Collaborating Authors


'AI algorithm can predict Alzheimer's disease in 1 minute'


A study by Vuno, a Korean artificial intelligence (AI) developer, showed that a deep learning algorithm could predict Alzheimer's disease (AD) within one minute. Jointly with Asan Medical Center, Vuno verified an AI algorithm using MRI scans of 2,727 patients registered at domestic medical institutions. Vuno found that the algorithm predicted AD and mild cognitive impairment (MCI) accurately. Vuno's deep learning-based algorithm used an area under the curve (AUC) to predict dementia. The closer the AUC value is, the higher the algorithm's performance is.

EEG Plus Machine-learning Identifies Mild Cognitive Impairment...


Using non-invasive, dry-electroencephalography (EEG) and machine-learning computer algorithms, researchers were able to distinguish between Parkinson's disease patients with and without mild cognitive impairments with 80% accuracy, early results of a study found. Follow-up assessments will be conducted after 12 months to validate the method's ability to predict cognitive impairment in this patient population. The results were presented in a poster, "The Identification of Mild Cognitive Impairment in Parkinson's disease using EEG and Machine Learning, "at the Alzheimer's Association International Conference 2020. The poster abstract was published in the journal Alzheimer's & Dementia. Cognitive impairment is a common symptom of Parkinson's disease characterized by difficulties in executive function, attention, vision, word-finding, and problems with learning and remembering information.

How artificial intelligence could ameliorate the diagnosis of patients with Alzheimer's


A recent study released in Nature Reviews: Neurology found that Alzheimer's disease could be diagnosed faster and efficiently using artificial intelligence (AI). The study, conducted by the University of Sheffield, looks at the use of AI technologies, like machine learning, in healthcare to reduce the workflow and economic effects of traditional methods for detecting neurodegenerative diseases. In their study, the use of machine learning in assessing cognitive function was initiated in conjunction with biotech company BenevolentAI. The Sheffield team, along with BenevolentAI, demonstrated in their findings how machine learning algorithms could be efficient for the detection of the brain regions implicated before the onset of rapid cognitive decline or development of Alzheimer's. "Widespread implementation of AI technologies can help, for example, predict which patients showing mild cognitive impairment will go on to develop Alzheimer's disease, or how severely their motor skills will decline over time," said Laura Ferraiuolo, the study's lead author.

Achievements and Challenges in Explaining Deep Learning based Computer-Aided Diagnosis Systems Artificial Intelligence

Remarkable success of modern image-based AI methods and the resulting interest in their applications in critical decision-making processes has led to a surge in efforts to make such intelligent systems transparent and explainable. The need for explainable AI does not stem only from ethical and moral grounds but also from stricter legislation around the world mandating clear and justifiable explanations of any decision taken or assisted by AI. Especially in the medical context where Computer-Aided Diagnosis can have a direct influence on the treatment and well-being of patients, transparency is of utmost importance for safe transition from lab research to real world clinical practice. This paper provides a comprehensive overview of current state-of-the-art in explaining and interpreting Deep Learning based algorithms in applications of medical research and diagnosis of diseases. We discuss early achievements in development of explainable AI for validation of known disease criteria, exploration of new potential biomarkers, as well as methods for the subsequent correction of AI models. Various explanation methods like visual, textual, post-hoc, ante-hoc, local and global have been thoroughly and critically analyzed. Subsequently, we also highlight some of the remaining challenges that stand in the way of practical applications of AI as a clinical decision support tool and provide recommendations for the direction of future research.

AI detects protein signatures for Alzeihmer's disease in the blood - Advanced Science News


Nanoparticles could make a reliable blood test for Alzheimer's disease a reality; image credit: National Cancer Institute, Daniel Sone Using nanoparticles with different surface properties, researchers are able to detect subtle changes in the composition of proteins in the plasma years before the presentation of clinical symptoms of Alzheimer's disease, which include memory loss, confusion, and cognitive difficulties. Owing to the unique properties of nanoparticles, different proteins in biological fluids selectively stick onto their surface forming a protein corona, which was found to change during disease. Researchers from the United States and Italy identify these subtle changes in plasma protein patterns to distinguish plasma samples from healthy individuals and those diagnosed with Alzheimer's disease. "Protein corona composition is both influenced by specific health conditions as well as the chemical and physical properties of the nanoparticles themselves," says Dr. Claudia Corbo of the University of Milano-Bicocca and lead author of the study published in Advanced Healthcare Materials. "Binding of proteins to the surface of particles is very precise and dependent on the chemistry and shape of the particles and the chemistry and structure of the proteins," says senior author Professor Omid Farokhzad of Brigham and Women's Hospital and Harvard Medical School.

