IU Indianapolis ScholarWorks :: Browsing by Author "Dunn, Kenneth W."

Browsing by Author "Dunn, Kenneth W."

Now showing 1 - 10 of 61

3D Centroidnet: Nuclei Centroid Detection with Vector Flow Voting
(IEEE, 2022-10) Wu, Liming; Chen, Alain; Salama, Paul; Dunn, Kenneth W.; Delp, Edward J.; Electrical and Computer Engineering, School of Engineering and Technology
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.
A multimodal and integrated approach to interrogate human kidney biopsies with rigor and reproducibility: guidelines from the Kidney Precision Medicine Project
(American Physiological Society, 2021) El-Achkar, Tarek M.; Eadon, Michael T.; Menon, Rajasree; Lake, Blue B.; Sigdel, Tara K.; Alexandrov, Theodore; Parikh, Samir; Zhang, Guanshi; Dobi, Dejan; Dunn, Kenneth W.; Otto, Edgar A.; Anderton, Christopher R.; Carson, Jonas M.; Luo, Jinghui; Park, Chris; Hamidi, Habib; Zhou, Jian; Hoover, Paul; Schroeder, Andrew; Joanes, Marianinha; Azeloglu, Evren U.; Sealfon, Rachel; Winfree, Seth; Steck, Becky; He, Yongqun; D’Agati, Vivette; Iyengar, Ravi; Troyanskaya, Olga G.; Barisoni, Laura; Gaut, Joseph; Zhang, Kun; Laszik, Zoltan; Rovin, Brad H.; Dagher, Pierre C.; Sharma, Kumar; Sarwal, Minnie M.; Hodgin, Jeffrey B.; Alpers, Charles E.; Kretzler, Matthias; Jain, Sanjay; Medicine, School of Medicine
Comprehensive and spatially mapped molecular atlases of organs at a cellular level are a critical resource to gain insights into pathogenic mechanisms and personalized therapies for diseases. The Kidney Precision Medicine Project (KPMP) is an endeavor to generate three-dimensional (3-D) molecular atlases of healthy and diseased kidney biopsies by using multiple state-of-the-art omics and imaging technologies across several institutions. Obtaining rigorous and reproducible results from disparate methods and at different sites to interrogate biomolecules at a single-cell level or in 3-D space is a significant challenge that can be a futile exercise if not well controlled. We describe a “follow the tissue” pipeline for generating a reliable and authentic single-cell/region 3-D molecular atlas of human adult kidney. Our approach emphasizes quality assurance, quality control, validation, and harmonization across different omics and imaging technologies from sample procurement, processing, storage, shipping to data generation, analysis, and sharing. We established benchmarks for quality control, rigor, reproducibility, and feasibility across multiple technologies through a pilot experiment using common source tissue that was processed and analyzed at different institutions and different technologies. A peer review system was established to critically review quality control measures and the reproducibility of data generated by each technology before their being approved to interrogate clinical biopsy specimens. The process established economizes the use of valuable biopsy tissue for multiomics and imaging analysis with stringent quality control to ensure rigor and reproducibility of results and serves as a model for precision medicine projects across laboratories, institutions and consortia.
An Ensemble Learning and Slice Fusion Strategy for Three-Dimensional Nuclei Instance Segmentation
(IEEE, 2022-06) Wu, Liming; Chen, Alain; Salama, Paul; Dunn, Kenneth W.; Delp, Edward J.; Electrical and Computer Engineering, School of Engineering and Technology
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.
