The use of Serious Games (SG) in post-stroke rehabilitation has been considered a promising therapeutic resource. Key elements in rehabilitation (intensity, repetitiveness and task orientation), are highlighted by using SG. Moreover, SG improve both patients’ attention and motivation during rehabilitation sessions contributing to patients’ adherence. However, existing commercial games have been crafted with no concern on clinical efficacy. SG conception orientated on clinical efficacy is fundamental for patients’ rehabilitation. Therefore, the aim of this work is to present a conceptual model for SG design/development for post-stroke rehabilitation.

A systematic literature was performed to identify main limitations in games used in the rehabilitation of post-stroke patients. From 309 articles identified 31 respected all inclusion/exclusion criteria and were selected for further analysis. The fields investigated were methodology and multidisciplinarity of design, clinical evaluation with patients and game metric properties.

No game design methodologies were observed. Almost half of the studies were conducted by joint authorship of the technology. No information was found about the development team. Only half of the studies evaluated therapeutic outcomes. The relationship between game score and clinical evaluation was not found. To overcome these drawbacks a conceptual model called STRONGER — SysTems for Rehabilitation based on GamE scoRe — for SG design for post-stroke rehabilitation was proposed (Fig. 1).

Regarding the inconsistency of studies in SG, a framework as STRONGER that establishes a relationship among the key stakeholders (experts and patient) and elements (biomedical device, SG and game score) should be used in SG design. The score aimed for clinical evaluation is an innovative aspect. The correlation between game score and clinical tests can lead the biomedical system to become an aid for treatment and evaluation. Although this study has been limited to post-stroke rehabilitation, the proposed model can be extended to other populations.