Classification of Alzheimer's disease: application of a transfer learning deep Q-network method.

Early diagnosis is crucial to slowing the progression of Alzheimer's disease (AD), so it is urgent to find an effective diagnostic method for AD. This study intended to investigate whether the transfer learning approach of deep Q-network (DQN) could effectively distinguish AD patients using local metrics of resting-state functional magnetic resonance imaging (rs-fMRI) as features. This study included 1310 subjects from the Consortium for Reliability and Reproducibility (CoRR) and 50 subjects from the Alzheimer's Disease Neuroimaging Initiative (ADNI) GO/2. The amplitude of low-frequency fluctuation (ALFF), fractional ALFF (fALFF) and percent amplitude of fluctuation (PerAF) were extracted as features using the Power 264 atlas. Based on gender bias in AD, we searched for transferable similar parts between the CoRR feature matrix and the ADNI feature matrix, resulting in the CoRR similar feature matrix served as the source domain and the ADNI similar feature matrix served as the target domain. A DQN classifier was pre-trained in the source domain and transferred to the target domain. Finally, the transferred DQN classifier was used to classify AD and healthy controls (HC). A permutation test was performed. The DQN transfer learning achieved a classification accuracy of 86.66% (p < 0.01), recall of 83.33% and precision of 83.33%. The findings suggested that the transfer learning approach using DQN could be an effective way to distinguish AD from HC. It also revealed the potential value of local brain activity in AD clinical diagnosis.

Effects of Fiber Shape on Mechanical Properties of Fiber Assemblies.

The effects of fiber shape on the mechanical responses of fiber assemblies under compression, tension, and shear deformations are numerically investigated using the discrete element method (DEM). Simulations of the compression of ring-shaped fibers are consistent with experimental results, verifying the discrete element method code. In the compressive tests of S-shaped fibers, pressure exhibits a nonmonotonic dependence on fiber curvature; while in the tensile tests, yield tensile stress generally decreases with increasing fiber curvature. In the shear tests, yield shear stress decreases with increasing fiber curvature for the S-shaped fibers, and the smallest yield shear stresses and the smallest coordination numbers are obtained for U-shaped and Z-shaped fibers. It is interesting to observe that for the assemblies of various fiber shapes, yield shear stress increases with increasing maximum Feret diameter of the fibers, which characterizes the largest dimension of a fiber between two parallel tangential lines. These novel observations of the effects of fiber shape provide some guidelines for material designs with the fibers.

Abnormal amygdala functional connectivity and deep learning classification in multifrequency bands in autism spectrum disorder: A multisite functional magnetic resonance imaging study.

Previous studies have explored resting-state functional connectivity (rs-FC) of the amygdala in patients with autism spectrum disorder (ASD). However, it remains unclear whether there are frequency-specific FC alterations of the amygdala in ASD and whether FC in specific frequency bands can be used to distinguish patients with ASD from typical controls (TCs). Data from 306 patients with ASD and 314 age-matched and sex-matched TCs were collected from 28 sites in the Autism Brain Imaging Data Exchange database. The bilateral amygdala, defined as the seed regions, was used to perform seed-based FC analyses in the conventional, slow-5, and slow-4 frequency bands at each site. Image-based meta-analyses were used to obtain consistent brain regions across 28 sites in the three frequency bands. By combining generative adversarial networks and deep neural networks, a deep learning approach was applied to distinguish patients with ASD from TCs. The meta-analysis results showed frequency band specificity of FC in ASD, which was reflected in the slow-5 frequency band instead of the conventional and slow-4 frequency bands. The deep learning results showed that, compared with the conventional and slow-4 frequency bands, the slow-5 frequency band exhibited a higher accuracy of 74.73%, precision of 74.58%, recall of 75.05%, and area under the curve of 0.811 to distinguish patients with ASD from TCs. These findings may help us to understand the pathological mechanisms of ASD and provide preliminary guidance for the clinical diagnosis of ASD.

The White Matter Functional Abnormalities in Patients with Transient Ischemic Attack: A Reinforcement Learning Approach.

