Characterization of cerebral autoregulation in stroke patients

• Grant number: 21/10288-8
• Support Opportunities: Scholarships in Brazil - Master
• Start date: July 01, 2022
• End date: August 23, 2023
• Field of knowledge: Engineering - Biomedical Engineering - Bioengineering
• Principal Investigator: João Loures Salinet Júnior
• Grantee: Renata Romanelli da Costa
• Host Institution: Centro de Engenharia, Modelagem e Ciências Sociais Aplicadas (CECS). Universidade Federal do ABC (UFABC). Ministério da Educação (Brasil). Santo André , SP, Brazil

Abstract

The Cerebral Autoregulation (CA) mechanism plays a crucial role in brain homeostasis, being responsible for keeping the Cerebral Blood Flow (CBF) constant despite variations in Blood Pressure (BP). The mechanism of CA is critically challenging in stroke, as this cerebrovascular disease significantly alters the entire brain hemodynamics. While remaining functional, CA can protect the brain from BP levels and minimize neurological damage, especially in the stroke phase. There is evidence that CA is compromised at this stage, but these results are based on single-center studies with a limited number of participants. A better understanding of the functioning of CA in stroke may contribute to the development of predictive indicators of neurological evolution, as well as therapeutic strategies. The general objective of this master's project is to investigate the relationship between CA methods and clinical data from post-stroke patients, in order to improve the assessment of CA in these patients, through an international multicenter study. Data from 120 post-stroke patients will be provided by six national and international research centers associated with the CARNet research network (Cerebral Autoregulation Research Network). For the evaluation of CA, pre-processing methods, Transfer Function Analysis (TFA), Auto-Regulation Index (ARI), Autoregressive ARI with Moving Average (ARI-ARMA) and Mean Flow Index (Mx) will be applied. After applying the methods, a comparison will be made of the normative values of CA obtained in the ARI method with other methods mentioned. In this comparison, the possibility of identifying a threshold in the metrics of each method will be investigated, allowing the classification between the groups of compromised and non-compromised CA. This can be done, for example, with the Receiver Operating Characteristic Curve (ROC curve). In addition, machine learning methods will be used to perform clustering and classification, which can reveal groups of post-stroke patients with similar behaviors, which had not been perceived only through the ARI, helping to translate the TFA, ARMA-ARI and Mx indices to the clinic. The results of this study will be pioneers for using a significant amount of data to evaluate the CA in a complex disease, with the aim of incorporating them into the first Open Source platform developed for CA calculation, called Cerebral Autoregulation Assessment open source Platform (CAAos).