Living well with diabetes in Alaska.

Many people with diabetes mellitus experience minimal or no complications. Our objective was to determine the proportion of Alaska Native people who experienced four major complications or mortality and to identify factors that may be associated with these outcomes. We used records in a diabetes registry and clinical and demographic variables in our analyses. We used logistic regression and Cox Proportional Hazards models to evaluate associations of these parameters with death and complications that occurred prior to 2013. The study included 591 Alaska Native people with non-type 1 diabetes mellitus, diagnosed between 1986 and 1992. Over 60% of people in this study remained free of four major diabetes-related complications for the remainder of life or throughout the approximately 20-year study period. Lower BMI, higher age at diagnosis of diabetes, and use of at least one diabetes medication were associated with death and a composite of four complications. A majority of Alaska Native people with DM had none of four major complications over a 20-year period. Lower BMI and use of diabetes medications were associated with higher hazard for some deleterious outcomes. This suggests that goals in care of elders should be carefully individualised. In addition, we discuss several programme factors that we believe contributed to favourable outcomes.

Evidence-based screening, clinical care and health education recommendations for Alaska Native peoples with prediabetes living in southcentral Alaska: findings from the Alaska EARTH follow-up study.

Pre-diabetes (pre-DM) is a strong predictor of diabetes (DM) over time. This study investigated how much of the recent increase in pre-DM identified among Alaska Native (AN) peoples living in urban southcentral Alaska may be due to changes in diagnostic methods. We used clinical and demographic data collected at baseline between 2004 and 2006 and at follow-up collected between 2015 and 2017 from the urban southcentral Alaska Education and Research Towards Health (EARTH) cohort. We used descriptive statistics and logistic regression to explore differences in demographic and clinical variables among the identified pre-DM groups. Of 388 participants in the follow-up study, 243 had A1c levels indicating pre-DM with only 20 demonstrating pre-DM also by fasting blood glucose (FBG). Current smoking was the sole predictor for pre-DM by A1c alone while abdominal obesity and elevated FBG-predicted pre-DM by A1c+FBG. No participants had an elevated FBG without an A1c elevation. A substantial portion of the rise in pre-DM found among urban southcentral AN peoples in the EARTH follow-up study was due to the addition of A1c testing. Pre-DM by A1c alone should be used to motivate behavioural changes that address modifiable risk factors, including smoking cessation, physical activity and weight management.

Gut microbiome association with brain imaging markers, APOE genotype, calcium and vegetable intakes, and obesity in healthy aging adults.

Introduction: Advanced age is a significant factor in changes to brain physiology and cognitive functions. Recent research has highlighted the critical role of the gut microbiome in modulating brain functions during aging, which can be influenced by various factors such as apolipoprotein E (APOE) genetic variance, body mass index (BMI), diabetes, and dietary intake. However, the associations between the gut microbiome and these factors, as well as brain structural, vascular, and metabolic imaging markers, have not been well explored. Methods: We recruited 30 community dwelling older adults between age 55-85 in Kentucky. We collected the medical history from the electronic health record as well as the Dietary Screener Questionnaire. We performed APOE genotyping with an oral swab, gut microbiome analysis using metagenomics sequencing, and brain structural, vascular, and metabolic imaging using MRI. Results: Individuals with APOE e2 and APOE e4 genotypes had distinct microbiota composition, and higher level of pro-inflammatory microbiota were associated higher BMI and diabetes. In contrast, calcium- and vegetable-rich diets were associated with microbiota that produced short chain fatty acids leading to an anti-inflammatory state. We also found that important gut microbial butyrate producers were correlated with the volume of the thalamus and corpus callosum, which are regions of the brain responsible for relaying and processing information. Additionally, putative proinflammatory species were negatively correlated with GABA production, an inhibitory neurotransmitter. Furthermore, we observed that the relative abundance of bacteria from the family Eggerthellaceae, equol producers, was correlated with white matter integrity in tracts connecting the brain regions related to language, memory, and learning. Discussion: These findings highlight the importance of gut microbiome association with brain health in aging population and could have important implications aimed at optimizing healthy brain aging through precision prebiotic, probiotic or dietary interventions.

Prebiotic inulin enhances gut microbial metabolism and anti-inflammation in apolipoprotein E4 mice with sex-specific implications.

