CurveBeam announces release of AI platform for complex foot and ankle measurements

CurveBeam announces release of AI platform for complex foot and ankle measurements

Press releases may be edited for formatting or style | February 24, 2020 Artificial Intelligence
HATFIELD, PENN. -- The advent of weight-bearing CT technology has led to new insights into functional foot and ankle alignment. CurveBeam, the leader in weight-bearing CT technology, is collaborating with orthopedic researchers to design artificial intelligence tools to take advantage of our expanding understanding of lower extremity biomechanics.

Almost all analysis requires a segmented CT scan, a dataset in which the bones are individually identified. Manual segmentation is time-consuming to the point that it is practical only for research applications in most circumstances.

Automatic segmentation promises to make the process much more efficient, and therefore clinically applicable. However, most current tools use a clustering method, which is semi-automatic at best. Human intervention is usually required to “clean up” mis-identified bones and joints, especially in anatomy that is not “normal.”

Servicing GE/Siemens Nuclear Medicine equipment with OEM trained engineers

Numed, a well established company in business since 1975 provides a wide range of service options including time & material service, PM only contracts, full service contracts, labor only contracts & system relocation. Call 800 96 Numed for more info.

The foot, with 26 interlocking bones, is notoriously one of the most difficult parts of the body to segment, both manually and automatically. In a quest to make this process more practical for everyday use, CurveBeam has partnered with an artificial intelligence group to use a deep machine learning model to accurately segment and identify complex foot bone structures with little or no human intervention.

The tool was trained using 300 datasets of healthy feet & ankles. The tool was taught to account for typical variations in anatomy, such as accessory ossicles.

CurveBeam will debut its segmentation platform at AAOS. The platform, called CubeVue Autometrics is a secure WebApp to which clinicians will be able to upload CurveBeam device datasets. The fully automated segmentation process will take about two minutes. Clinicians will be able to view the segmented datasets in CubeVue Autometrics as 3D renderings and via multi-planar reconstruction (MPR) presentations. Users will also be able to apply a suite of measurement tools to the dataset, including CurveBeam’s Smart M1-M2 angle measurement tool.

Smart M1-M2 computes the intermetatarsal angle (IMA), a common angle used in presurgical planning. IMA is a radiographic angle measurement between the long axes of the first and second metatarsals. Identification and measurement of the IMA represents a near universal tenet of determining hallux valgus (HV) deformity severity. IMA is traditionally measured on two-dimensional plain radiographs.

CubeVue Autometrics licenses for research applications are anticipated to be available in April 2020. This tool is anticipated to be available for clinical diagnosis by December 2020.

Smart M1-M2 is also available as a feature in CubeVue, CurveBeam’s custom visualization software. The Smart M1-M2 feature is currently a research tool and is not intended for diagnostic evaluation.

Providers will have the option to generate a PDF report of a patient’s Smart M1-M2 measurements via CubeVue Autometics or export an STL file of the segmented dataset.

Additional automatic measurements on the segmentation platform will include:
- TALAS™ (Torque Ankle Lever Arm System), a novel method to calculate hindfoot alignment in three-dimensions.
- Syndesmetric, a novel method to calculate syndesmosis injury based on automatic volumetric measurement of air space between the tibia and fibula.

To learn more and see the HiRise on display, visit CurveBeam at AAOS Booth #2909 in Hall B. Visit www.curvbeam.com, call 267-483-8081, or email info@curvebeam.com to set up a demo.

You Must Be Logged In To Post A Comment