Recalibration of Neural Networks for Point Cloud Analysis Artificial Intelligence

Spatial and channel re-calibration have become powerful concepts in computer vision. Their ability to capture long-range dependencies is especially useful for those networks that extract local features, such as CNNs. While re-calibration has been widely studied for image analysis, it has not yet been used on shape representations. In this work, we introduce re-calibration modules on deep neural networks for 3D point clouds. We propose a set of re-calibration blocks that extend Squeeze and Excitation blocks and that can be added to any network for 3D point cloud analysis that builds a global descriptor by hierarchically combining features from multiple local neighborhoods. We run two sets of experiments to validate our approach. First, we demonstrate the benefit and versatility of our proposed modules by incorporating them into three state-of-the-art networks for 3D point cloud analysis: PointNet++, DGCNN, and RSCNN. We evaluate each network on two tasks: object classification on ModelNet40, and object part segmentation on ShapeNet. Our results show an improvement of up to 1% in accuracy for ModelNet40 compared to the baseline method. In the second set of experiments, we investigate the benefits of re-calibration blocks on Alzheimer's Disease (AD) diagnosis. Our results demonstrate that our proposed methods yield a 2% increase in accuracy for diagnosing AD and a 2.3% increase in concordance index for predicting AD onset with time-to-event analysis. Concluding, re-calibration improves the accuracy of point cloud architectures, while only minimally increasing the number of parameters.

Autosomal dominant VCP hypomorph mutation impairs disaggregation of PHF-tau


Neurodegeneration in Alzheimer's disease dementia is associated with neurofibrillary tangles composed of aggregated tau protein. Darwich et al. describe an additional form of autosomal-dominant dementia with neurofibrillary tangles linked to a hypomorph mutation in valosin-containing protein (VCP). VCP was found to disaggregate pathologic tau, and the hypomorph mutation increased tau accumulation in cells and mice. These findings highlight the role of protein turnover in maintaining neuronal health and suggest that VCP may provide a therapeutic target for Alzheimer's disease. Science , this issue p. [eaay8826][1] ### INTRODUCTION Alzheimer’s disease (AD) is a fatal neurodegenerative disease in which progressive brain degeneration compromises cognitive function. AD neurodegeneration is tightly associated with abnormal neuronal inclusions called neurofibrillary tangles, which are composed of aggregated tau protein. The importance of tau protein in dementia is highlighted by various known autosomal-dominant mutations in MAPT (microtubule-associated protein tau), the gene that encodes for tau, that are associated with frontotemporal lobar degeneration with tau inclusions (FTLD-tau). Identifying additional disease-causing genes that affect tau accumulation, including genes involved in protein quality control and tau clearance, has the potential to reveal previously unknown mechanisms that maintain neuronal health. ### RATIONALE Two families were identified with autosomal-dominant dementia linked to a p.Asp395Gly mutation in VCP . Valosin-containing protein (VCP) is a AAA+ [adenosine triphosphatases (ATPases) associated with diverse cellular activities] protein and uses energy from adenosine 5′-triphosphate (ATP) hydrolysis to unfold substrates to assist the dismantling of macromolecular complexes. Other VCP mutations have been identified in a disease called multisystem proteinopathy (MSP), which is associated with neuronal TDP-43 [TAR DNA-binding protein 43] protein inclusions. The mechanisms by which p.Asp395Gly VCP leads to neurodegeneration are unknown. ### RESULTS Two kindred were identified with an autosomal-dominant inheritance pattern of frontotemporal degeneration linked to a p.Asp395Gly VCP mutation. We have named this disease vacuolar tauopathy because of the presence of neuronal vacuoles and tau aggregates. Tau aggregates were morphologically and biochemically similar to AD neurofibrillary tangles. Moreover, the presence of vacuoles and neurofibrillary tangles in the brain were inversely correlated. Degenerating brain regions such as the frontal neocortex exhibited tau aggregation, whereas nondegenerating brain regions such as the visual cortex exhibited vacuolization. To further characterize the p.Asp395Gly VCP mutation, we assessed recombinant VCP proteins for ATPase activity in an in vitro assay. This approach demonstrated that p.Asp395Gly VCP exhibited a partial loss of ATPase activity, in contrast with MSP mutations, which increase ATPase activity. Given that VCP unfolds protein substrates, we hypothesized that VCP may disaggregate pathologic tau aggregates. Indeed, VCP appeared to partially disaggregate pathologic tau aggregates derived from AD human brain tissues, and the p.Asp395Gly mutation impaired this activity. VCP activity against pathologic tau was energy (ATP) dependent and required polyubiquitination of the tau substrate. In addition, expression of p.Asp395Gly VCP in a cell culture model of tau aggregation was associated with enhanced accumulation of cellular tau aggregates. Last, we generated mice in which the p.Asp395Gly mutation was knocked in, which exhibited a minimal phenotype when unchallenged. However, upon initiating tau aggregation through microinjection of pathologic AD tau extracts into the mouse brain, mutant VCP mice showed an increase in tau accumulation compared with that of wild-type animals. ### CONCLUSION We describe a partial loss-of-function mutation in VCP that was associated with a neurodegenerative disease, which we named vacuolar tauopathy. VCP appeared to exhibit activity that promoted the disruption of tau aggregates in an energy- and polyubiquitin-dependent manner. Furthermore, the p.Asp395Gly VCP mutation enhanced tau aggregation in a cell culture model system and in a knock-in mouse model. These results highlight a potential role for protein disaggregation in the maintenance of neuronal health. Furthermore, our findings suggest that VCP may provide a potential therapeutic target in the development of future AD therapies. ![Figure][2] Vacuolar tauopathy and VCP function. Model of tau aggregation, highlighting different mechanisms that lead to neurodegeneration in AD and FTLD-tau. Tau is a soluble protein (green) that forms highly structured fibrils (purple) in AD. Tau aggregates are also a hallmark of FTLD-tau. Autosomal dominant FTLD-tau mutations in MAPT , the gene that encodes tau protein, can alter splicing or enhance tau protein aggregation, which results in the accumulation of insoluble tau inclusions in neurons and glia. VCP (blue) appears to exhibit disaggregase activity against pathologic tau. This activity is dependent on the availability of energy (ATP) and the presence of polyubiquitin on tau protein aggregates. Vacuolar tauopathy is an autosomal-dominant form of dementia linked to a p.Asp395Gly VCP mutation (orange-red). Neurodegeneration in vacuolar tauopathy was associated with tau aggregates of similar biochemical composition and morphology as those seen in Alzheimer’s disease. The accumulation of insoluble tau aggregates in vacuolar tauopathy appeared to be in part due to a partial loss of tau disaggregase function associated with the p.Asp395Gly VCP mutation. WT, wild-type. Neurodegeneration in Alzheimer’s disease (AD) is closely associated with the accumulation of pathologic tau aggregates in the form of neurofibrillary tangles. We found that a p.Asp395Gly mutation in VCP (valosin-containing protein) was associated with dementia characterized neuropathologically by neuronal vacuoles and neurofibrillary tangles. Moreover, VCP appeared to exhibit tau disaggregase activity in vitro, which was impaired by the p.Asp395Gly mutation. Additionally, intracerebral microinjection of pathologic tau led to increased tau aggregates in mice in which p.Asp395Gly VCP mice was knocked in, as compared with injected wild-type mice. These findings suggest that p.Asp395Gly VCP is an autosomal-dominant genetic mutation associated with neurofibrillary degeneration in part owing to reduced tau disaggregation, raising the possibility that VCP may represent a therapeutic target for the treatment of AD. [1]: /lookup/doi/10.1126/science.aay8826 [2]: pending:yes