Gut dysbiosis has been identified as a crucial factor of Alzheimer's disease (AD) development for apolipoprotein E4 (APOE4) carriers. Inulin has shown the potential to mitigate dysbiosis. However, it remains unclear whether the dietary response varies depending on sex. In the study, we fed 4-month-old APOE4 mice with inulin for 16 weeks and performed shotgun metagenomic sequencing to determine changes in microbiome diversity, taxonomy, and functional gene pathways. We also formed the same experiments with APOE3 mice to identify whether there are APOE-genotype dependent responses to inulin. We found that APOE4 female mice fed with inulin had restored alpha diversity, significantly reduced Escherichia coli and inflammation-associated pathway responses. However, compared with APOE4 male mice, they had less metabolic responses, including the levels of short-chain fatty acids-producing bacteria and the associated kinases, especially those related to acetate and Erysipelotrichaceae. These diet- and sex- effects were less pronounced in the APOE3 mice, indicating that different APOE variants also play a significant role. The findings provide insights into the higher susceptibility of APOE4 females to AD, potentially due to inefficient energy production, and imply the importance of considering precision nutrition for mitigating dysbiosis and AD risk in the future.

Self-Supervised Learning Application on COVID-19 Chest X-ray Image Classification Using Masked AutoEncoder.

The COVID-19 pandemic has underscored the urgent need for rapid and accurate diagnosis facilitated by artificial intelligence (AI), particularly in computer-aided diagnosis using medical imaging. However, this context presents two notable challenges: high diagnostic accuracy demand and limited availability of medical data for training AI models. To address these issues, we proposed the implementation of a Masked AutoEncoder (MAE), an innovative self-supervised learning approach, for classifying 2D Chest X-ray images. Our approach involved performing imaging reconstruction using a Vision Transformer (ViT) model as the feature encoder, paired with a custom-defined decoder. Additionally, we fine-tuned the pretrained ViT encoder using a labeled medical dataset, serving as the backbone. To evaluate our approach, we conducted a comparative analysis of three distinct training methods: training from scratch, transfer learning, and MAE-based training, all employing COVID-19 chest X-ray images. The results demonstrate that MAE-based training produces superior performance, achieving an accuracy of 0.985 and an AUC of 0.9957. We explored the mask ratio influence on MAE and found ratio = 0.4 shows the best performance. Furthermore, we illustrate that MAE exhibits remarkable efficiency when applied to labeled data, delivering comparable performance to utilizing only 30% of the original training dataset. Overall, our findings highlight the significant performance enhancement achieved by using MAE, particularly when working with limited datasets. This approach holds profound implications for future disease diagnosis, especially in scenarios where imaging information is scarce.

Formyl peptide receptor 1 is involved in surgery-induced neuroinflammation and dysfunction of learning and memory in mice.

Postoperative cognitive dysfunction (POCD) is a common complication after surgery. Peripheral immune cells may contribute to the development of POCD. However, molecules that are important for this contribution are not known. We hypothesize that formyl peptide receptor 1 (FPR1), a molecule critical for the migration of the monocytes and neutrophils into the brain after brain ischemia, is central to the development of postoperative neuroinflammation and dysfunction of learning and memory. Male C57BL/6 (wild-type) mice and FPR1-/- mice received right carotid artery exposure surgery. Some wild-type mice received cFLFLF, an FPR1 antagonist. Mouse brains were harvested 24 h after the surgery for biochemical analysis. Mice were subjected to the Barnes maze and fear conditioning tests to determine their learning and memory from 2 weeks after the surgery. We found that surgery increased FPR1 in the brain and proinflammatory cytokines in the blood and brain of wild-type mice. Surgery also impaired their learning and memory. cFLFLF attenuated these effects. Surgery did not induce an increase in the proinflammatory cytokines and impairment of learning and memory in FPR1-/- mice. These results suggest that FPR1 is important for the development of neuroinflammation and dysfunction of learning and memory after surgery. Specific interventions that inhibit FPR1 may be developed to reduce POCD.

The impact of preloaded intraocular lens implantation system (TECNIS iTec®) in routine cataract surgery in China: a time-motion analysis.

