104 - A novel, cost-effective, anatomically accurate knee phantom for practicing ultrasound-guided injections

Friday, February 20, 2026

5:00 PM - 6:30 PM AST

Presenting Author(s)

BO

Bjorn Olmanson, MD

PM&R Resident
University of Minnesota
Eagan, Minnesota, United States

Co-Author(s)

JB

Jessica Bailey, DO

Resident
University of Minnesota
Minneapolis, Minnesota, United States

Corresponding Author(s)

MM

Marlena Rose Mueller, DO

University of Minnesota Physical Medicine and Rehabilitation Faculty
University of Minnesota
Minneapolis, Minnesota, United States

Submitter(s)

BO

Bjorn Olmanson, MD

University of Minnesota
Eagan, Minnesota, United States

Objectives: Ultrasound-guided knee injection practice is essential for trainees. While residents traditionally practice ultrasound on each other, it is impractical to perform injections on peers. Commercially available phantoms exist, but are often cost-prohibitive. 3D-printed components may offer a low-cost alternative, though previous knee models lack anatomical detail and the ability to practice injections. We present a novel, affordable, and anatomically accurate knee phantom using gelatin, 3D-printed components, and household materials to simulate the tactile and imaging characteristics of a real knee to perform ultrasound-guided injections.

Design: The model was constructed using a combined femur/patella bone unit, an anterior knee mold, and a quadriceps tendon mold developed from an open-source CT-based knee model. These components were 3D-printed in PLA. Silicone caulk was applied to both the knee mold, to form a thin skin layer, and to the bone unit, to form prefemoral and quadriceps fat pads. Silicone was mixed with psyllium powder in a tendon mold and attached to the patella. A rubber sphere filled with water was placed between the fat pads and connected to a posteriorly exiting tube to simulate the suprapatellar recess. The femur was pinned inside the knee mold, which was then filled with a 10% ballistic gelatin and psyllium powder mixture and refrigerated.

Results: The completed phantom model was used for multiple successful ultrasound-guided knee injections, confirmed by injectate exiting the posterior tube. Ultrasound visualization of the skin, soft tissue, bone, quadriceps tendon, fat pads, and suprapatellar recess allowed for accurate needle guidance. The psyllium powder provided realistic echogenicity, and the ballistic gelatin simulated the resistance of soft tissue. Material cost per phantom was approximately $15, with the ability to reuse most components for future models.

Conclusions: The authors propose an affordable, anatomically accurate, and reproducible knee phantom for practicing ultrasound-guided knee injections.