Banca de DEFESA: RENATA ROMANELLI DA COSTA

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : RENATA ROMANELLI DA COSTA
DATE: 31/10/2023
TIME: 10:00
LOCAL: HeartLab, Bloco Zeta, Campus São Bernardo do Campo - UFABC
TITLE:

Characterization of cerebral autoregulation in patients with stroke and hypertension

PAGES: 90
BIG AREA: Engenharias
AREA: Engenharia Biomédica
SUBÁREA: Bioengenharia
SPECIALTY: Processamento de Sinais Biológicos
SUMMARY:

The human brain relies on a consistent supply of oxygen and energy substrates through cerebral blood flow (CBF). The occurrence of ischemic stroke can severely impact this supply by altering cerebral perfusion, which depends on cerebral arteries, collateral circulation, and compensatory mechanisms. The mechanism of cerebral autoregulation (CA) maintains a constant CBF during changes in systemic blood pressure (BP) between 60 and 140 mmHg. However, when compromised, CBF becomes dependent on BP, resulting in critical conditions for the brain. There is evidence that CA is compromised in the acute phase of ischemic stroke (AIS), but these findings are based on single-center studies with a limited number of participants. A better understanding of CA function in AIS could contribute to the development of neurological outcome prediction indicators as well as therapeutic strategies. The research project aims to investigate the relationship between CA methods and clinical and hemodynamic data of AIS patients. Data from 110 patients (50 control and 60 AIS) were provided by two research centers, one national and one international, associated with the Cerebral Autoregulation Research Network (CARNet). To assess CA, preprocessing methods such as resampling, calibration, and filtering are applied, followed by the application of four CA calculation metrics: Transfer Function Analysis (TFA), Classic Cerebral Autoregulation Index (Classic ARI), Cerebral Autoregulation Index calculated by the Autoregressive with Moving Average model (ARI-ARMA), and non-invasive Mean Flow Index (nMx). After applying the metrics, a comparison of normative CA values obtained from the mentioned methods is made, dividing the results into groups of mild, moderate, and severe AIS patients. These results are then compared with the clinical data of patients to investigate their relationship. In addition, machine learning methods (Support Vector Machine, Decision Tree, Naive Bayes, and K-Nearest Neighbors) are used for CA classification, potentially revealing groups of AIS patients with similar behaviors that were not apparent through Classic ARI alone, assisting in translating TFA, ARI-ARMA, and nMx indices into clinical practice. The results of this study are pioneering in their use of a significant amount of data to assess CA in a complex disease, with the goal of incorporating them into the first Open Source platform developed for CA calculation, called the Cerebral Autoregulation Assessment Open Source Platform (CAAos). It should be noted that this platform was developed by the research group at the Federal University of ABC, of which the student is a part and participated in the development of three of the four CA metrics present in the platform today. Up to the present moment, this study has encompassed various stages, from preprocessing to the application of CA metrics, statistical analysis of clinical and hemodynamic features, and the development of a classification model using machine learning. By combining patient data from two clinical centers, the study has enhanced the generality and robustness of its findings, providing insights into the influence of clinical and hemodynamic factors on AIS. The analysis has revealed that the severity of AIS, assessed by the modified Ranking scale (mRS), is associated with functional disability, while differences in hemodynamic metrics suggest possible associations between blood pressure and AIS. Notably, the absence of differences in CA between affected and unaffected hemispheres can be attributed to collateral blood flow. CA metrics also showed significant differences between AIS subgroups and the control group, highlighting the importance of CA in pathophysiology and outcome prediction. Although further refinement is required for result classification, the findings point to the feasibility of developing more precise diagnostic tools. In summary, this study underscores the relevance of CA in AIS and validates the CAAos platform as a reliable tool for CA measurements.

COMMITTEE MEMBERS:
Presidente - Interno ao Programa - 2231661 - JOAO LOURES SALINET JUNIOR
Membro Titular - Examinador(a) Externo ao Programa - 1188948 - JOAO LAMEU DA SILVA JUNIOR
Membro Titular - Examinador(a) Externo à Instituição - RICARDO DE CARVALHO NOGUEIRA
Membro Suplente - Examinador(a) Interno ao Programa - 2188954 - ERICK DARIO LEON BUENO DE CAMARGO
Membro Suplente - Examinador(a) Externo à Instituição - MATHEUS CARDOSO MORAES - UNIFESP
Membro Suplente - Examinador(a) Externo à Instituição - ALESSANDRO PEREIRA DA SILVA