Narrow Results

Subjective cognitive decline (SCD), characterized by self-perceived subtle cognitive impairment ahead of the appearance of explicit and measurable cognitive deficits, is regarded as the preclinical manifestation of the pathological change continuum of Alzheimer’s disease (AD). We were committed to exploring the amyloid and glucose metabolic signatures related to imminent brain metabolic changes in SCD subjects. This study included 39 subjects (mean age = 71.9 years; 14 males and 25 females) diagnosed with SCD disease and 39 gender-matched healthy controls (HCs) (mean age = 75.2; 16 males and 23 females) with brain [18F] fluorodeoxyglucose positron emission tomography (PET) images and [18F] florbetapir PET images. The standardized uptake value ratios (SUVRs) of PET images within the regions of interest (ROIs) were calculated. Inter-group SUVR differences were assessed by two-sample $t$-testing and receiver operating characteristic curve (ROC) analyses. A generalized linear model (GLM) was employed to evaluate the correlations between amyloid and FDG uptake. Compared with HCs, SCD subjects showed significantly increased amyloid SUVR, as well as significantly increased glucose SUVR in the olfactory, amygdala, thalamus, heschl gyrus, superior and middle temporal gyrus and temporal pole (all $P<0.05$). The amyloid SUVR of thalamus was found to have a better ROC result (area under the curve (AUC): 0.77, 95% confidence interval (CI): 0.66–0.86) in the HC group, as was the case with the glucose SUVR of the middle temporal gyrus (AUC: 0.83, 95% CI: 0.73–0.91). There were significant positive correlations between amyloid and glucose SUVRs ($P<0.05$). The amyloid SUVR of the thalamus showed a significantly better main effect (odd ratio $=$ 2.91, 95% CI: 1.44–6.7, $P<0.001$), and the glucose SUVR of the heschl gyrus indicated an enhanced main effect (odd ratio $=$ 5.08, 95% CI: 1.86–18.15, $P<0.001$). SCD subjects demonstrated significant amyloid accumulation and glucose hypermetabolism in specific brain regions, and amyloid pathology overlapped with regions of glucose abnormality. These findings may advance the understanding of imminent pathological changes in the SCD stage and help to provide clinical guidelines for interventional management.

Multi-modality image registration and fusion play an increasingly important role in medicine. 3-D image reconstruction, object recognition, and medical imaging analysis, are just a few examples. Medical images provide essential information for clinical diagnosis. Good image quality can yield more accurate patient information, which can then be used for better clinical decision making.

X-ray computed tomography (CT), magnetic resonance imaging (MRI), positron emission tomography (PET), and single photon emission computed tomography (SPECT) are clinically established imaging modalities. Among them, CT and MRI images are anatomical imaging with high spatial resolutions. However their physiological information is limited. On the other hand, although SPECT and PET can provide physiological information, both spatial resolutions are too poor to provide clear anatomical information. Thus, it would be advantageous to combine images from multi-modality images so that the resulting image can provide both physiological and anatomical information with high spatial resolution for use in clinical diagnosis and therapy.

In order to generate an overlapping image from different-modality images, image registration and various fusion techniques have been employed. In the past studies, artificial neural network (ANN) has been used in registration and fusion of multi-modality remote sensing data for the reconstruction of 3-D models of terrain regions and medical images. In this study, we apply multilayer back propagation (BP) and conjugate gradient back propagation (CGBP) NN model to train PET and MRI images. We use Matlab and Borland C++ Builder to implement the network training frame and find out the optimal training model for accurate image registration. Further, we evaluate the accuracy of the image registration using NN by phantom images. Our study shows that, the registration errors are less than 5mm, and using appropriate opacity is better to show the fusion image than using interleaved pixels with independent color scale. Finally, we use Borland C++ Builder to implement a user-friendly window-based interface for image registration and fusion system.

The quality of traditional two-dimensional image reconstruction for PET has been efficiently improved by three-dimensional image reconstruction, but the sensitivity of the data and the quality of the image are restricted by the limit of modality physics. In analytical image reconstruction algorithm, 3DRP method compensates the unmeasured events by forward projection based on the initial direct image estimate. However, the original 3DRP method merely depends on the parallel projections without taking into account the oblique projections. In our proposed 3DRP-SSRB method, we improve the first image estimate by incorporating the rebinned oblique data. SSRB method was used to perform the rebinning operation to make uses of the oblique projection data to improve the sensitivity information. And then project the improved image estimate forward and reconstruct the final image. Conflicting parameters of reconstructed image quality of 3DRP are experimented by simulated three-dimensional phantom study with regard to both system sensitivity and image quality factors. PET simulation software package was used to conduct the experiment along with the MATLAB software to evaluate the effectiveness of two-dimensional FBP, 3DRP, and our proposed 3DRP-SSRB methods. The result demonstrated its better image quality by having better mean squared error numbers in most of output image slices.