OBJECTIVE: To evaluate the impact on surgical efficiency and labor time cost of preloaded intraocular lens (IOL) implantation system compared with manual IOL implantation system in age-related cataract surgery in China. METHODS: This study was an observational, multicenter, prospective time-motion analysis. IOL preparation time, operation time, cleaning time, number and cost of cataract surgeries in eight participating hospitals were collected. The linear mixed model was used to explore factors associated with the difference in operation time between the preloaded IOL implantation system and the manual IOL implantation system. A time-motion model was constructed to convert the operation time cost saved by using preloaded IOL into economic benefits from hospital and social perspective, respectively. RESULTS: There were 2,591 cases included in the study (preloaded IOL: 1,591 cases; manual IOL: 1,000 cases). The preloaded IOL implantation system was significant time-saving in both preparation time and operation time compared to the manual IOL implantation system (25.48s vs. 47.04s, P < 0.001 and 353.84s vs. 367.46s, P = 0.004, respectively). An average total of 35.18s can be saved by using preloaded IOL per procedure. The results of linear mixed model showed that the type of IOL was the main factor leading to the difference in preparation time between preloaded IOL and manual IOL implantation system. By switching from manual IOL to preloaded IOL, the model projected additional 392 surgeries can be performed each year and an increase in revenue of $565,282 per hospital, a 9% increase from hospital perspective. And the annual productivity loss saved by using preloaded IOL was $3,006 in eight hospitals from perspective of society. CONCLUSION: Compared with manual IOL implantation system, the preloaded IOL implantation system reduces lens preparation time and operation time, which increases potential surgical volume and revenue, and reduces the loss of work productivity. This study provides real-world evidence to support the advantages of the preloaded IOL implantation system in improving efficiency of ophthalmic surgery in China.

Association of Sarcopenia with Cognitive Function and Dementia Risk Score: A National Prospective Cohort Study.

The relationship between skeletal muscle and cognitive disorders has drawn increasing attention. This study aims to examine the associations of sarcopenia with cognitive function and dementia risk score. Data on 1978 participants (aged 65 years and older) from the 2011 wave of the China Health and Retirement Longitudinal Study, with four follow-up waves to 2018, were used. Cognitive function was assessed by four dimensions, with a lower score indicating lower cognitive function. Dementia risk was assessed by a risk score using the Rotterdam Study Basic Dementia Risk Model (BDRM), with a higher score indicating a greater risk. Sarcopenia was defined when low muscle mass plus low muscle strength or low physical performance were met. We used generalized estimating equations to examine the associations of sarcopenia. In the fully adjusted models, sarcopenia was significantly associated with lower cognitive function (standardized, ß = -0.15; 95% CIs: -0.26, -0.04) and a higher BDRM score (standardized, ß = 0.42; 95% CIs: 0.29, 0.55). Our findings may provide a new avenue for alleviating the burden of cognitive disorders by preventing sarcopenia.

Efficient Training on Alzheimer's Disease Diagnosis with Learnable Weighted Pooling for 3D PET Brain Image Classification.

Three-dimensional convolutional neural networks (3D CNNs) have been widely applied to analyze Alzheimer's disease (AD) brain images for a better understanding of the disease progress or predicting the conversion from cognitively impaired (CU) or mild cognitive impairment status. It is well-known that training 3D-CNN is computationally expensive and with the potential of overfitting due to the small sample size available in the medical imaging field. Here we proposed a novel 3D-2D approach by converting a 3D brain image to a 2D fused image using a Learnable Weighted Pooling (LWP) method to improve efficient training and maintain comparable model performance. By the 3D-to-2D conversion, the proposed model can easily forward the fused 2D image through a pre-trained 2D model while achieving better performance over different 3D and 2D baselines. In the implementation, we chose to use ResNet34 for feature extraction as it outperformed other 2D CNN backbones. We further showed that the weights of the slices are location-dependent and the model performance relies on the 3D-to-2D fusion view, with the best outcomes from the coronal view. With the new approach, we were able to reduce 75% of the training time and increase the accuracy to 0.88, compared with conventional 3D CNNs, for classifying amyloid-beta PET imaging from the AD patients from the CU participants using the publicly available Alzheimer's Disease Neuroimaging Initiative dataset. The novel 3D-2D model may have profound implications for timely AD diagnosis in clinical settings in the future.

Targeting Astrocyte Signaling Alleviates Cerebrovascular and Synaptic Function Deficits in a Diet-Based Mouse Model of Small Cerebral Vessel Disease.

