CMIV Publications

Background Stroke now represents the condition with the highest need for physical rehabilitation worldwide, with only low or moderate-level evidence testing telerehabilitation compared to in-person care. We compared functional ambulation in subacute patients with stroke following telerehabilitation and matched in-person controls with no biopsychosocial differences at baseline.Methods We conducted a matched case-control study to compare functional ambulation between individuals with stroke following telerehabilitation and in-person rehabilitation, assessed using the Functional Ambulation Categories (FAC) and the Functional Independence Measure (TM) (FIM).Results The telerehabilitation group (n = 38) achieved significantly higher FAC gains (1.5 (1.3) vs 1.0 (1.0)) than the in-person rehabilitation group, with no differences in ambulation efficiency, in individuals: admitted to rehabilitation within 60 days after stroke onset; aged 49.8 (+/- 11.4) years at admission; 55.3% female sex; moderate stroke severity; 42.1% with 'good' motor FIM at baseline; mostly living with sentimental partner (73.7%); with 21.1% holding an university education degree.Conclusions The groups showed no significant differences in ambulation efficiency, though the telerehabilitation group achieved higher FAC gains. Our results suggest that home telerehabilitation can be considered a good alternative to in-person rehabilitation when addressing ambulation in patients with moderate stroke severity and whose home situation mostly includes a cohabiting partner.

Purpose: To implement and evaluate deep learning-based methods for the classification of pediatric brain tumors in MR data.

Materials and methods: A subset of the “Children’s Brain Tumor Network” dataset was retrospectively used (n=178 subjects, female=72, male=102, NA=4, age-range [0.01, 36.49] years) with tumor types being low-grade astrocytoma (n=84), ependymoma (n=32), and medulloblastoma (n=62). T1w post-contrast (n=94 subjects), T2w (n=160 subjects), and ADC (n=66 subjects) MR sequences were used separately. Two deep-learning models were trained on transversal slices showing tumor. Joint fusion was implemented to combine image and age data, and two pre-training paradigms were utilized. Model explainability was investigated using gradient-weighted class activation mapping (Grad-CAM), and the learned feature space was visualized using principal component analysis (PCA).

Results: The highest tumor-type classification performance was achieved when using a vision transformer model pre-trained on ImageNet and fine-tuned on ADC images with age fusion (MCC: 0.77 ± 0.14 Accuracy: 0.87 ± 0.08), followed by models trained on T2w (MCC: 0.58 ± 0.11, Accuracy: 0.73 ± 0.08) and T1w post-contrast (MCC: 0.41 ± 0.11, Accuracy: 0.62 ± 0.08) data. Age fusion marginally improved the model’s performance. Both model architectures performed similarly across the experiments, with no differences between the pre-training strategies. Grad-CAMs showed that the models’ attention focused on the brain region. PCA of the feature space showed greater separation of the tumor-type clusters when using contrastive pre-training.

Conclusion: Classification of pediatric brain tumors on MR-images could be accomplished using deep learning, with the top-performing model being trained on ADC data, which is used by radiologists for the clinical classification of these tumors.