Our research partners

The Moovment software suite by Qinematic is intended to be used by health and wellness service providers in workplaces, gyms, clinics and retail outlets. It is an automated service that records, measures and reports carefully selected movements that are essential to optimal performance and independent living.

Moovment markerless scanning is affordable and accessible for everyone because we use easy to find off-the-shelf hardware, such as a gaming PC and the Microsoft Kinect sensor. The sensor records a 3D video and creates a life-like avatar of a person performing essential movements. Moovment software performs sofisticated analytics in the backgound to provide amazing visualisation of static and dynamic posture and important metrics for monitoring progress and exercise prescription.

Read the conclusions from lab testing of Moovment Scan (previously called Posture Scan) here.

Deep learning analyses 3D-video recorded at 30 frames per second. The Microsoft Kinect sensor has been validated by international research as a tool for posture and balance assessment in peer-reviewed scientific literature (1, 2, 3), even with the native skeletal tracking (4), which was intended for gaming.  It has demonstrated reliability in measurement, and inter-reliability in assessment and re-assessment.

The Qinematic team of human movement experts and 3D software engineers have gone one step further to fine-tune the tracking algorithms for improved accuracy, and a service design that makes implementation simple and easy to operate - no technical expertise required to calibrate and start the system, and no personnel required to run a scan. The scan is self-service and takes just 4 minutes.

The Kinect sensor does have some limitations and does not accurately scan all movement. For example, the frame rate (30fps) is not high enough for recording a golf swing, jumping or running. Furthermore, there are some movements that cannot be scanned with just one sensor because the body parts are not always visible. Many people ask for gait analysis, however the range of the camera is not sufficient for recording normal walking speed. In these cases, high-end laboratory equipment should be used, such as Qualysis and Vicon. However, not many people have the budget, time and knowledge to operate these labs.

Qinematic is the first line of assessment for health providers to triage clients and decide on resource allocation, and monitor progress over time - objectively. Some clients may need some home-exercises, which can be sent through the Moovment Pro app. Because Moovment is so fast and convenient, it is used by researchers to record and analyse simple functional movements that are the building blocks of more complex movements. Large samples of objective data about human function from Moovment Scan as well as self-reported clinical surveys from Moovment Pro enable measurement of rates of improvement/decline, intervention outcomes, and take the world of human movement closer to defining ‘ideal’, ‘normal variation’ and ‘dysfunctional’ movement patterns. 

Balance, posture and movement patterns in tasks such as side bending and squat are essential for activities of daily living, performing work and sports performance. These movements are assessed daily in gyms and clinics around the world, and the quality and magnitude of the movements can be associates with existing or potential problems - for decision support, prevention, and for rehabilitation. Qinematic hopes to bridge the gap between practitioners and researchers by collecting large amounts of movement patterns generated from ‘practice-based’ evidence and using applied science to turn big data insights into health innovation. We still need the small ‘evidence-based’ studies that typically involve 40-500 people, but measurement outside of the lab is more likely to reflect ‘real life’ and the amount of data necessary for powerful machine learing insights.

Licensed health professionals in western countries are increasingly obliged to show intervention outcomes to justify reimbursement, and to show they are working ‘evidence based’. We want to help them with that. There is still a role for ‘intuitive health’ however, it needs the support of objective, and preferably ‘digital health’ measures. Studies (5) show that both experienced and novice health professionals have some difficulty in observing and documenting function consistently. Observational assessment involving multiple body parts (eg. movement in the trunk, hips and the knees at the same time) is particularly difficult for humans. To define a problem (diagnosis) and choose a course of action (further assessment or intervention), the extent to which a movement is optimal (or dysfunctional) should be determined. With Moovment 3D avatars available in the cloud, this can be done face-to-face or remotely, and documented immediately with quantitative information - no more paperwork taking up valuable time and allowing human error to creep in!

Moovment does not score performance, diagnose problems or offer advice about interventions. That is for the health professionals to decide, with due consideration for the many other important contextual factors. Moovment simply provides accurate and consistent measures of intersegmental kinematics for the whole body. The assessment and the reporting is unbiased. It involves the client by communicating to them movement visually, reinforcing engagement and motivation through the Moovment Pro app. The 3D-video can be played back multiple time to assist with team meetings, researcher analysis and client education. Moovment Lab and Moovment Pro finally make it possible for the health seeker and the health provider to see movement from different perspectives. After all, movement is three dimensional. 

For repeat scanning to enable monitoring of progress over time, it needs to be reliable. The Moovment Scan instructions are standardised. The automated calibration procedure and movement error detection minimise bad data. Portability and ease of use make it possible to scan in natural environments such as homes, gyms and workplaces. 

Please note – the measures are accurate and repeatable - we have seen that in the lab. However, people do vary the way they move, which is the very reason why we need to measure them often and using a standardised methodology - to quantify the variation in the way they move, and not the variation in the measurement methodology. This is a key challenge for other observational measures such as posture grids, goniometers and popular Functional Movement Screening. Stereotypical movement makes researching movement easier, statistically speaking, but it is not considered healthy, and probably more indicitive of an underlying problem. There is mounting evidence of the importance of variability in normal movement, which reveals variation not as error, but as a necessary condition for function (6). Our own experience shows that there may be exceptions such as well-rehearsed dancers, or people who have had intensive repetitive knee rehabilitation - in that case, a deviation from their ‘normal’ may in fact be a sign of injury or risk of injury.

What is considered normal healthy variation (behavioral variation and not statistical variation) will emerge as we analyse intersegmental kinematics using big data analysis, instead of small data statistics. Qinematic will re-define ideal, normal and dysfunctional movement patterns, and thereafter consider scoring performance with the help of researchers. In the meantime, although the literature supports the Kinect sensor as a valid measurement tool, it is still up to the health provider to decide what to do with the measures, and what they mean for their clients. Moovment software simply makes measurement and exercise prescription easy, objective and entertaining.

 Link to conclusions from studies about Kinect.

Glenn Bilby

Human Movement Scientist
Physiotherapist
Course Manager – Transforming Healthcare, Karolinska Institute, Sweden
Founding CEO – Qinematic AB