Biology subjects

Dehkharghanian, T.

Publications and source records attributed to Dehkharghanian, T..

4 recordsLinked to original sources

bioRxiv2023

The Spatial Patterns and Determinants of Cerebrospinal Fluid Circulation in the Human Brain

The circulation of cerebrospinal fluid (CSF) is essential for maintaining brain homeostasis and clearance, and impairments in its flow can lead to various brain disorders. Recent studies have shown that CSF circulation can be interrogated using low b-value diffusion magnetic resonance imaging (low-b dMRI). Nevertheless, the spatial organization of intracranial CSF flow dynamics remains largely elusive. Here, we developed a whole-brain voxel-based analysis framework, termed CSF pseudo-diffusion spatial statistics (C{Psi}SS), to examine CSF mean pseudo-diffusivity (M{Psi}), a measure of CSF flow magnitude derived from low-b dMRI. We showed that intracranial CSF M{Psi} demonstrates characteristic covariance patterns by employing seed-based correlation analysis. Importantly, we applied non-negative matrix factorization analysis to further elucidate the covariance patterns of CSF M{Psi} in a hypothesis-free, data-driven way. We identified distinct CSF spaces that consistently displayed unique pseudo-diffusion characteristics across multiple imaging datasets. Our study revealed that age, sex, brain atrophy, ventricular anatomy, and cerebral perfusion differentially influence M{Psi} across these CSF spaces. Notably, individuals with anomalous CSF flow patterns displayed incidental findings on multimodal neuroradiological examinations. Our work sets forth a new paradigm to study CSF flow, with potential applications in clinical settings.

neuroscience↗

bioRxiv2022

Whole slide image representation in bone marrow cytology

One of the goals of AI-based computational pathology is to generate compact WSI representations, identifying the essential information required for diagnosis. While such approaches have been applied to histopathology, few applications have been reported in cytology. Bone marrow aspirate cytology is the basis for key clinical decisions in hematology. However, visual inspection of aspirate specimens is a tedious and complex process subject to variation in interpretation, and hematopathology expertise is scarce. The ability to generate a compact representation of an aspirate specimen may form the basis for clinical decision support tools in hematology. We have previously published an end-to-end AI-based system for counting and classifying cells from bone marrow aspirate WSI. Using deep embeddings from this model, we construct bags of individual cell features from each WSI, and apply multiple instance learning to extract vector representations for each WSI. Using these representations in vector search, we achieved 0.58 {+/-} 0.02 mAP@10 in WSI-level image retrieval, which outperforms the Random baseline (0.39 {+/-} 0.1). Using a weighted k-nearest-neighbours (k-NN) model on these slide vectors, we predict five broad diagnostic labels on individual aspirate WSI with a weighted-macro-average F1 score of 0.57 {+/-} 0.03 on the test set of 278 randomly sampled WSIs, which outperforms a classifier using empirical class prior probabilities (0.26 {+/-} 0.02). We present the first example of exploring trainable mechanisms to generate compact, slide-level representations in bone marrow cytology with deep learning. This method has the potential to summarize complex semantic information in WSIs toward improved diagnostics in hematology, and may eventually support AI-assisted computational pathology approaches.

bioinformatics↗

bioRxiv2022

Cell projection plots: a novel visualization of bone marrow aspirate cytology

Deep models for cell detection have demonstrated utility in bone marrow cytology, showing impressive results in terms of accuracy and computational efficiency. However, these models have yet to be implemented in the clinical diagnostic workflow. Additionally, the metrics used to evaluate cell detection models are not necessarily aligned with clinical goals and targets. In order to address these issues, we introduce cell projection plots (CPPs), which are novel, automatically generated visual summaries of bone marrow aspirate specimens. CPPs provide a compact summary of bone marrow aspirate cytology, and encompass relevant biological patterns such as neutrophil maturation. To gauge clinical relevance, CPPs were shown to three hematopathologists, who decided whether shown diagnostic synopses matched with generated CPPs. Pathologists were able to match CPPs to the correct synopsis with 85% accuracy. Our finding suggests CPPs can compactly represent clinically relevant information from bone marrow aspirate specimens, and may be used to efficiently summarize bone marrow cytology to pathologists. CPP could be a step toward human-centered implementation of artificial intelligence (AI) in hematopathology, and a basis for a diagnostic support tool for digital pathology workflows.

pathology↗

bioRxiv2021

Semi-automated analysis of an optical ATP indicator in neural synapses

SignificanceThe firefly enzyme luciferase has been used in a wide range of biological assays, including bioluminescence imaging of ATP. The biosensor Syn-ATP utilizes subcellular targeting of luciferase to nerve terminals for optical measurement of ATP in this compartment. Manual analysis of Syn-ATP signals is challenging due to signal heterogeneity and cellular motion in long imaging sessions. Here, we have leveraged machine learning tools to develop a method for analysis of bioluminescence images. AimOur goal was to create a semi-automated pipeline for analysis of bioluminescence imaging to improve measurements of ATP content in nerve terminals. ApproachWe developed an image analysis pipeline that applies machine learning toolkits to distinguish neurons from background signals, and excludes neural cell bodies, while also incorporating user input. ResultsSide-by-side comparison of manual and semi-automated image analysis demonstrated that the latter improves precision and accuracy of ATP measurements. ConclusionsOur method streamlines data analysis and reduces user-introduced bias, thus enhancing the reproducibility and reliability of quantitative ATP imaging in nerve terminals.

neuroscience↗