Despite the indispensable role that astrocytes play in the neurovascular unit, few studies have investigated the functional impact of astrocyte signaling in cognitive decline and dementia related to vascular pathology. Diet-mediated induction of hyperhomocysteinemia (HHcy) recapitulates numerous features of vascular contributions to cognitive impairment and dementia (VCID). Here, we used astrocyte targeting approaches to evaluate astrocyte Ca2+ dysregulation and the impact of aberrant astrocyte signaling on cerebrovascular dysfunction and synapse impairment in male and female HHcy diet mice. Two-photon imaging conducted in fully awake mice revealed activity-dependent Ca2+ dysregulation in barrel cortex astrocytes under HHcy. Stimulation of contralateral whiskers elicited larger Ca2+ transients in individual astrocytes of HHcy diet mice compared with control diet mice. However, evoked Ca2+ signaling across astrocyte networks was impaired in HHcy mice. HHcy also was associated with increased activation of the Ca2+/calcineurin-dependent transcription factor NFAT4, which has been linked previously to the reactive astrocyte phenotype and synapse dysfunction in amyloid and brain injury models. Targeting the NFAT inhibitor VIVIT to astrocytes, using adeno-associated virus vectors, led to reduced GFAP promoter activity in HHcy diet mice and improved functional hyperemia in arterioles and capillaries. VIVIT expression in astrocytes also preserved CA1 synaptic function and improved spontaneous alternation performance on the Y maze. Together, the results demonstrate that aberrant astrocyte signaling can impair the major functional properties of the neurovascular unit (i.e., cerebral vessel regulation and synaptic regulation) and may therefore represent a promising drug target for treating VCID and possibly Alzheimer's disease and other related dementias.SIGNIFICANCE STATEMENT The impact of reactive astrocytes in Alzheimer's disease and related dementias is poorly understood. Here, we evaluated Ca2+ responses and signaling in barrel cortex astrocytes of mice fed with a B-vitamin deficient diet that induces hyperhomocysteinemia (HHcy), cerebral vessel disease, and cognitive decline. Multiphoton imaging in awake mice with HHcy revealed augmented Ca2+ responses in individual astrocytes, but impaired signaling across astrocyte networks. Stimulation-evoked arteriole dilation and elevated red blood cell velocity in capillaries were also impaired in cortex of awake HHcy mice. Astrocyte-specific inhibition of the Ca2+-dependent transcription factor, NFAT, normalized cerebrovascular function in HHcy mice, improved synaptic properties in brain slices, and stabilized cognition. Results suggest that astrocytes are a mechanism and possible therapeutic target for vascular-related dementia.

Impact of local anesthesia block on pain medication use and length of hospital stay in elderly indigenous patients in Alaska hospitalized for fragility fracture.

Introduction: Fragility fractures (low-energy, minimal-trauma fractures) are common in the aging population and can lead to decreased function, increased mortality, and long-lasting pain. Although opioids are helpful in reducing acute postoperative pain, they present risks that may lead to increased morbidity and mortality. Materials and Methods: This was a retrospective review of medical records of all Alaska Native and American Indian people older than 50 years, who received surgery for hip fracture repair between January 2018 and June 2019 (n = 128). Results: We found that receipt of a peripheral nerve block (PNB) is a predictor for decreased length of hospital stay. However, receipt of PNB did not predict a reduction in postoperative morphine milligram equivalents opioid doses. Discussion: Further study is required to determine whether one PNB method is superior to others based on individual-level characteristics.

Functional recovery outcomes following acute stroke is associated with abundance of gut microbiota related to inflammation, butyrate and secondary bile acid.