Background: Transient ischemic attack (TIA) is a known risk factor for stroke. Abnormal alterations in the low-frequency range of the gray matter (GM) of the brain have been studied in patients with TIA. However, whether there are abnormal neural activities in the low-frequency range of the white matter (WM) in patients with TIA remains unknown. The current study applied two resting-state metrics to explore functional abnormalities in the low-frequency range of WM in patients with TIA. Furthermore, a reinforcement learning method was used to investigate whether altered WM function could be a diagnostic indicator of TIA. Methods: We enrolled 48 patients with TIA and 41 age- and sex-matched healthy controls (HCs). Resting-state functional magnetic resonance imaging (rs-fMRI) and clinical/physiological/biochemical data were collected from each participant. We compared the group differences between patients with TIA and HCs in the low-frequency range of WM using two resting-state metrics: amplitude of low-frequency fluctuation (ALFF) and fractional ALFF (fALFF). The altered ALFF and fALFF values were defined as features of the reinforcement learning method involving a Q-learning algorithm. Results: Compared with HCs, patients with TIA showed decreased ALFF in the right cingulate gyrus/right superior longitudinal fasciculus/left superior corona radiata and decreased fALFF in the right cerebral peduncle/right cingulate gyrus/middle cerebellar peduncle. Based on these two rs-fMRI metrics, an optimal Q-learning model was obtained with an accuracy of 82.02%, sensitivity of 85.42%, specificity of 78.05%, precision of 82.00%, and area under the curve (AUC) of 0.87. Conclusion: The present study revealed abnormal WM functional alterations in the low-frequency range in patients with TIA. These results support the role of WM functional neural activity as a potential neuromarker in classifying patients with TIA and offer novel insights into the underlying mechanisms in patients with TIA from the perspective of WM function.

Altered Temporal Dynamics of Brain Activity in Multiple-Frequency Bands in Non-Neuropsychiatric Systemic Lupus Erythematosus Patients with Inactive Disease.

PURPOSE: In this study, we seek to investigate dynamic changes of brain activity in non-neuropsychiatric systemic lupus erythematosus (non-NPSLE) patients with inactive disease. PATIENTS AND METHODS: Thirty-one non-NPSLE patients with inactive disease and 20 matched healthy controls underwent the blood oxygenation level-dependent fMRI examination. Dynamic regional homogeneity (ReHo) and fractional amplitude of low-frequency fluctuations (fALFF) were used to analyze the brain activity in typical band (0.01-0.08 Hz), slow-4 (0.027-0.073 Hz) and slow-5 (0.01-0.027 Hz). Pearson's correlation analysis was performed to correlate dynamic regional homogeneity (dReHo) and dynamic fractional amplitude of low-frequency fluctuations (dfALFF) values for clusters of voxels where significant group differences were found with clinical variables in non-NPSLE patients with inactive disease. RESULTS: In typical band, non-NPSLE patients showed increased dReHo in left middle occipital gyrus (MOG) compared to healthy controls. Meanwhile, patients showed decreased dfALFF in right superior frontal gyrus (SFG) and bilateral middle frontal gyrus (MFG) in typical band. In slow-4, increased dReHo in left MOG was found in non-NPSLE patients. In slow-5, non-NPSLE patients showed increased dReHo in left MOG, left calcarine fissure and surrounding cortex, right precentral gyrus (PreCG) and left postcentral gyrus (PoCG). Meanwhile, non-NPSLE patients showed decreased dfALFF in left SFG, right MFG, and right PreCG in slow-5. Moreover, the glucocorticoid dose showed significantly negative correlations with dReHo values in right PreCG in slow-5, left PoCG in slow-5, and left MOG in typical band. CONCLUSION: dReHo and dfALFF abnormalities in different frequency bands may be the key characteristics in the pathogenesis mechanism of non-NPSLE.

Resting-state abnormalities in functional connectivity of the default mode network in autism spectrum disorder: a meta-analysis.

Increasing evidence has shown that the resting state brain connectivity of default mode network (DMN) which are important for social cognition are disrupted in autism spectrum disorder (ASD). However, previous neuroimaging studies did not present consistent results. Therefore, we performed a meta-analysis of resting-state functional connectivity (rsFC) studies of DMN in the individuals with ASD and healthy controls (HCs) to provide a new perspective for investigating the pathophysiology of ASD. We carried out a search using the terms: ("ASD" OR "Autism") AND ("resting state" OR "rest") AND ("DMN" OR "default mode network") in PubMed, Web of Science and Embase to identify the researches published before January 2020. Ten resting state datasets including 203 patients and 208 HCs were included. Anisotropic Effect Size version of Signed Differential Mapping (AES-SDM) method was applied to identify group differences. In comparison with the HCs, the patients with ASD showed increased connectivity in cerebellum, right middle temporal gyrus, superior occipital gyrus, right supramarginal gyrus, supplementary motor area and putamen. Decreased connectivity was discovered in some nodes of DMN, such as medial prefrontal cortex, precuneus and angular gyrus. These results may help us to further clarify the neurobiological mechanisms in patients with ASD.

Identify abnormal functional connectivity of resting state networks in Autism spectrum disorder and apply to machine learning-based classification.

