Dynamic 3D images for strong shoulders

After a shoulder injury has been treated, patients are often left with a feeling of insecurity – many of them report that their shoulder “doesn't hold” or “slips out easily”. When diagnosing shoulder instabilities, doctors often have to rely on these subjective assessments. The reason: Conventional imaging methods do not capture the movement of the shoulder. In contrast, the newly developed 4D analysis by Empa researcher Ameet Aiyangar goes beyond static imaging: “We combine high-precision X-ray videos from two perspectives and use them to reconstruct a four-dimensional motion analysis – in other words, a 3D image while the shoulder is moving.”

A biplanar radiographic imaging system (DBRI) installed at sitem-insel, the Swiss Institute for Translational and Entrepreneurial Medicine in Bern, is used for the dynamic 3D images of the shoulder. It was developed in close collaboration with Empa and Inselspital Bern. First, the test subjects perform specific shoulder movements, which are recorded from two different perspectives using synchronized X-ray images. Additional CT scans are used to reconstruct detailed 3D models of the bones with anatomical landmarks. A tracking procedure based on this determines the exact position of the shoulder joint. Finally, optimization software calculates the movement sequences of the shoulder.

This enables the researchers to record not only the rough movement patterns of a joint, but also the smallest rolling and sliding movements that are crucial for stability – with an accuracy in the range of 0.1 to 0.5 millimeters. “This is a decisive step forward, because conventional motion sensors with infrared cameras and markers on the skin are inaccurate by 20 to 40 millimeters – far too imprecise to reliably detect instabilities,” says Aiyangar.

The next step will be a joint study with Inselspital in Bern, with which there is a long-standing research collaboration. Around 40 patients with untreated shoulder instability are being sought to be examined before and after targeted muscle strengthening. Individual musculoskeletal models will play a central role to analyze the interactions between muscles, joints and forces. “In the long term, we hope that our four-dimensional movement analysis will find its way into clinical practice. This is because the problem with treating joint instability lies mainly in the dynamics,” explains Aiyangar. Based on the biomechanical study data, the Empa researcher also wants to develop force-controlled kinematic models that take into account muscle forces and joint loads in addition to movement. In future, these should enable a patient-specific dynamic analysis of shoulder instability – and thus a comprehensive clinical diagnosis without radiation-based imaging.