| Resumo: | The aim of this work was to implement an automatic application (AA) for the quantification of the striatum transaxial slices based on [123 I]FP-CIT SPECT (Single Photon Emission Tomography) images. A sample of 68 subjects with and without pathology was collected from the nuclear medicine department of the Hospital Particular de Almada. In order to compare the striatum specific uptake ratios (SUR), a semi-automatic application (SAA) was also developed allowing manual adjustments by a specialist operator. The obtained results were compared with the DaTSCAN ® V4 application, from General Electric Healthcare. The experimental results based on SUR demonstrate that the AA can successfully discriminate healthy patients from the pathological subjects. Additionally, a strong correlation was verified between the AA and the semi-automatic methods (DaTSCAN ® V4 and the SAA). These methods also evidence intra and inter-operator variability, suppressed by the AA. The proposed methodology for segmentation and quantification of the striatum transaxial images obtained by [123 I]FP-CIT SPECT demonstrated that the developed application can accurately complement the visual analysis, requiring future optimization.
|
|---|