Reducing variations in multi-center Alzheimer's disease classification with convolutional adversarial autoencoder.

Based on brain magnetic resonance imaging (MRI), multiple variations ranging from MRI scanners to center-specific parameter settings, imaging protocols, and brain region-of-interest (ROI) definitions pose a big challenge for multi-center Alzheimer's disease characterization and classification. Existing approaches to reduce such variations require intricate multi-step, often manual preprocessing pipelines, including skull stripping, segmentation, registration, cortical reconstruction, and ROI outlining. Such procedures are time-consuming, and more importantly, tend to be user biased. Contrasting costly and biased preprocessing pipelines, the question arises whether we can design a deep learning model to automatically reduce these variations from multiple centers for Alzheimer's disease classification? In this study, we used T1 and T2-weighted structural MRI from Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset based on three groups with 375 subjects, respectively: patients with Alzheimer's disease (AD) dementia, with mild cognitive impairment (MCI), and healthy controls (HC); to test our approach, we defined AD classification as classifying an individual's structural image to one of the three group labels. We first introduced a convolutional adversarial autoencoder (CAAE) to reduce the variations existing in multi-center raw MRI scans by automatically registering them into a common aligned space. Afterward, a convolutional residual soft attention network (CRAT) was further proposed for AD classification. Canonical classification procedures demonstrated that our model achieved classification accuracies of 91.8%, 90.05%, and 88.10% for the 2-way classification tasks using the RAW aligned MRI scans, including AD vs. HC, AD vs. MCI, and MCI vs. HC, respectively. Thus, our automated approach achieves comparable or even better classification performance by comparing it with many baselines with dedicated conventional preprocessing pipelines. Furthermore, the uncovered brain hotpots, i.e., hippocampus, amygdala, and temporal pole, are consistent with previous studies.

Relationship between the maximum standardized uptake value of fluoro-2-deoxyglucose-positron emission tomography/computed tomography and clinicopathological characteristics in tongue squamous cell carcinoma.

AIM: Tongue carcinoma is one of the most common oral and maxillofacial malignant tumors worldwide, maximum standardized uptake value (SUVmax) in 18 F-fluoro-2-deoxyglucose-positron emission tomography/computed tomography (PET/CT) has been widely used in cancer research; however, there are few systematical reports on the relationship between SUVmax and clinicopathological characteristics in tongue squamous cell carcinoma (TSCC). This study aimed to investigate the relationship between them and whether SUV parameters can predict lymph node metastasis. MATERIALS AND METHODS: PET/CT manifestations and clinicopathological features of 52 patients with TSCC confirmed by pathology were retrospectively analyzed. Single-factor and multiple regression analyses were conducted on possible factors influencing TSCC SUVmax, including sex, age, smoking history, tumor location and size, histological differentiation, and tumor node metastasis (TNM) stages, T stages, and N stages. Diagnostic performance of SUVmax for lymph node metastasis was measured by the area under the receiver operating characteristic curve, and sensitivity and specificity were determined by Youden's J statistic. RESULTS: SUVmax was correlated with sex, tumor location and size, and TNM stages, T stages, and N stages (P < 0.05) but was not correlated with histological differentiation, smoking history, and age (P > 0.05). Sex, tumor location, tumor size, and N stage were influencing factors independent of TSCC SUVmax (P < 0.05). TSCC SUVmax had predictive value for lymph node metastasis. When the cutoff value was 6.57, the diagnostic efficiency was the highest, with the sensitivity being 79.2% and the specificity being 85.7%. CONCLUSIONS: SUVmax was higher among male patients with TSCC with posterior tumor location, larger tumor size, and lymph node metastasis, and TSCC SUVmax was important in predicting lymph node metastasis.

Correlation analysis between the SUVmax of FDG-PET/CT and clinicopathological characteristics in oral squamous cell carcinoma.

OBJECTIVES:: To investigate the relationship between maximum standardized uptake value (SUVmax) of 18F-FDG PET/CT and clinicopathological features of oral squamous cell carcinoma (OSCC), in order to formulate a better clinical guideline. METHODS:: In 104 patients with OSCC confirmed by pathology, there were 67 males and 37 females (age, 33-76 years; mean age, 56 years).18FDG, 18-fludeoxyglucose (18F-FDG) PET/CT manifestations and the clinicopathological features of the 104 patients were retrospectively analysed. Single-factor analysis and multiple regression analysis were conducted on possible factors influencing primary tumour SUVmax, including gender, age, smoking history, tumour location, tumour size, histological differentiation, TNM stage, T stage, N stage. Diagnostic performance of SUVmax for invading peri-tissue of OSCC was measured by the area under receiver operating characteristic curve, and sensitivity and specificity were determined at the Youdons index. RESULTS:: The single-analysis results showed that SUVmax was correlated with the histological differentiation, tumour size, TNM stage, T stage, N stage(p < 0.05), yet it was not correlated with gender, age, smoking history, tumour location (p > 0.05). Multivariate liner regression analysis showed that tumour size, TNM stage were influencing factors independent of primary tumour SUVmax (p < 0.05). Primary tumour SUVmax had predictive value for invading peri-tissue of OSCC. When the cutoff value was 7.98, the diagnostic efficiency was the highest, with the sensitivity being 90.0％ and the specificity being 76.2%. CONCLUSIONS:: OSCC 18F-FDG PET/CT SUVmax is higher among patients with larger tumour size, poorer stage, and that primary tumour SUVmax is of important significance in predicting invading peri-tissue.

Low-rank network signatures in the triple network separate schizophrenia and major depressive disorder.

Brain imaging studies have revealed that functional and structural brain connectivity in the so-called triple network (i.e., default mode network (DMN), salience network (SN) and central executive network (CEN)) are consistently altered in schizophrenia. However, similar changes have also been found in patients with major depressive disorder, prompting the question of specific triple network signatures for the two disorders. In this study, we proposed Supervised Convex Nonnegative Matrix Factorization (SCNMF) to extract distributed multi-modal brain patterns. These patterns distinguish schizophrenia and major depressive disorder in a latent low-dimensional space of the triple brain network. Specifically, 21 patients of schizophrenia and 25 patients of major depressive disorder were assessed by T1-weighted, diffusion-weighted, and resting-state functional MRIs. Individual structural and functional connectivity networks, based on pre-defined regions of the triple network were constructed, respectively. Afterwards, SCNMF was employed to extract the discriminative patterns. Experiments indicate that SCNMF allows extracting the low-rank discriminative patterns between the two disorders, achieving a classification accuracy of 82.6% based on the extracted functional and structural abnormalities with support vector machine. Experimental results show the specific brain patterns for schizophrenia and major depressive disorder that are multi-modal, complex, and distributed in the triple network. Parts of the prefrontal cortex including superior frontal gyri showed variation between patients with schizophrenia and major depression due to structural properties. In terms of functional properties, the middle cingulate cortex, inferior parietal lobule, and cingulate cortex were the most discriminative regions.