Till sidans topp

Sidansvarig: Webbredaktion
Sidan uppdaterades: 2012-09-11 15:12

Tipsa en vän
Utskriftsversion

Improved detection rate a… - Göteborgs universitet Till startsida
Webbkarta
Till innehåll Läs mer om hur kakor används på gu.se

Improved detection rate and visualization of liver uptake foci in diagnostic 111In-octreotide SPECT/CT investigations with a novel segmentation analysis

Konferensbidrag (offentliggjort, men ej förlagsutgivet)
Författare Tobias Magnander
Emma Wikberg
Johanna Svensson
Peter Gjertsson
Bo Wängberg
Magnus Båth
Peter Bernhardt
Publicerad i Optimisation in X-ray and Molecular Imaging 2015 - the Fourth Malmö Conference on Medical Imaging, Gothenburg, Sweden, 28-30 May 2015
Publiceringsår 2015
Publicerad vid Institutionen för kliniska vetenskaper, Avdelningen för radiofysik
Institutionen för kliniska vetenskaper, Avdelningen för kirurgi
Institutionen för kliniska vetenskaper, Avdelningen för onkologi
Institutionen för medicin, avdelningen för molekylär och klinisk medicin
Språk en
Länkar www.delegia.com/app/Data/ProjectIma...
https://gup.ub.gu.se/file/191738
Ämneskategorier Radiofysik, Radiologi, Cancer och onkologi

Sammanfattning

Purpose: Detection of liver tumors will change the course of treatment of neuroendocrine tumours. In nuclear medicine 111In-octreoscan is of high value for detection of neuroendocrine tumours. However, neuroendocrine tumours disseminated to the livers is often challenging to detect from 111In-octreoscan SPECT images due to low uptake, high noise, poor resolution and low contrast. The aim of the present study was to develop a segmentation analysis method for increased diagnostic accuracy of neuroendocrine liver tumours. Methods: For the SPECT reconstruction 120 projections are acquired with 3 degrees spacing around the patient injected with 111In-octreoscan. The projections are reconstructed into a 128x128x128 voxel matrix using OSEM with CT based attenuation correction. The liver is segmented from the SPECT or CT using either an isosurface, region growing or a GPU accelerated level set algorithm. Manual editing finishes the segmentation of the liver. The segmented liver volume of interest, liver VOI, is thresholded at 125 equidistant threshold values between 0 and the maximum voxel value. At each threshold value a connected component labeling algorithm is used to calculate the number of uptake foci (NUF). The normalized NUF (nNUF) is then plotted against the threshold index (ThI), defined as ThI=(cmax-cthr)/cmax, where cmax is the maximal voxel value in the VOI, and cthr is the voxel threshold value. The method is named nNUFTI - normalized Number of Uptake Foci vs ThI. The ThI at 0.25 nNUF was used for analysis of liver tumour involvement. SPECT images from 53 patients without tumour involvement (i.e SPECT negative) in the liver were analysed with nNUFTI. A three year follow up with MRI, SPECT, PET/CT and CT was used to separate the patients into two groups: the healthy group, with still no liver tumours, and the malignant group, shown to have developed tumours in the liver. Results: 40 patients ended up in the healthy group and 13 in the malignant group. The ThI at 0.25 nNUF was significantly different between the groups (p<0.01). A probability function for the ThI values was constructed from the obtained data. This relationship might be a useful guide in the diagnostic decision making. Conclusions: Our new developed method nNUFTI has been shown to perform well. More studies on the nNUFTI method are needed.

Sidansvarig: Webbredaktion|Sidan uppdaterades: 2012-09-11
Dela:

På Göteborgs universitet använder vi kakor (cookies) för att webbplatsen ska fungera på ett bra sätt för dig. Genom att surfa vidare godkänner du att vi använder kakor.  Vad är kakor?