logo SBA

ETD

Digital archive of theses discussed at the University of Pisa

 

Thesis etd-01072025-161907


Thesis type
Tesi di specializzazione (4 anni)
URN
etd-01072025-161907
Thesis title
MIP-CT in the selection of stroke patients for endovascular therapy
Department
RICERCA TRASLAZIONALE E DELLE NUOVE TECNOLOGIE IN MEDICINA E CHIRURGIA
Course of study
RADIODIAGNOSTICA
Supervisors
.
relatore Prof. Cosottini, Mirco
Keywords
  • Mechanical thrombectomy
  • MIP
  • Perfusion
  • Selection criteria
  • Stroke
Graduation session start date
27/01/2025
Availability
Withheld
Release date
27/01/2095
Abstract (Inglese)
Abstract (Italiano)
Acute ischemic stroke (AIS) requires rapid intervention, with ischemic core volume assessment being critical for patients’ selection in medical and endovascular treatments. Currently, perfusion CT (CTP) is often used for this purpose, but it requires additional contrast, radiation, and specialized software. This study explores the potential of 8mm
maximum-intensity projection (MIP) reformats of non-contrast CT (NCCT) as an alternative for ischemic core volume estimation.
This retrospective study analyzed data from 42 patients treated with mechanical thrombectomy between 2021 and 2022. Ischemic core volumes were measured using three methods: manual segmentation of NCCT-MIP, automated CTP, and follow-up CT scans for final infarct volume (FIV). Correlation and concordance between the methods were assessed using Kendall’s tau correlation coefficient, Generalized Linear Models regression, and Bland-Altman plots.
MIP-based ischemic core correlated better with FIV (tau=0.43) compared to CTP (tau=0.39) and NCCT (tau=0.37). Bland-Altman Analysis showed improved concordance between MIP and FIV, with more regular systematic bias compared to CTP. MIP also demonstrated greater sensitivity in detecting ischemic hypodensity, suggesting it may be more reliable than CTP for early stroke assessment.
MIP on NCCT provides a reliable, contrast- and radiation-sparing method for ischemic core volume measurement in AIS patients, outperforming CTP in predicting final infarct volume. Further research is warranted to validate these findings in larger, prospective cohorts.
File