Stroke is a network disease



Neuroimaging-based analysis of a large group of stroke patients provides important insight in the complex interplay within the motor network during motor recovery after a stroke. The study, carried out by EPFL scientists with South Korean and German colleagues, shows the importance of changing stroke-rehabilitation strategies from one-size-fits-all towards more personalized, patient-tailored approaches. The work is published in Stroke.

Stroke is the main cause of long-term disability, with only 15% of patients recovering enough to re-enter their normal professional life. The unsatisfactory treatment might be the one-size-fits-all treatment strategy that does not take into account the heterogeneity of stroke patients and their recovery. It is therefore crucial to better understand the mechanisms of stroke recovery and develop innovative treatment strategies that can predict individual courses and degrees of recovery, and stratify patients for personalized treatments.

The brain’s proper function depends on a well-connected and orchestrated neuronal network. Single brain lesions, such as those following a stroke, do not only impair the function of the affected brain region itself, but also leads to changes in the underlying neuronal network. Therefore, stroke must be considered as a network disease, causing changes in structural and functional connectivity.

Stroke-related network disturbances are not uniform across patients, but heterogeneous depending on individual stroke-related (e.g., lesion site and location) and stroke-unrelated (e.g., individual structural connectivity) factors that significantly impact the functional role of specific network hubs or connections for the course and degree of recovery. However, the interactions between different network connections are largely unexplored, although they have great potential as ‘biomarkers’ for treatment stratification for individual patients.

In this study, the lab of Friedhelm Hummel at EPFL, with colleagues in South Korea and Hamburg, shows that stroke damage to the cortico-spinal tract (CST), can determine recovery, as it is the main connecting tract between the motor cortex (M1) and the spinal cord.

Furthermore, they show that stroke recovery is influenced by the degree of damage to the CST determines reorganization patterns and the functional relevance of specific network interactions, such as the ventral premotor (PMv) - motor cortex (M1) connection.

Future steps towards personalized precision medicine for stroke recovery

The study’s findings can pave the way towards the use of ‘biomarkers’ based on structural imaging to predict outcome and treatment responses of innovative, neurotechnology-based treatments such as brain stimulation.

“This study is an important step towards a better understanding of the network character of stroke, the underlying mechanisms, and reorganization patterns of recovery,” says Hummel. “It can pave the way towards precision-medicine approaches to enhance stroke recovery to a larger degree than current one-size-fits-all therapies.”

Such strategies are now being explored by Hummel’s lab in cooperation with the Hôpital de Valais Sion and the Clinique Romande de Réadaptation Sion.

Contributing institutions

  • EPFL
  • University Hospital Geneva, CH
  • University Medical Center Hamburg-Eppendorf, Germany
  • Sungkyunkwan University School of Medicine Seoul, Republic of Korea
  • Kyungpook National University Medical Center, Daegu, Republic of Korea


  • Defitech Foundation
  • National Research Foundation of Korea (NRF)


Robert Schulz, Eunhee Park, Jungsoo Lee, Won Hyuk Chang, Ahee Lee, Yun-Hee Kim, Friedhelm C. Hummel. Interactions Between the Corticospinal Tract and Premotor–Motor Pathways for Residual Motor Output After Stroke. Stroke 13 September 2017, 48:2805-2811. DOI: 10.1161/STROKEAHA.117.016834