Accumulating evidence suggests that gut microbes modulate brain plasticity via the bidirectional gut-brain axis and play a role in stroke rehabilitation. However, the microbial species alterations associated with stroke and their correlation with functional outcome measures following acute stroke remain unknown. Here we measure post-stroke gut dysbiosis and how it correlates with gut permeability and cognitive functions in 12 stroke participants, 18 controls with risk factors for stroke, and 12 controls without risk factors. Stool samples were used to measure the microbiome with whole genome shotgun sequencing and leaky gut markers. We genotyped APOE status and measured diet composition and motor, cognitive, and emotional status using NIH Toolbox. We used linear regression methods to identify gut microbial associations with cognitive and emotional assessments. We did not find significance differences between the two control groups. In contrast, the bacteria populations of the Stroke group were statistically dissimilar from the control groups. Relative abundance analysis revealed notable decreases in butyrate-producing microbial taxa, secondary bile acid-producing taxa, and equol-producing taxa. The Stroke group had higher levels of the leaky gut marker alpha-1-antitrypsin in the stool than either of the groups and several taxa including Roseburia species (a butyrate producer) were negatively correlated with alpha-1-antitrypsin. Stroke participants scored lower on memory testing than those in the two control groups. Stroke participants with more Roseburia performed better on the picture vocabulary task; more Bacteroides uniformis (a butyrate producer) and less Escherichia coli (a pro-inflammatory species) reported higher levels of self-efficacy. Intakes of fiber, fruit and vegetable were lower, but sweetened beverages were higher, in the Stroke group compared with controls. Vegetable consumption was correlated with many bacterial changes among the participants, but only the species Clostridium bolteae, a pro-inflammatory species, was significantly associated with stroke. Our findings indicate that stroke is associated with a higher abundance of proinflammatory species and a lower abundance of butyrate producers and secondary bile acid producers. These altered microbial communities are associated with poorer functional performances. Future studies targeting the gut microbiome should be developed to elucidate whether its manipulation could optimize rehabilitation and boost recovery.

Inulin Supplementation Mitigates Gut Dysbiosis and Brain Impairment Induced by Mild Traumatic Brain Injury during Chronic Phase.

Mild traumatic brain injury (mTBI) has been shown to acutely alter the gut microbiome diversity and composition, known as dysbiosis, which can further exacerbate metabolic and vascular changes in the brain in both humans and rodents. However, it remains unknown how mTBI affects the gut microbiome in the chronic phase recovery (past one week post injury). It is also unknown if injury recovery can be improved by mitigating dysbiosis. The goal of the study is to fill the knowledge gap. First, we aim to understand how mTBI alters the gut microbiome through the chronic period of recovery (3 months post injury). In addition, as the gut microbiome can be modulated by diet, we also investigated if prebiotic inulin, a fermentable fiber that promotes growth of beneficial bacteria and metabolites, would mitigate dysbiosis, improve systemic metabolism, and protect brain structural and vascular integrity when administered after 3 months post closed head injury (CHI). We found that CHI given to male mice at 4 months of age induced gut dysbiosis which peaked at 1.5 months post injury, reduced cerebral blood flow (CBF) and altered brain white matter integrity. Interestingly, we also found that Sham mice had transient dysbiosis, which peaked 24 hours after injury and then normalized. After 8 weeks of inulin feeding, CHI mice had increased abundance of beneficial/anti-inflammatory bacteria, reduced abundance of pathogenic bacteria, enriched levels of short-chain fatty acids, and restored CBF in both hippocampi and left thalamus, compared to the CHI-control fed and Sham groups. Using machine learning, we further identified top bacterial species that separate Sham and CHI mice with and without the diet. Our results indicate that there is an injury- and time-dependent dysbiosis between CHI and Sham mice; inulin is effective to mitigate dysbiosis and improve brain injury recovery in the CHI mice. As there are currently no effective treatments for mTBI, the study may have profound implications for developing therapeutics or preventive interventions in the future.

Extended Middle Cerebral Artery Occlusion (MCAO) Model to Mirror Stroke Patients Undergoing Thrombectomy.

Stroke remains a leading global cause of death and disability. In the last decade, the therapeutic window for mechanical thrombectomy has increased from a maximum of 6 to 24 h and beyond. While endovascular advancements have improved rates of recanalization, no post-stroke pharmacotherapeutics have been effective in enhancing neurorepair and recovery. New experimental models are needed to closer mimic the human patient. Our group has developed a model of transient 5-h occlusion in rats to mimic stroke patients undergoing thrombectomy. Our procedure was designed specifically in aged rats and was optimized based on sex in order to keep mortality and extent of injury consistent between aged male and female rats. This model uses a neurological assessment modeled after the NIH Stroke Scale. Finally, the potential for translation between our rat model of stroke and humans was assessed using comparative gene expression for key inflammatory genes. This model will be useful in the evaluation of therapeutic targets to develop adjuvant treatments for large vessel occlusion during the thrombectomy procedure.

