New Studies on Total Joint Arthroplasty and Trauma Solutions
15 new studies were presented describing new research and advancements in total joint arthroplasty and trauma solutions at the 2025 Orthopedic Research Society (ORS) Annual Meeting in Phoenix, Ariz. The works explored the company’s latest innovations leveraging artificial intelligence (AI) and machine learning (ML) image segmentation and predictive modeling, CT image-based radiomic analysis, computer-assisted surgical navigation, clinical outcomes research and enhanced implant materials.
Analysis
One ML-based study analyzed pre-operative CT scan data of the deltoid muscle to assess the influence of image acquisition and processing parameters on the extraction of radiomic measurements.
One study:
- Used ML framework to identify the most stable radiomic features across scanning protocols
- Found that shape-based features were the most robust and least affected
- Second-order texture features were the most impacted by variations in image processing parameters
Another ML-based Study:
- Examined the relationship between these stable radiomic features and the amount of shoulder motion a patient has before shoulder arthroplasty
- Reported several deltoid radiomic features were predictive preoperative and post-operative motion
...Our new research not only provides a repeatable method to quantitatively assess the deltoid, but we also identified the specific radiomic features that are clinically meaningful and predictive of shoulder arthroplasty outcomes...
“These radiomic studies provide new insights that can be incorporated into pre-operative planning software with applications for improved clinical decision support tools,” said Chris Roche, Advita Senior Vice President of Extremities. “Shoulder surgeons do not currently assess the deltoid using any quantitative method when considering their patients for shoulder arthroplasty.”
Other studies confirmed Activit-E® polyethylene‘s superior oxidation resistance and mechanical durability for joint arthroplasty. Compared to conventional ultra-high-molecular-weight polyethylene (UHMWPE), Activit-E outperformed in aggressive knee wear testing and simulated in-vivo aging conditions. Its peroxide crosslinked, vitamin E-stabilized composition is designed to enhance implant longevity and patient outcomes.
Additional research covered the development of a novel 3D classification of the TKA patients based on unsupervised machine learning clustering models, which can be applied to the company’s Newton® Knee balancing technology.
