Title :
Anatomical District Recognition in CT Bio-images
Author :
Angelo Ostinelli;Giuseppe Pozzi;Antonio Vassena
Author_Institution :
Dept. of Med. Phys., Ospedale S. Anna, Como, Italy
Abstract :
Use of X-ray radiations in bio-medical imaging, including computerized tomography (CT) examinations, may trigger biological effects to be taken in the proper account. In order to do that, the pure amount of radiations absorbed by the patient may not suffice, as the same quantity of radiations may trigger different biological effects according to the type of tissue actually irradiated. Despite most standards for bio-images include the storage of alpha-numerical tags (as the DICOM format does), the description of the anatomical district covered by the image may sometimes be incomplete, missing, or even erroneous. Anatomical District Recognition (ADR) aims at properly identifying the anatomical district depicted by an X-ray bio-image, without requiring any additional information but the image itself. By this paper, we propose a novel approach which reads the scout image of a CT examination and identifies the involved anatomical district(s), so that proper evaluations, including dose estimation, can then be performed.
Keywords :
"Computed tomography","Histograms","DICOM","Feature extraction","Indexes","Taxonomy"
Conference_Titel :
Healthcare Informatics (ICHI), 2015 International Conference on
DOI :
10.1109/ICHI.2015.45