EVALUASI PENGUKURAN VOLUME PENDARAHAN TERHADAP METODE AUTOMATIC VOXEL CALCUATION HOUNSFIELD UNIT PADA MULTI SLICE COMPUTED TOMOGRAPHY BRAIN NON CONTRAST DI RUMAH SAKIT UMUM CUT MEUTIA ACEH UTARA
ABSTRACT
BACHELOR PROGRAM OF APPLIED
RADIOLOGY TECHNOLOGY IMAGING
RADIODIAGNOSTIC TECHNIQUES AND RADIOTHERAPY
POLTEKKES MINISTRY OF HEALTH JAKARTA II
THESIS, 2025
ASNAWIAH
EVALUATION OF HAEMORRHAGE VOLUME MEASUREMENT USING THE
AUTOMATIC VOXEL CALCULATION HOUNSFIELD UNIT METHOD IN NONCONTRAST MSCT BRAIN IMAGING UNDERTAKEN AT A HOSPITAL
V Chapters + 50 Pages + 17 Images + 4 Tables + 9 Attachments
Background: quantitatively, this research explores the quantitative measurement of tissue density
values, defined as hyperdense and hypodense on CT scans, using Hounsfield Unit (HU) values. As
hematomas transition from hyperdense to hypodense over the time, HU measurement provides a
quantitative representation of this transformation. However, HU measurement may lead to
misinterpretation in the patients diagnosed with cerebral contusions or subarachnoid
haemorrhages.
Objective: purposefully, the objective of this research is to evaluate the haemorrhage volume in
non-contrast MSCT brain imaging using the automatic voxel calculation method based on
Hounsfield Units (HU). This research also aims to assess MSCT Brain non-contrast examination
technique for brain haemorrhage diagnosis, evaluate the measurement process using HU-based
method for haemorrhage information, and determine the optimal HU range for accurate
assessment.
The research design used in this study is a qualitative descriptive study. The study was conducted
at the Radiology Installation of Cut Meutia Hospital, North Aceh. Conducted in September -
December 2024. The population of this study was all MSCT Brain examinations with a diagnosis
of brain hemorrhage, while the sample was 10 samples. The data collection methods used were
literature studies, direct observation and interviews. The research instruments were worksheets and
interview sheets. Data processing and analysis carried out are from observations and data collected
in documentation. Data processing and analysis in this study use descriptive analysis data. by
processing qualitative data obtained from the results of observations, evaluations and reinforced
with the results of expertise from radiologists then conducting interviews. all image data results are
then evaluated descriptively to obtain conclusions.
Research Findings : this research finds that each type of haemorrhage exhibits a distinct HU
range, especially in the cases with heterogeneous HU values. The automatic segmentation software
used detects haemorrhages based on the defined HU range. Furthermore, discrepancies between
the software's predefined HU range and the actual HU range of the haemorrhage can lead to
underestimation of haemorrhage volume. For that reason, this underestimation, caused by
mismatched HU ranges, can be corrected using additional software tools such as Smart Mesh.
In conclusion, based on an observation it revealed that the accuracy of haemorrhage volume
measurements increase in the cases where the haemorrhage is localized. However, the accuracy
decreases when the haemorrhage spreads near the bone or extends into the ventricles.
Keywords: Hounsfield Unit, Underestimation, Smart Mesh
Tidak ada salinan data
Tidak tersedia versi lain