Bonelogic

Bonelogic 2.2 is a software system designed to support orthopedic healthcare professionals in diagnosing and planning surgical interventions for foot and ankle conditions. It utilizes CT image data to perform semi-automatic segmentation of bone structures, create 3D anatomical models, and provide tools for radiographic measurements, optionally employing an artificial neural network for assisted bone identification to enhance clinical workflows.

Bonelogic 2.2 is a software tool intended for orthopedic healthcare professionals to aid in the diagnosis of conditions and the planning of surgical procedures for the foot and ankle within a hospital or clinic setting. The system processes computed tomography (CT) imaging input to perform semi-automatic segmentation of bony structures, allowing for either manual input or an assisted input method for identifying these structures. It then generates three-dimensional mathematical models of the patient's foot and ankle anatomy. Furthermore, Bonelogic 2.2 provides measurement templates containing standard radiographic measures relevant to the foot and ankle, alongside tools that enable users to manually obtain precise linear and angular measurements. These 3D models, combined with the derived measurements, are utilized for diagnosing orthopedic conditions, while the dedicated surgical planning application employs these models and measurements to help plan treatments and operations to correct various foot and ankle pathologies. The device's technical capabilities include segmenting foot and ankle bones from CT scans to create patient-specific 3D anatomical models. The semi-automatic segmentation workflow can be guided by a healthcare professional's manual marking or by an optional assisted input feature that utilizes a locked artificial neural network (ANN) model for initial bone identification, which the user then acknowledges. Performance evaluations indicated high accuracy, with software verification meeting criteria of 95% model conformance within a 1.0mm distance to a reference model and a 2.0-degree standard deviation for angular measurements. A clinical data-based assessment of the AI algorithms, using 82 CT image series, demonstrated 100% accuracy in identifying foot and ankle bones and 98.8% accuracy in identifying the presence of metal. The AI models were developed and validated using a diverse dataset of 357 CT studies, encompassing various patient demographics, foot alignments, clinical conditions (including fractures and metal implants), and imaging systems, with ground truth established by three orthopedic surgeons and independent datasets for training, tuning, and validation.

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 performance of the semi-automatic segmentation with ANN-assisted input compare to the purely manual input workflow in terms of processing time and accuracy for complex foot and ankle deformities?; What specific quantitative benefits (e.g., reduced surgical time, improved implant placement accuracy) have been observed when using Bonelogic 2.2 for surgical planning compared to traditional methods?; Are there any known limitations in the software's ability to accurately segment bones or identify metal in the presence of severe artifacts in CT images or with specific implant materials?; How does the system account for variations in patient anatomy due to age (e.g., pediatric patients, though current IFU mentions adults over 16) or rare congenital conditions of the foot and ankle?; What is the typical learning curve for an orthopedic surgeon to achieve proficiency in utilizing all functionalities of Bonelogic 2.2, including the 3D modeling and surgical planning modules?