Application of Laser Microdissection to Uncover Regional Transcriptomics in Human Kidney Tissue
(MyJove Corporation, 2020-06-09) Barwinska, Daria; Ferkowicz, Michael J.; Cheng, Ying-Hua; Winfree, Seth; Dunn, Kenneth W.; Kelly, Katherine J.; Sutton, Timothy A.; Rovin, Brad H.; Parikh, Samir V.; Phillips, Carrie L.; Dagher, Pierre C.; El-Achkar, Tarek M.; Eadon, Michael T.; Medicine, School of Medicine
Gene expression analysis of human kidney tissue is an important tool to understand homeostasis and disease pathophysiology. Increasing the resolution and depth of this technology and extending it to the level of cells within the tissue is needed. Although the use of single nuclear and single cell RNA sequencing has become widespread, the expression signatures of cells obtained from tissue dissociation do not maintain spatial context. Laser microdissection (LMD) based on specific fluorescent markers would allow the isolation of specific structures and cell groups of interest with known localization, thereby enabling the acquisition of spatially-anchored transcriptomic signatures in kidney tissue. We have optimized an LMD methodology, guided by a rapid fluorescence-based stain, to isolate five distinct compartments within the human kidney and conduct subsequent RNA sequencing from valuable human kidney tissue specimens. We also present quality control parameters to enable the assessment of adequacy of the collected specimens. The workflow outlined in this manuscript shows the feasibility of this approach to isolate sub-segmental transcriptomic signatures with high confidence. The methodological approach presented here may also be applied to other tissue types with substitution of relevant antibody markers.
Automatic segmentation of intravital fluorescence microscopy images by K-means clustering of FLIM phasors
(OSA, 2019-08) Zhang, Yide; Hato, Takashi; Dagher, Pierre C.; Nichols, Evan L.; Smith, Cody J.; Dunn, Kenneth W.; Howard, Scott S.; Medicine, School of Medicine
Fluorescence lifetime imaging microscopy (FLIM) provides additional contrast for fluorophores with overlapping emission spectra. The phasor approach to FLIM greatly reduces the complexity of FLIM analysis and enables a useful image segmentation technique by selecting adjacent phasor points and labeling their corresponding pixels with different colors. This phasor labeling process, however, is empirical and could lead to biased results. In this Letter, we present a novel and unbiased approach to automate the phasor labeling process using an unsupervised machine learning technique, i.e., K-means clustering. In addition, we provide an open-source, user-friendly program that enables users to easily employ the proposed approach. We demonstrate successful image segmentation on 2D and 3D FLIM images of fixed cells and living animals acquired with two different FLIM systems. Finally, we evaluate how different parameters affect the segmentation result and provide a guideline for users to achieve optimal performance.
Boundary Segmentation For Fluorescence Microscopy Using Steerable Filters
(SPIE, 2017) Ho, David Joon; Salama, Paul; Dunn, Kenneth W.; Delp, Edward J.; Electrical and Computer Engineering, School of Engineering and Technology
Fluorescence microscopy is used to image multiple subcellular structures in living cells which are not readily observed using conventional optical microscopy. Moreover, two-photon microscopy is widely used to image structures deeper in tissue. Recent advancement in fluorescence microscopy has enabled the generation of large data sets of images at different depths, times, and spectral channels. Thus, automatic object segmentation is necessary since manual segmentation would be inefficient and biased. However, automatic segmentation is still a challenging problem as regions of interest may not have well defined boundaries as well as non-uniform pixel intensities. This paper describes a method for segmenting tubular structures in fluorescence microscopy images of rat kidney and liver samples using adaptive histogram equalization, foreground/background segmentation, steerable filters to capture directional tendencies, and connected-component analysis. The results from several data sets demonstrate that our method can segment tubular boundaries successfully. Moreover, our method has better performance when compared to other popular image segmentation methods when using ground truth data obtained via manual segmentation.
Center-Extraction-Based Three Dimensional Nuclei Instance Segmentation of Fluorescence Microscopy Images
(IEEE, 2019-05) Ho, David Joon; Han, Shuo; Fu, Chichen; Salama, Paul; Dunn, Kenneth W.; Delp, Edward J.; Electrical and Computer Engineering, School of Engineering and Technology
Fluorescence microscopy is an essential tool for the analysis of 3D subcellular structures in tissue. An important step in the characterization of tissue involves nuclei segmentation. In this paper, a two-stage method for segmentation of nuclei using convolutional neural networks (CNNs) is described. In particular, since creating labeled volumes manually for training purposes is not practical due to the size and complexity of the 3D data sets, the paper describes a method for generating synthetic microscopy volumes based on a spatially constrained cycle-consistent adversarial network. The proposed method is tested on multiple real microscopy data sets and outperforms other commonly used segmentation techniques.
Cilia Associated Signaling in Adult Energy Homeostasis
(2022-05) Bansal, Ruchi; Berbari, Nicolas F.; Perrin, Benjamin J.; Mastracci, Teresa L.; Baucum, Anthony J.; Dunn, Kenneth W.
