieved component orientation, measured on a post-operative CT scan. Mean absolute deviation was 1.7 degrees for glenoid version (SD 1.5, range 0.1 to 5.7), 1.4 degrees for glenoid inclination (SD 1.3, range 0.0 to 7.0), 2.2 degrees for humeral neck-shaft angle (SD 1.4, range 0.1 to 5.4) and 1.3 mm for humeral resection thickness (SD 1.2, range 0.0 to 5.1). Humeral version could be measured in only 22 patients whose CT scans reached the epicondyles: mean absolute deviation 2.9 degrees, with a signed bias of 2.5 degrees towards anteversion that differed from zero (p = 0.001). Resection was 0.7 mm thinner than planned on average (p = 0.035). Using thresholds of 5 degrees for neck-shaft angle and glenoid angles, 10 degrees for version and 3 mm for resection, 8 patients (17.8 per cent) were outliers on at least one measure: 5 on the humeral side and 3 on the glenoid. The paper's summary figure is that 89 per cent of humeral and 93 per cent of glenoid components were placed within those limits. The comparisons the paper draws are against published figures for standard instruments (Rojas et al for the humerus, Hendel et al for the glenoid), not against a control arm in the same patients. Subgroups were exploratory and unpowered: only glenoid version differed, with larger deviation in men (p = 0.007). Variance did not differ between surgeons for the four measures tested. The 24 anatomic cases used the Mathys Affinis Short prosthesis, which is not on the FDA-cleared implant list; the 21 reverse cases used the Enovis AltiVate Reverse, which is. The authors describe their paper as the first in vivo accuracy study of a humeral patient-specific guide; that is their claim. The authors disclosed no funding or grants, but the acknowledgements thank Enovis ANZ for supporting the research and the limitations state that the software manufacturer provided support. Two authors are Enovis ANZ employees, one receives institutional funding from an Enovis ANZ subsidiary, three surgeons are Enovis ANZ consultants (one of them the senior author, who co-founded Precision AI and sat on its guide design team), and one surgeon is a paid consultant to Mathys, maker of the anatomic implant used. The original 510(k) summary (K233992, 2024) also refers to a post-market clinical case series of 35 patients in Australia, evaluated by post-operative CT under ethics approval. It reports no figures, and the entry does not treat it as the same cohort as the 2025 paper. The two later clearances rested on bench testing, software verification and usability validation; clinical testing was not required. Evidence grade: a prospectively approved single-arm accuracy cohort with a post-operative CT acquired for the study is graded prospective. It is not a randomised comparison and it measures placement, not patient outcome. AurisMark reasoning: three FDA clearances and one peer-reviewed study of this device earn one mark. A second mark is withheld because the only study is industry-authored with no independent replication, no functional or revision outcome, no published deployment scale, and no approval outside the United States could be located. Re-evaluate on an independent accuracy or outcome study or a non-US register entry.

Inside the Auris+ Listing

Regulatory Approvals

Safety Record

No recall, enforcement action or MAUDE adverse-event report naming Precision AI, whether as device name, manufacturer, product description or recalling firm, was found in the openFDA device databases queried directly on 15 and 26 September 2026, including a recall search by the three K-numbers. The ASTRA trade name was not searched in MAUDE because it collides with unrelated products. The study did not report complications or adverse events; its only endpoint was placement accuracy. The clinical risk sits in the outliers: about one patient in six exceeded a placement threshold on at least one measure, and humeral version showed a consistent bias towards anteversion. The glenoid guide does not control reaming depth, and the study did not record intraoperative adjustments by the surgeon. The guides are cleared for a delto-pectoral approach only.

Intended Use & Indications

Per the FDA indications for use (K251558): the Precision AI Planning Software is intended to be used as a pre-surgical planner for simulation of surgical interventions for shoulder joint arthroplasty. It assists in the positioning of shoulder components by creating a 3D bone construct of the joint and allows the surgeon to visualize, measure, reconstruct, annotate and edit pre-surgical plan data, leading to a surgery report and a plan data file used as input to design the Precision AI Shoulder Guide and Biomodels. The Guides and Biomodels are patient-specific surgical instruments that assist intraoperative positioning of shoulder implant components in total and reverse shoulder arthroplasty by referencing anatomic landmarks identifiable on preoperative CT. The Glenoid Guide places the k-wire; the Humeral Guide places the humeral pins for humeral head resection. Single use only. Indicated for adult patients consented for shoulder joint arthroplasty; both guides suit a delto-pectoral approach only. Indicated for use with the listed Enovis (AltiVate Anatomic, AltiVate Anatomic Augmented, Turon, Reverse Shoulder Prosthesis, AltiVate Reverse, AltiVate Reverse Glenoid) and Lima (SMR and PRIMA) implant systems and their compatible components.