Autism spectrum disorder (ASD) patients are often reported altered patterns of functional connectivity (FC) on resting-state functional magnetic resonance imaging (rsfMRI) scans. However, the results in similar brain regions were inconsistent. In this study, we first investigated statistical differences in large-scale resting-state networks (RSNs) on 192 healthy controls (HCs) and 103 ASD patients by using independent component analysis (ICA). Second, an image-based meta-analysis (IBMA) was applied to discover the consistency of spatial patterns from different sites. Last, utilizing these patterns as features, we used Support Vector Machine (SVM) classifier to identify whether a subject was suffering from ASD or not. As a result, six RSNs were obtained with ICA. In each RSN, we identified altered functional connectivity between ASD and HC across the multi-site data. We calculated the area under the receiver operating characteristic curve plots (AUC) to determine the classification performance. The AUC value of classification reaches 0.988. In conclusion, the present study indicates that intrinsic connectivity patterns produced from rsfMRI data could yield a possible biomarker of ASD and contributed to the neurobiology of ASD.

The Predictive Value of Dynamic Intrinsic Local Metrics in Transient Ischemic Attack.

BACKGROUND: Transient ischemic attack (TIA) is known as "small stroke." However, the diagnosis of TIA is currently difficult due to the transient symptoms. Therefore, objective and reliable biomarkers are urgently needed in clinical practice. OBJECTIVE: The purpose of this study was to investigate whether dynamic alterations in resting-state local metrics could differentiate patients with TIA from healthy controls (HCs) using the support-vector machine (SVM) classification method. METHODS: By analyzing resting-state functional MRI (rs-fMRI) data from 48 patients with and 41 demographically matched HCs, we compared the group differences in three dynamic local metrics: dynamic amplitude of low-frequency fluctuation (d-ALFF), dynamic fractional amplitude of low-frequency fluctuation (d-fALFF), and dynamic regional homogeneity (d-ReHo). Furthermore, we selected the observed alterations in three dynamic local metrics as classification features to distinguish patients with TIA from HCs through SVM classifier. RESULTS: We found that TIA was associated with disruptions in dynamic local intrinsic brain activities. Compared with HCs, the patients with TIA exhibited increased d-fALFF, d-fALFF, and d-ReHo in vermis, right calcarine, right middle temporal gyrus, opercular part of right inferior frontal gyrus, left calcarine, left occipital, and left temporal and cerebellum. These alternations in the dynamic local metrics exhibited an accuracy of 80.90%, sensitivity of 77.08%, specificity of 85.37%, precision of 86.05%, and area under curve of 0.8501 for distinguishing the patients from HCs. CONCLUSION: Our findings may provide important evidence for understanding the neuropathology underlying TIA and strong support for the hypothesis that these local metrics have potential value in clinical diagnosis.

Altered activation in sensorimotor network after applying rTMS over the primary motor cortex at different frequencies.

INTRODUCTION: Repetitive transcranial magnetic stimulation (rTMS) over the primary motor cortex (M1) can modulate brain activity both in the stimulated site and remote brain areas of the sensorimotor network. However, the modulatory effects of rTMS at different frequencies remain unclear. Here, we employed finger-tapping task-based fMRI to investigate alterations in activation of the sensorimotor network after the application of rTMS over the left M1 at different frequencies. MATERIALS AND METHODS: Forty-five right-handed healthy participants were randomly divided into three groups by rTMS frequency (HF, high-frequency, 3 Hz; LF, low-frequency, 1 Hz; and SHAM) and underwent two task-fMRI sessions (RH, finger-tapping with right index finger; LH, finger-tapping with left index finger) before and after applying rTMS over the left M1. We defined regions of interest (ROIs) in the sensorimotor network based on group-level activation maps (pre-rTMS) from RH and LH tasks and calculated the percentage signal change (PSC) for each ROI. We then assessed the differences of PSC within HF or LF groups and between groups. RESULTS: Application of rTMS at different frequencies resulted in a change in activation of several areas of the sensorimotor network. We observed the increased PSC in M1 after high-frequency stimulation, while we detected the reduced PSC in the primary sensory cortex (S1), ventral premotor cortex (PMv), supplementary motor cortex (SMA), and putamen after low-frequency stimulation. Moreover, the PSC in the SMA, dorsal premotor cortex (PMd), and putamen in the HF group was higher than in the LF group after stimulation. CONCLUSION: Our findings suggested that activation alterations within sensorimotor network are dependent on the frequency of rTMS. Therefore, our findings contribute to understanding the effects of rTMS on brain activation in healthy individuals and ultimately may further help to suggest mechanisms of how rTMS could be employed as a therapeutic tool.

The low-frequency BOLD signal oscillation response in the insular associated to immediate analgesia of ankle acupuncture in patients with chronic low back pain.

