Biology subjects

Salama, P.

Publications and source records attributed to Salama, P..

9 recordsLinked to original sources

bioRxiv2026

Multidomain Analysis of Clinical Cognitive Assessments and Imaging Data in Alzheimer's Disease Accurately Predicts Disease Stage and Grade Independent of Amyloid and Tau

BackgroundIndividual clinical cognitive assessments (CCA) for Alzheimers disease (AD) provide broad disease stratification but are limited in sensitivity and specificity, requiring integration of multiple CCA for optimal disease staging. Recent work from our lab suggests that neuro-metabolic and vascular dysregulation (MVD) occurs early in AD, prior to clinical symptoms, and may provide higher sensitivity and specificity than CCA alone. In this study, we combined three widely accepted CCA with MVD readouts and developed a multimodal ensemble machine learning approach across the AD spectrum to predict disease stage and grade. MethodsAD subjects (N=372) across the disease spectrum with imaging (PET:18F-FDG, MRI:T1w, T2 FLAIR, ASL) and CCAs (ADAS-Cog, CDR, MoCA) data were analyzed from ADNI. Imaging data were registered to MNI152+, z-scored relative to cognitively normal controls, and processed for MVD. A clinical-set-enrichment analysis (CSEA) was developed to link regional brain changes with CCA scores, map changes to functional categories, project them into a 3D Cartesian space, and model trajectories, thus revealing at-risk and resilient regions. In addition, an ensemble machine-learning approach was utilized for disease stage classification, and a disease grading scheme across the AD spectrum was developed to further stratify within disease stages. FindingsRegional data followed an MVD pattern across AD stages stratified by CSEA scores. Females showed greater stage separation along the CCA axis within each region, indicating faster disease progression. Moreover, progression in at-risk brain regions (e.g., mid- and inf-temporal gyri, amygdala) was associated with longer disease path lengths, whereas progression in resilient brain regions (supramarginal gyrus) was not. Moreover, our classification and grading approach can predict AD stage and grade independent of amyloid-beta and tau with high precision and accuracy. InterpretationA framework was developed to evaluate MVD and CCA variations across the AD spectrum, thereby distinguishing at-risk and resilient brain regions. Distinct disease trajectories were identified, and a new data-driven grading scheme was proposed to highlight the potential for precision medicine and therapeutic evaluation. FundingNIH T32AG071444

neuroscience↗

bioRxiv2024

A Quantitative Metric of Confidence For Segmentation of Nuclei in Large Spatially Variable Image Volumes

Nuclei segmentation is an important step for quantitative analysis of fluorescence microscopy images. A large volume generally has many different regions containing nuclei with varying spatial characteristics. Automatically identifying nuclei that are challenging to segment can speed up the analysis of biological tissues. Here we show a segmentation technique that provides a metric of segmentation "confidence" for each segmented object in an image volume. This confidence metric can be used either to generate a "confidence map" for visual distinction of reliable from unreliable regions, or in the data space to identify questionable measurements that can be analyzed separately or eliminated from analysis. In an analysis of nuclei in a 3-dimensional image volume, we show that the confidence map correlates well with visual evaluations of segmentation quality, and that the confidence metric correlates well with F1 scores within subregions of the image volume. In addition, we also describe three visualization methods that can visualize the segmentation differences between a segmented volume and a reference volume.

cell biology↗

bioRxiv2022

3D Ground Truth Annotations of Nuclei in 3D Microscopy Volumes

In this paper we describe a set of 3D microscopy volumes we have partially manually annotated. We describe the volumes annotated and the tools and processes we use to annotate the volumes. In addition, we provide examples of annotated subvolumes. We also provide synthetically generated 3D microscopy volumes that can be used for training segmentation methods. The full set of annotations, synthetically generated volumes, and original volumes can be accessed as described in the paper.

cell biology↗

bioRxiv2022

3D CentroidNet: Nuclei Centroid Detection With Vector Flow Voting

Automated microscope systems are increasingly used to collect large-scale 3D image volumes of biological tissues. Since cell boundaries are seldom delineated in these images, detection of nuclei is a critical step for identifying and analyzing individual cells. Due to the large intra-class variability in nuclei morphology and the difficulty of generating ground truth annotations, accurate nuclei detection remains a challenging task. We propose a 3D nuclei centroid detection method by estimating the "vector flow" volume where each voxel represents a 3D vector pointing to its nearest nuclei centroid in the corresponding microscopy volume. We then use a voting mechanism to estimate the 3D nuclei centroids from the "vector flow" volume. Our system is trained on synthetic microscopy volumes and tested on real microscopy volumes. The evaluation results indicate our method outperforms other methods both visually and quantitatively.

cell biology↗

bioRxiv2022

NISNet3D: Three-Dimensional Nuclear Synthesis and Instance Segmentation for Fluorescence Microscopy Images

The primary step in tissue cytometry is the automated distinction of individual cells (segmentation). Since cell borders are seldom labeled, researchers generally segment cells by their nuclei. While effective tools have been developed for segmenting nuclei in two dimensions, segmentation of nuclei in three-dimensional volumes remains a challenging task for which few tools have been developed. The lack of effective methods for three-dimensional segmentation represents a bottleneck in the realization of the potential of tissue cytometry, particularly as methods of tissue clearing present researchers with the opportunity to characterize entire organs. Methods based upon deep-learning have shown enormous promise, but their implementation is hampered by the need for large amounts of manually annotated training data. In this paper we describe 3D Nuclei Instance Segmentation Network (NISNet3D), a deep learning-based approach in which training is accomplished using synthetic data, profoundly reducing the effort required for network training. We compare results obtained from NISNet3D with results obtained from eight existing techniques.

