Dataset Information

Hypothesis Testing in Unsupervised Domain Adaptation with Applications in Alzheimer's Disease.

ABSTRACT: Consider samples from two different data sources [Formula: see text] and [Formula: see text]. We only observe their transformed versions [Formula: see text] and [Formula: see text], for some known function class h(·) and g(·). Our goal is to perform a statistical test checking if Psource = Ptarget while removing the distortions induced by the transformations. This problem is closely related to domain adaptation, and in our case, is motivated by the need to combine clinical and imaging based biomarkers from multiple sites and/or batches - a fairly common impediment in conducting analyses with much larger sample sizes. We address this problem using ideas from hypothesis testing on the transformed measurements, wherein the distortions need to be estimated in tandem with the testing. We derive a simple algorithm and study its convergence and consistency properties in detail, and provide lower-bound strategies based on recent work in continuous optimization. On a dataset of individuals at risk for Alzheimer's disease, our framework is competitive with alternative procedures that are twice as expensive and in some cases operationally infeasible to implement.

SUBMITTER: Zhou HH

PROVIDER: S-EPMC5754041 | biostudies-literature | 2016

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Hypothesis Testing in Unsupervised Domain Adaptation with Applications in Alzheimer's Disease.

Zhou Hao Henry HH Ravi Sathya N SN Ithapu Vamsi K VK Johnson Sterling C SC Wahba Grace G Singh Vikas V

Advances in neural information processing systems 20160101

Consider samples from two different data sources [Formula: see text] and [Formula: see text]. We only observe their transformed versions [Formula: see text] and [Formula: see text], for some known function class h(·) and g(·). Our goal is to perform a statistical test checking if Psource = Ptarget while removing the distortions induced by the transformations. This problem is closely related to domain adaptation, and in our case, is motivated by t ...[more]

PMID: 29308004

Similar Datasets

Project description:BackgroundCell counting from cell cultures is required in multiple biological and biomedical research applications. Especially, accurate brightfield-based cell counting methods are needed for cell growth analysis. With deep learning, cells can be detected with high accuracy, but manually annotated training data is required. We propose a method for cell detection that requires annotated training data for one cell line only, and generalizes to other, unseen cell lines.ResultsTraining a deep learning model with one cell line only can provide accurate detections for similar unseen cell lines (domains). However, if the new domain is very dissimilar from training domain, high precision but lower recall is achieved. Generalization capabilities of the model can be improved with training data transformations, but only to a certain degree. To further improve the detection accuracy of unseen domains, we propose iterative unsupervised domain adaptation method. Predictions of unseen cell lines with high precision enable automatic generation of training data, which is used to train the model together with parts of the previously used annotated training data. We used U-Net-based model, and three consecutive focal planes from brightfield image z-stacks. We trained the model initially with PC-3 cell line, and used LNCaP, BT-474 and 22Rv1 cell lines as target domains for domain adaptation. Highest improvement in accuracy was achieved for 22Rv1 cells. F1-score after supervised training was only 0.65, but after unsupervised domain adaptation we achieved a score of 0.84. Mean accuracy for target domains was 0.87, with mean improvement of 16 percent.ConclusionsWith our method for generalized cell detection, we can train a model that accurately detects different cell lines from brightfield images. A new cell line can be introduced to the model without a single manual annotation, and after iterative domain adaptation the model is ready to detect these cells with high accuracy.

Project description:Alzheimer's disease (AD) is the most common form of dementia in the elderly. Hallmark AD neuropathology includes extracellular amyloid plaques composed largely of the amyloid-? protein (A?), intracellular neurofibrillary tangles (NFTs) composed of hyper-phosphorylated microtubule-associated protein tau (MAP-tau), and microtubule destabilization. Early-onset autosomal dominant AD genes are associated with excessive A? accumulation, however cognitive impairment best correlates with NFTs and disrupted microtubules. The mechanisms linking A? and NFT pathologies in AD are unknown. Here, we propose that sequestration of zinc by A?-amyloid deposits (A? oligomers and plaques) not only drives A? aggregation, but also disrupts zinc homeostasis in zinc-enriched brain regions important for memory and vulnerable to AD pathology, resulting in intra-neuronal zinc levels, which are either too low, or excessively high. To evaluate this hypothesis, we 1) used molecular modeling of zinc binding to the microtubule component protein tubulin, identifying specific, high-affinity zinc binding sites that influence side-to-side tubulin interaction, the sensitive link in microtubule polymerization and stability. We also 2) performed kinetic modeling showing zinc distribution in extra-neuronal A? deposits can reduce intra-neuronal zinc binding to microtubules, destabilizing microtubules. Finally, we 3) used metallomic imaging mass spectrometry (MIMS) to show anatomically-localized and age-dependent zinc dyshomeostasis in specific brain regions of Tg2576 transgenic, mice, a model for AD. We found excess zinc in brain regions associated with memory processing and NFT pathology. Overall, we present a theoretical framework and support for a new theory of AD linking extra-neuronal A? amyloid to intra-neuronal NFTs and cognitive dysfunction. The connection, we propose, is based on ?-amyloid-induced alterations in zinc ion concentration inside neurons affecting stability of polymerized microtubules, their binding to MAP-tau, and molecular dynamics involved in cognition. Further, our theory supports novel AD therapeutic strategies targeting intra-neuronal zinc homeostasis and microtubule dynamics to prevent neurodegeneration and cognitive decline.

Dataset Information

Hypothesis Testing in Unsupervised Domain Adaptation with Applications in Alzheimer's Disease.

Publications

Hypothesis Testing in Unsupervised Domain Adaptation with Applications in Alzheimer's Disease.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets