IB Lab LAMA

IB Lab LAMA is a software device employing deep learning technology to automatically analyze full leg AP radiographs for aiding healthcare professionals. Its purpose is to provide precise, automated geometric measurements of limb-length discrepancy and knee alignment parameters in adult patients, thereby assisting in the quantitative assessment of lower extremity deformities.

IB Lab LAMA is a fully-automated radiological image processing software device that utilizes deep learning technology to analyze uni- and bilateral AP full leg X-ray images. It is intended to aid healthcare professionals trained in radiology in the precise measurement of limb-length discrepancy and quantitative knee alignment parameters for adult patients aged 22 years and older, particularly those with suspected or present lower extremity deformities. The software automatically calculates thirteen geometric parameters, including mechanical axis deviation, femur/tibia/leg lengths, leg length discrepancy, hip knee ankle angle, and various other joint angles. The results are presented in DICOM reports for review on standard medical viewers, and the system compares measurements to pre-set norm-ranges. It is not indicated for use on radiographs where Ankle Arthroplasties or Unicompartmental Knee Arthroplasties are present and should not replace a full patient evaluation. A clinical data-based standalone performance study was conducted in the U.S. using 189 radiographs (325 legs) from adult patients to validate IB Lab LAMA's outputs. Ground truth measurements were established by two board-certified musculoskeletal radiologists. The study demonstrated strong agreement between the software's measurements and the ground truth for all parameters, assessed via Bland-Altman plots and Intra-class Correlation Coefficients; for example, mean differences were -1.37mm for mechanical axis deviation and 0.13mm for leg length discrepancy. The software also exhibited high sensitivity (95.05%) and specificity (99.80%) for detecting total knee and hip arthroplasties, a low failure rate of approximately 2.8%, and perfect repeatability in tests, indicating it provides clinically relevant and reliable measurements.

AI Regulatory Concerns

Powered by

Jev reads the full FDA letter and returns calibrated concern probabilities per regulatory category. These are AI estimates, not regulatory advice.

• Q1. How does the deep learning algorithm's performance in IB Lab LAMA vary across different demographic groups (e.g., body mass index, severity of deformity) within the indicated adult population?; What specific image quality parameters (e.g., resolution, contrast, patient positioning variability) are most critical for the optimal performance of IB Lab LAMA, and how does the system manage or flag suboptimal images?; Beyond the explicitly stated contraindications of Ankle Arthroplasties and Unicompartmental Knee Arthroplasties, are there other types of orthopedic implants, surgical alterations, or severe pathological conditions that could significantly impact the accuracy of the measurements?; What was the detailed inter-observer variability between the two musculoskeletal radiologists who established the ground truth, and how does this compare to the device-versus-ground truth variability for each of the 13 reported measurements?; Could the IB Lab LAMA software be adapted or further trained for use in pediatric populations (under 22 years) or for analyzing different radiographic views or imaging modalities relevant to lower limb assessment?