Skip to main content

Intuitive Natural Prosthesis UTilization

Periodic Reporting for period 3 - INPUT (Intuitive Natural Prosthesis UTilization)

Reporting period: 2018-08-01 to 2020-01-31

The INPUT project builds upon two previous FP7 research projects, AMYO and MYOSENS. In those projects, the academic state of the art in upper limb prosthesis control was significantly advanced. Most noticeably, simultaneous and proportional control of two degrees of freedom through multi-channel muscle recordings was achieved. As Innovation Action, INPUT will have less focus on inventing novel approaches, but rather take the output of its preceding projects and improve them to a state which makes them usable for a large population of amputees in everyday life. This will be achieved by concentrating on 3 key components: 1) Optimized muscle signal acquisition hardware 2) More robust machine learning methods and 3) Encouraging training methods for amputees to get better and higher focused rehabilitation training. All of this will be validated in clinical settings with amputees in real life scenarios. To this end, the INPUT consortium is composed of two industrial partners specialized in prosthetics and signal acquisition, two academic partners specialized in signal processing, and two clinical partners for outcome assessment. Half of this balanced team has already collaborated within AMYO and MYOSENS, while the other half are new partners. This careful composition of partners will allow INPUT to achieve its primary goal of making highly advanced prosthetic control ready for everyday clinical use. The results of INPUT will significantly increase the quality of life for amputees.
The consortium of INPUT is made up of two academic, two clinical and two industrial partners. This balanced combination of expertise represents also the operative combination of all work packages in the project.
In short, the developments of the academic partners are iteratively tested in tight collaboration with the clinics and the industrial partners develop and manufacture the necessary prosthetic hardware, while coordinating the overall goal of preparing market-readiness for all components.
Research in the project is focused on robust control methods and signal processing methods and EMG signal acquisition knowledge generation. Clinical aspects cover patient testing and clinical study evaluations of the developed control methods in real-life settings. Additionally, the development of optimal user training is part of the clinical work, in order to prepare users optimally for using their prosthesis. Industrial aspects cover the assembly of the prostheses for the patient tests and the overall project coordination towards viable market solutions. In the first 18 months of the project’s implementation progress was as planned. The results achieved so far in WPs 4 to 6 were very valuable for improving the user interface towards reliability and better control for the end-user. We advanced significantly in establishing materials and processing methods for the signal acquisition liners. Together with the new data processing methods based on robust regression for 3 degrees of freedom, the foundations for a more reliable detection of movement intent were laid. Despite initial concerns we were also able to adapt commercially available hardware for this project to be used for state of the art patient evaluations. Furthermore, we investigated and evaluated different patient training regimes which allow users to train in a controlled and motivating way.
In the second half of the project, all separately developed components (liner, control methods and training games) will be united to one system and then tested extensively in clinical studies.
"End-user impact:
» The benefit for the end-users achieved with the INPUT system is the most important impact this project will have.
In our project we will concentrate on the application of established technologies and improve them only where necessary for making simple control of a complex prosthesis daily reality for the wearers. ""Don and play"" will become reality for the wearer - easy to use, reliable and without annoying frequent malfunctions. Reducing pure research to a minimum will allow us to dedicate a maximum of time to interact with end-users. This will help us to quantitatively assess the effectiveness of the implemented measures, but even more importantly puts users in the design loop. Their satisfaction is the ultimate measure of success for INPUT. In the course of this project we will, for the first time, have a large population of amputees wear advanced prostheses for a prolonged time. The integration of the feedback by these first-hand experiences in our development will be the enabling factor for INPUT to create a device that actually satisfies user needs - and not just those of engineers and reviewers.

Industrial impact:
» The industrial partners of INPUT (two entities of Ottobock) have all facilities to exploit the project's results after its end as the market leader in exoprosthetics. A concrete exploitation plan will be generated as part of WP2 in Task 2.3.
The results of INPUT will thus have direct application in the market. The advancement of the practicality level of current solutions is important before attempting market introduction, since the previously proposed systems do not yet possess the required technology readiness levels. By developing an integrated, autonomous, wearable system, the industrial impact of INPUT will be to advance this technology to the robust control of 4 DOF simultaneously with a TRL 8 - ""System complete and qualified"" - preparing market introduction.
To reduce market introduction efforts after the end of INPUT to a minimum, since the initial phases of the project attention will be paid to the fulfillment of ISO13485 as quality management standard for design and manufacture of medical devices.

Scientific impact:
The most relevant innovations INPUT will contribute will be targeting market applications relating to novel signal acquisition hardware, machine learning methods including novel robustness promoting strategies, end-user training and end-user evaluation paradigms for upper limb prosthetic control. While industrial impact will be prioritized over scientific impact in INPUT, all of these topics will also be of very high interest for the scientific community. We therefore plan to showcase our advancements towards ""don and play"" multi-DOF prosthesis control solutions in a number of scientific congresses and journal papers.