Apolipoprotein E genotype-dependent nutrigenetic effects to prebiotic inulin for modulating systemic metabolism and neuroprotection in mice via gut-brain axis.

OBJECTIVE: The goal of the study was to identify the potential nutrigenetic effects to inulin, a prebiotic fiber, in mice with different human apolipoprotein E (APOE) genetic variants. Specifically, we compared responses to inulin for the potential modulation of the systemic metabolism and neuroprotection via gut-brain axis in mice with human APOE ϵ3 and ϵ4 alleles. METHOD: We performed experiments with young mice expressing the human APOE3 (E3FAD mice and APOE4 gene (E4FAD mice). We fed mice with either inulin or control diet for 16 weeks starting from 3 months of age. We determined gut microbiome diversity and composition using16s rRNA sequencing, systemic metabolism using in vivo MRI and metabolomics, and blood-brain barrier (BBB) tight junction expression using Western blot. RESULTS: In both E3FAD and E4FAD mice, inulin altered the alpha and beta diversity of the gut microbiome, increased beneficial taxa of bacteria and elevated cecal short chain fatty acid and hippocampal scyllo-inositol. E3FAD mice had altered metabolism related to tryptophan and tyrosine, while E4FAD mice had changes in the tricarboxylic acid cycle, pentose phosphate pathway, and bile acids. Differences were found in levels of brain metabolites related to oxidative stress, and levels of Claudin-1 and Claudin-5 BBB tight junction expression. DISCUSSION: We found that inulin had many similar beneficial effects in the gut and brain for both E3FAD and E4FAD mice, which may be protective for brain functions and reduce risk for neurodegeneration. . E3FAD and E4FAD mice also had distinct responses in several metabolic pathways, suggesting an APOE-dependent nutrigenetic effects in modulating systemic metabolism and neuroprotection.

Inulin supplementation prior to mild traumatic brain injury mitigates gut dysbiosis, and brain vascular and white matter deficits in mice.

Introduction: Mild traumatic brain injury (mTBI) has been shown to negatively alter bacterial diversity and composition within the gut, known as dysbiosis, in rodents and humans. These changes cause secondary consequences systemically through decreased bacterial metabolites such as short chain fatty acids (SCFAs) which play a role in inflammation and metabolism. The goal of the study was to identify if giving prebiotic inulin prior to closed head injury (CHI) could mitigate gut dysbiosis, increase SCFAs, and improve recovery outcomes, including protecting cerebral blood flow (CBF) and white matter integrity (WMI) in young mice. Methods: We fed mice at 2 months of age with either inulin or control diet (with cellulose as fiber source) for two months before the CHI and continued till the end of the study. We analyzed gut microbiome composition and diversity, determined SCFAs levels, and measured CBF and WMI using MRI. We compared the results with Naïve and Sham-injury mice at 24 hours, 1.5 months, and 3-4 months post-injury. Results: We found that both CHI and Sham mice had time-dependent changes in gut composition and diversity after surgery. Inulin significantly reduced the abundance of pathobiont bacteria, such as E. coli, Desulfovibrio spp and Pseudomonas aeruginosa, in Sham and CHI mice compared to mice fed with control diet. On the other hand, inulin increased SCFAs-producing bacteria, such as Bifidobacterium spp and Lactobacillus spp, increased levels of SCFAs, including butyrate and propionate, and significantly altered beta diversity as early as 24 hours post-injury, which lasted up to 3-4 months post-injury. The mitigation of dysbiosis is associated with protection of WMI in fimbria, internal and external capsule, and CBF in the right hippocampus of CHI mice, suggesting protection of memory and cognitive functions. Discussion: The results indicate that giving inulin prior to CHI could promote recovery outcome through gut microbiome modulation. As inulin, microbiome analysis, and MRI are readily to be used in humans, the findings from the study may pave a way for a cost-effective, accessible intervention for those at risk of sustaining a head injury, such as military personnel or athletes in contact sports.

Gut microbial dysbiosis correlates with stroke severity markers in aged rats.

