Ortho AI

Ortho AI is image-processing software that assists orthopedic surgeons in preoperative and intraoperative planning for hip replacement, knee replacement, and lumbar spine fusion surgery. It uses artificial intelligence to automatically detect anatomical landmarks and measure lengths and angles on radiological images, which surgeons can then review and edit before surgical planning.

Ortho AI accepts multiple image formats (JPEG, PNG, BMP, DICOM) whereas the predicate (KOALA) accepted only DICOM-compliant images. Ortho AI is semi-automated, requiring surgeon review and editing of AI-generated landmarks, whereas KOALA was fully automatic. Both perform length and angle measurements on radiological images using AI-driven algorithms and run on servers. Ortho AI extends to hip, knee, and lumbar spine, while KOALA was limited to leg imaging.

Not stated in this summary.

Although Ortho AI differs from KOALA in anatomical scope (hip, knee, lumbar spine vs. leg only), measurement approach (semi-automatic vs. automatic), and supported image formats, these differences do not raise new safety or effectiveness questions. Both are AI-based radiological image-processing systems for measurement-assisted surgical planning by trained professionals. The performance testing demonstrates accuracy within clinically acceptable tolerances (measurements within ±2 degrees and ±1.96 mm of human measurement), with Dice coefficients exceeding the 0.85 acceptance threshold across all anatomical models and x-ray machine types, establishing that the device performs safely and effectively as intended despite these incremental technological variations.