cell biology↗

bioRxiv2022

DINAVID: A Distributed and Networked Image Analysis System for Volumetric Image Data

BackgroundThe advancement of high content optical microscopy has enabled the acquisition of very large 3D image datasets. Image analysis tools and three dimensional visualization are critical for analyzing and interpreting 3D image volumes. The analysis of these volumes require more computational resources than a biologist may have access to in typical desktop or laptop computers. This is especially true if machine learning tools are being used for image analysis. With the increased amount of data analysis and computational complexity, there is a need for a more accessible, easy-to-use, and efficient network-based/cloud-based 3D image processing system. ResultsThe Distributed and Networked Analysis of Volumetric Image Data (DINAVID) system was developed to enable remote analysis of 3D microscopy images for biologists. DINAVID is a server/cloud-based system with a simple web interface that allows biologists to upload 3D volumes for analysis and visualization. DINAVID is designed using open source tools and has two main sub-systems, a computational system for 3D microscopy image processing and analysis as well as a 3D visualization system. ConclusionsIn this paper, we will present an overview of the DINAVID system and compare it to other tools currently available for microscopy image analysis.

cell biology↗

bioRxiv2022

Deep multi-omic network fusion for marker discovery of Alzheimer's Disease

Multi-omic data spanning from genotype, gene expression to protein expression have been increasingly explored to interpret findings from genome wide association studies of Alzheimers disease (AD) and to gain more insight of the disease mechanism. However, each -omics data type is usually examined individually and the functional interactions between genetic variations, genes and proteins are only used after discovery to interpret the findings, but not beforehand. In this case, multi-omic findings are likely not functionally related and therefore give rise to challenges in interpretation. To address this problem, we propose a new interpretable deep neural network model MoFNet to jointly model the prior knowledge of functional interactions and multi-omic data set. It aims to identify a subnetwork of functional interactions predictive of AD evidenced by multi-omic measures. Particularly, prior functional interaction network was embedded into the architecture of MoFNet in a way that it resembles the information flow from DNA to gene and protein. The proposed model MoFNet significantly outperformed all other state-of-art classifiers when evaluated using multi-omic data from the ROS/MAP cohort. Instead of individual markers, MoFNet yielded multi-omic sub-networks related to innate immune system, clearance of misfolded proteins, and neurotransmitter release respectively. Around 50% of these findings were replicated in another independent cohort. Our identified gene/proteins are highly related to synaptic vesicle function. Altered regulation or expression of these genes/proteins could cause disruption in neuron-neuron or neuron-glia cross talk and further lead to neuronal and synapse loss in AD. Further investigation of these identified genes/proteins could possibly help decipher the mechanisms underlying synaptic dysfunction in AD, and ultimately inform therapeutic strategies to modify AD progression in the early stage.

systems biology↗

bioRxiv2022

An Ensemble Learning and Slice Fusion Strategy for Three-Dimensional Nuclei Instance Segmentation

Automated microscopy image analysis is a fundamental step for digital pathology and computer aided diagnosis. Most existing deep learning methods typically require post-processing to achieve instance segmentation and are computationally expensive when directly used with 3D microscopy volumes. Supervised learning methods generally need large amounts of ground truth annotations for training whereas manually annotating ground truth masks is laborious especially for a 3D volume. To address these issues, we propose an ensemble learning and slice fusion strategy for 3D nuclei instance segmentation that we call Ensemble Mask R-CNN (EMR-CNN) which uses different object detectors to generate nuclei segmentation masks for each 2D slice of a volume and propose a 2D ensemble fusion and a 2D to 3D slice fusion to merge these 2D segmentation masks into a 3D segmentation mask. Our method does not need any ground truth annotations for training and can inference on any large size volumes. Our proposed method was tested on a variety of microscopy volumes collected from multiple regions of organ tissues. The execution time and robustness analyses show that our method is practical and effective.

cell biology↗

bioRxiv2019

TSUNAMI: Translational Bioinformatics Tool Suite For Network Analysis And Mining

Gene co-expression network (GCN) mining identifies gene modules with highly correlated expression profiles across samples/conditions. It helps to discover latent gene/molecular interactions, identify novel gene functions, and extract molecular features from certain disease/condition groups, thus help to identify disease biomarkers. However, there lacks an easy-to-use tool package for users to mine GCN modules that are relatively small in size with tightly connected genes that can be convenient for downstream Gene Ontology (GO) enrichment analysis, as well as modules that may share common members. To address this need, we develop a GCN mining tool package TSUNAMI (Tools SUite for Network Analysis and MIning) which incorporates our state-of-the-art lmQCM algorithm to mine GCN modules in public and user-input data (microarray, RNA-seq, or any other numerical omics data), then performs downstream GO and enrichment analysis based on the modules identified. It has several features and advantages: (i) user friendly interface and the real-time co-expression network mining through web server; (ii) direct access and search of GEO and TCGA databases as well as user-input expression matrix (microarray, RNA-seq, etc.) for GCN module mining; (iii) multiple co-expression analysis tools to choose with highly flexible of parameter selection options; (iv) identified GCN modules are summarized to eigengenes, which are convenient for user to check their correlation with other clinical traits; (v) integrated downstream Enrichr enrichment analysis and links to other GO tools; (vi) visualization of gene loci by Circos plot in any step. The web service is freely accessible through URL: http://spore.ph.iu.edu:3838/zhihuan/TSUNAMI/. Source code is available at https://github.com/huangzhii/TSUNAMI/.

bioinformatics↗