Image analysis for Alzheimer's disease prediction: Embracing pathological hallmarks for model architecture design Machine Learning

Alzheimer's disease (AD) is associated with local (e.g. brain tissue atrophy) and global brain changes (loss of cerebral connectivity), which can be detected by high-resolution structural magnetic resonance imaging. Conventionally, these changes and their relation to AD are investigated independently. Here, we introduce a novel, highly-scalable approach that simultaneously captures $\textit{local}$ and $\textit{global}$ changes in the diseased brain. It is based on a neural network architecture that combines patch-based, high-resolution 3D-CNNs with global topological features, evaluating multi-scale brain tissue connectivity. Our local-global approach reached competitive results with an average precision score of $0.95\pm0.03$ for the classification of cognitively normal subjects and AD patients (prevalence $\approx 55\%$).

A new AI program can listen to you cough and discern whether you have the coronavirus. Researchers hope to turn it into an app.


At least one out of every five people who get the coronavirus doesn't show symptoms and can unknowingly spread the virus to others. Those who don't feel sick and aren't notified of exposure can't know that they should get tested. But researchers at the Massachusetts Institute of Technology may have found a way to identify these silent coronavirus carriers without a test. A study published in September describes an artificial-intelligence model that can distinguish between the coughs of people with the coronavirus and those who are healthy. It can even tell from voluntary, forced coughs whether people were healthy or were asymptomatic carriers, based on sound variations too subtle for the human ear to discern.

How a machine learning algorithm could identify the early stages of Alzheimer's in patients - Mental Daily


A study published in the Journal of Medical Imaging unveils the use of machine learning to detect the early stages of Alzheimer's disease (AD) by functional magnetic resonance imaging. Alzheimer's disease is a neurodegenerative condition primarily occurring in late-adulthood and begins with symptoms of cognitive decline. Researchers from Texas Tech University developed a deep-learning algorithm called a convolutional neural network able to distinguish between the fMRI signals of healthy individuals, patients with mild cognitive impairment (MCI), and patients with Alzheimer's. "We present one such synergy of fMRI and deep learning, where we apply a simplified yet accurate method using a modified 3D convolutional neural networks to resting-state fMRI data for feature extraction and classification of Alzheimer's disease," the co-authors explained in their findings. "The convolutional neural network is designed in such a way that it uses the fMRI data with much less preprocessing, preserving both spatial and temporal information."