manDAAD
Manual Skills: Digitised Assessment, Analysis and Documentation
Project duration: 01/04/2018 - 30/09/2020
Project leader: Prof. Dr. Elke Kraus (ASH Berlin), Prof. Dr. Peter Hufnagl (HTW Berlin)
Project staff: Cornelia Schübl (ASH Berlin), Sebastian Lohmann (HTW Berlin)
Partners:
- Tembit GmbH
- Stabilo International GmbH
- Verband Deutscher Ergotherapeuten
- AOK Nordost - GeWINO - Gesundheitswissenschaftliches lnstitut Nordost
Abstract:
Project goals:
1) At the micro level:
Completion of a digitised diagnostic instrument together with the Digi-Pen from Stabilo, as an app / software. Digitisation spans implementation, evaluation, documentation and interpretation as a best practice example for users working with children or adults with or without motor problems (especially occupational therapists in paediatrics).
2) At the meso level:
Modularisation of the individual motor tests of the handedness profile, specifically tracing, dotting, hammering, tapping and drawing simultaneous bimanual circles. These modules can be implemented with other conditions (e.g. in psychiatry, in people with dementia or Alzheimer's, after a stroke or traumatic brain injury) as a best practice example of diagnostic e-tests. For example, a baseline performance can be established, and progress can be evaluated regularly from home in the form of e-tests.
3) At the macro level:
The multiple and diverse data of both qualitative and quantitative nature are summarised and analysed using analytic and interpretation tools. These additional systematic processes extend beyond a mere assessment to assist users in interpreting the results and making justified clinical decisions. They can serve as a basis for a best practice example with a deep-learning approach, as an alternative structure for patient records.
Workflow:
Phase 1 - creation of the manDAAD function prototype (software) In this phase, the content of the existing dig-TEMA software will be modified from a tablet application to the application of the Digipen – the so-called manDAAD software. At the same time, the stand-alone application is to be used to create an app with a connection to patient records. Phase 2 - application and evaluation of the manDAAD software In Phase 2, the software will be applied in practice. For this purpose, test subjects are trained on the use of the manDAAD software/functional prototype I with Stabilopen, in order to evaluate the usability in practice during implementation - a practicality study. In parallel, the software will be further developed as an app with a focus on data transfer, integration into the electronic patient records and integration of deep learning procedures. This will be carried out by HTW using a prototype I for the use of deep learning, in order to present a second function prototype at the end of this stage. Phase 3 - adaptation and revision of manDAAD software and publications In this phase, the manDAAD function prototype II will be revised and finalised on the basis of the existing study results and presented to the partners. In addition, the results will be published. The focus here is on optimising the app and in particular user control (usability adaptation) as well as initial analyses from the deep learning approach, using the analytic process to establish a handedness type as an example. In addition, the software will be made available to other potential users (e.g. BMBF project AID, to develop an app for the treatment of depression). The manDAAD software can be used by different professional groups such as therapists, paediatricians, neurologists and psychologists in the diagnostic field. It could also be a suitable tool for school entrance examinations in 5-6-year-old children. In addition, it could potentially be adapted as an intervention tool and offer training programmes to improve motor skills. Overall, it could be a best practice example of how to digitise common assessment procedures, thereby increasing the reliability and informative value of the results and expanding their scope of application.
Funding: Berlin Institute for Applied Research - Institut für angewandte Forschung Berlin (IFAF Berlin)
Keywords: Digitised Assessment, Deep Learning, Best Practice Example