Background: An imbalanced gut microbial community, or dysbiosis, has been shown to occur following stroke. It is possible that this dysbiosis negatively impacts stroke recovery and rehabilitation. Species level resolution measurements of the gut microbiome following stroke are needed to develop and test precision interventions such as probiotic or fecal microbiota transplant therapies that target the gut microbiome. Previous studies have used 16S rRNA amplicon sequencing in young male mice to obtain broad profiling of the gut microbiome at the genus level following stroke, but further investigations will be needed with whole genome shotgun sequencing in aged rats of both sexes to obtain species level resolution in a model which will better translate to the demographics of human stroke patients. Methods: Thirty-nine aged male and female rats underwent middle cerebral artery occlusion. Fecal samples were collected before stroke and 3 days post stroke to measure gut microbiome. Machine learning was used to identify the top ranked bacteria which were changed following stroke. MRI imaging was used to obtain infarct and edema size and cerebral blood flow (CBF). ELISA was used to obtain inflammatory markers. Results: Dysbiosis was demonstrated by an increase in pathogenic bacteria such as Butyricimonas virosa (15.52 fold change, p < 0.0001), Bacteroides vulgatus (7.36 fold change, p < 0.0001), and Escherichia coli (47.67 fold change, p < 0.0001). These bacteria were positively associated with infarct and edema size and with the inflammatory markers Ccl19, Ccl24, IL17a, IL3, and complement C5; they were negatively correlated with CBF. Conversely, beneficial bacteria such as Ruminococcus flavefaciens (0.14 fold change, p < 0.0001), Akkermansia muciniphila (0.78 fold change, p < 0.0001), and Lactobacillus murinus (0.40 fold change, p < 0.0001) were decreased following stroke and associated with all the previous parameters in the opposite direction of the pathogenic species. There were not significant microbiome differences between the sexes. Conclusion: The species level resolution measurements found here can be used as a foundation to develop and test precision interventions targeting the gut microbiome following stroke. Probiotics that include Ruminococcus flavefaciens, Akkermansia muciniphila, and Lactobacillus murinus should be developed to target the deficit following stroke to measure the impact on stroke severity.

Alzheimer's Disease Classification Using 2D Convolutional Neural Networks.

Alzheimer's disease (AD) is a non-treatable and non-reversible disease that affects about 6% of people who are 65 and older. Brain magnetic resonance imaging (MRI) is a pseudo-3D imaging technology that is widely used for AD diagnosis. Convolutional neural networks with 3D kernels (3D CNNs) are often the default choice for deep learning based MRI analysis. However, 3D CNNs are usually computationally costly and data-hungry. Such disadvantages post a barrier of using modern deep learning techniques in the medical imaging domain, in which the number of data that can be used for training is usually limited. In this work, we propose three approaches that leverage 2D CNNs on 3D MRI data. We test the proposed methods on the Alzheimer's Disease Neuroimaging Initiative dataset across two popular 2D CNN architectures. The evaluation results show that the proposed method improves the model performance on AD diagnosis by 8.33% accuracy or 10.11% auROC compared with the ResNet-based 3D CNN model, while significantly reducing the training time by over 89%. We also discuss the potential causes for performance improvement and the limitations. We believe this work can serve as a strong baseline for future researchers.

Multi-Modal Data Analysis for Alzheimer's Disease Diagnosis: An Ensemble Model Using Imagery and Genetic Features.

Alzheimer's disease (AD) is a devastating neurological disorder primarily affecting the elderly. An estimated 6.2 million Americans age 65 and older are suffering from Alzheimer's dementia today. Brain magnetic resonance imaging (MRI) is widely used for the clinical diagnosis of AD. In the meanwhile, medical researchers have identified 40 risk locus using single-nucleotide polymorphisms (SNPs) information from Genome-wide association study (GWAS) in the past decades. However, existing studies usually treat MRI and GWAS separately. For instance, convolutional neural networks are often trained using MRI for AD diagnosis. GWAS and SNPs are frequently used to identify genomic traits. In this study, we propose a multi-modal AD diagnosis neural network that uses both MRIs and SNPs. The proposed method demonstrates a novel way to use GWAS findings by directly including SNPs in predictive models. We test the proposed methods on the Alzheimer's Disease Neuroimaging Initiative dataset. The evaluation results show that the proposed method improves the model performance on AD diagnosis and achieves 93.5% AUC and 96.1% AP, respectively, when patients have both MRI and SNP data. We believe this work brings exciting new insights to GWAS applications and sheds light on future research directions.