PURPOSE: The present study evaluated the modulation effect of a painless acupuncture technique, ankle acupuncture (AA), on resting-state functional change in patients with chronic low back pain (CLBP). PATIENTS AND METHODS: Fourteen participants diagnosed with CLBP received AA and underwent one brain functional image scan after tactile stimulation and another one following the insertion of the needle. The needling sensations and clinical pain intensities were evaluated after the end of each functional image scan. The significance levels of Visual Analog Scales/Scores (VASs) before and after acupuncture were determined using paired t-test. The brain areas showing differences in the amplitude of low-frequency fluctuation (ALFF) and fractional ALFF (fALFF) between the two scans were identified. We also explored the relationship between mean ALFF values in brain areas identified and VAS scores based on Pearson correlation coefficient. RESULTS: A complete-case analysis was performed on 12 participants. Neither different needling sensations nor any local sensations during the two scans was found. The clinical findings indicated that the scores of VAS scores were significantly lower after AA intervention (P<0.001). Compared with those after tactile stimulation, ALFF decreased in the left insular and increased in the left precuneus and right precentral gyrus, and fALFF decreased in the left insular, during retaining of AA (corrected). Moreover, there was a positive correlation found between mean ALFF change in the left insular and that of VAS values (P<0.05). CONCLUSION: The present study demonstrated the low-frequency BOLD signal oscillation response in the left insular in brain activity was associated with an immediate analgesia of AA in patients with CLBP, which provides new insights into intrinsic connections between low-frequency brain signals and analgesic effects of acupuncture.

Deregulation of Snai2 is associated with metastasis and poor prognosis in tongue squamous cell carcinoma.

The members of the Snail superfamily of zinc-finger transcription factors, including Snai1 and Snai2, are involved in essential biological processes, such as epithelial-mesenchymal transition (EMT). Although Snai1 has been investigated in a number of cancers, our knowledge on Snai2 and its role(s) in squamous cell carcinoma of oral tongue (SCCOT) is limited. In this study, we confirmed the previous observation that over-expression of Snai2 is a frequent event in SCCOT. We further demonstrated that Snai2 over-expression is associated with lymph node metastasis in two independent SCCOT patient cohorts (total n = 129). Statistical analysis revealed that Snai2 over-expression was correlated with reduced overall survival. Furthermore, over-expression of Snai2 was correlated with reduced E-cadherin expression and enhanced Vimentin expression, suggesting a functional role of Snai2 in EMT. These observations were confirmed in vitro, in which knockdown of Snai2 induced a switch from a mesenchymal-like morphology to an epithelial-like morphology in SCCOT cell lines, and suppressed the cell invasion and migration. In contrast, ectopic transfection of Snai2 led to enhanced cell invasion and migration. Furthermore, Snai2 knockdown attenuated TGFß1-induced EMT in SCCOT cell lines. Taken together, these data suggest that Snai2 plays major roles in EMT and the progression of SCCOT and may serve as a therapeutic target for patients at risk of metastasis.

The expression and significance of MRP1, LRP, TOPOIIß, and BCL2 in tongue squamous cell carcinoma.

BACKGROUND: Multiple drug resistance protein 1 (MRP1), lung resistance protein (LRP), topoisomerase IIß (TOPOIIß) and B-cell lymphoma 2 (BCL2) are well known in the development of drug resistance in cancer cells. The aim of this study was to evaluate the relationship between them and the clinicopathological features, their expression differences between tumor tissue and experimental drug-resistant model in tongue carcinoma. MATERIALS AND METHODS: Multiple drug resistance protein 1, LRP, TOPOIIß, and BCL2 expression was examined by immunohistochemistry in specimens from radical surgeries of 65 patients with tongue carcinoma. A cisplatin-resistance cell line, SCC-15/cisplatin, was established from a cisplatin-sensitive cell line, SCC-15. A MTT-based method was used to analyze drug potencies. Immunofluorescence was used to detect protein expression in both cell lines. Western blot was used to compare the protein expressions in specimens and SCC-15/cisplatin cells. RESULTS: We found higher expression of MRP1, LRP, and BCL2 and lower expression of TOPOIIß in tongue carcinoma compared with adjacent non-neoplastic tongue tissues (P < 0.05). In addition, MRP1 and TopoIIß expression were significantly associated with clinical stage, lymph node metastasis and histologic grade, and LRP was significantly associated with histologic grade in the samples (P < 0.05). Finally, Western blot showed that higher expressions of MRP1, LRP, and BCL2 and lower expression of TopoIIß were observed in SCC-15/cisplatin cells than in clinical samples. CONCLUSION: Our results suggest that the high expressions of MRP1, LRP, and BCL2 and low expression of TOPOIIß in patients with tongue carcinoma indicates that intrinsic drug resistance may exist in tongue carcinoma, and is associated with tumor differentiation and cisplatin resistance in tongue carcinoma.