Keywords:
Ischaemia / Infarction, Comparative studies, Image manipulation / Reconstruction, CT, Head and neck, Emergency
Authors:
K. A. R. Balalta; Dasmarinas city/PH
DOI:
10.26044/ecr2019/C-2422
Methods and materials
A retrospective cross-sectional design was utilized using the review of record database of the Department of Radiology,
De La Salle University Medical Center in Dasmarinas,
Cavite,
Philippines,
from the period of January 2015 to December 2017.
Sixty patients aged 30 to 80 years old,
who were diagnosed with a hyperacute infarction on initial CT scan examination and had subsequent follow-up CT or MRI study confirming infarction,
were included in the study. The head CT scan images of these patients were reviewed using "brain windows" settings (window level of 35 HU and width of 100 HU) and the adjusted “stroke windows” (window level of 35 HU and width of 35 HU) settings.
All scans were obtained using an 8-slice CT-scan machine using the following protocol:
Scan type:
|
Axial
|
Rotation time:
|
2 seconds
|
Pitch:
|
1.65 seconds
|
Slice thickness (mm):
|
1.25
|
Interval (mm):
|
1.25
|
Three radiologists,
all of whom are fellows of the CT-MRI Society of the Philippines,
were asked to answer a questionnaire Fig. 2 created for this study.
Each page displayed a slice from each patient's CT scan examination wherein the hyperacute infarction was identified.
This slice was displayed using both "brain windows" and "stroke windows" on two separate pages,
for a total of two pages in the questionnaire per patient.
Below the image,
the responder was required to answer whether a hyperacute infarction was present.
Fig. 3 Eighty images of normal cranial CT scans,
forty of which were displayed in "brain windows" and forty in "stroke windows," were also included in the questionnaire to avoid response bias (that all images in the questionnaire were otherwise positive for hyperacute infarction).
These pages were then shuffled and shown to each respondent in random order.