Primary cilia are solitary cellular appendages that function as signaling centers for cells in adult energy homeostasis. Here in chapter 1, I introduce cilia and how dysfunction of these conserved organelles results in ciliopathies, such as Bardet-Biedl Syndrome (BBS), which present with childhood obesity. Furthermore, conditional loss of primary cilia from neurons in the hypothalamus leads to hyperphagia and obesity in mouse models of ciliopathies. Classically, cilia coordinate signaling often through specific G-protein coupled receptors (GPCRs) as is the case in both vision and olfaction. In addition, neurons throughout the brain including hypothalamic neurons possess primary cilia whose dysfunction contributes to ciliopathy-associated obesity. How neuronal cilia regulate the signaling of GPCRs remains unclear and many fundamental cell biology questions remain about cilia mediated signaling. For example, how cilia coordinate signaling to influence neuronal activity is unknown. To begin to address some of these cell biology questions around neuronal cilia, chapter 2, describes the development and use of a system for primary neuronal cultures from the hypothalamus. Using this system, we found that activation of the cilia regulated hedgehog pathway, which is critical in development, influenced the ability of neurons to respond to GPCR ligands. This result highlights the role of the developmentally critical hedgehog pathway on terminally differentiated hypothalamic neurons. One challenge facing the cilia field is our ability to assess cilia in large numbers without potential bias. This is especially true in tissues like the brain, where cilia appear to have region-specific characteristics. Work included in Chapter 3 describes the use of a computer-assisted artificial intelligence (Ai) approach to analyze cilia composition and morphology in a less biased and high throughput manner. Cilia length and intensities are important parameters for evaluation of cilia signaling. Evidence suggests that activation of some ciliary GPCRs results in shortening of cilia whereas deviations from normal cilia length in mutant phenotypes affects normal physiological processes such as decreased mucociliary clearance. Therefore, to analyze a large number of cilia, we describe the use of the Ai module from in vitro and in vivo samples in a reproducible manner that minimizes user bias. Using this approach, we identified that Mchr1 expression is significantly stronger in the cilia of paraventricular nucleus than that in the arcuate nucleus of adult mice. Work in Chapter 4 continues to explore the integration between hedgehog pathway and ciliary GPCR signaling in the central nervous system, and its relevance with energy homeostasis. We evaluated the hedgehog ligand in the plasma of mice in acute and long-term metabolic changes and identified that the activity of the ligand changed under altered metabolic conditions. We also developed a genetic mouse model where hedgehog signaling was constitutively active in neuronal cilia. These mice become hyperphagic and obese. These results further emphasize the potential role of the hedgehog signaling pathway in regulation of feeding behavior in adult vertebrates. Overall, results from this work will provide a better understanding of the defects not only underlying ciliopathy-associated obesity but may also reveal more common mechanisms of centrally mediated obesity. In addition, the tools I have developed will help in understanding how neuronal cilia are used for intercellular communications and ultimately how they regulate behaviors like feeding.
Clinical, histopathologic and molecular features of idiopathic and diabetic nodular mesangial sclerosis in humans
(Oxford University Press, 2021) Eadon, Michael T.; Lampe, Sam; Baig, Mirza M.; Collins, Kimberly S.; Ferreira, Ricardo Melo; Mang, Henry; Cheng, Ying-Hua; Barwinska, Daria; El-Achkar, Tarek M.; Schwantes-An, Tae-Hwi; Winfree, Seth; Temm, Constance J.; Ferkowicz, Michael J.; Dunn, Kenneth W.; Kelly, Katherine J.; Sutton, Timothy A.; Moe, Sharon M.; Moorthi, Ranjani N.; Phillips, Carrie L.; Dagher, Pierre C.; Medicine, School of Medicine
Background: Idiopathic nodular mesangial sclerosis, also called idiopathic nodular glomerulosclerosis (ING), is a rare clinical entity with an unclear pathogenesis. The hallmark of this disease is the presence of nodular mesangial sclerosis on histology without clinical evidence of diabetes mellitus or other predisposing diagnoses. To achieve insights into its pathogenesis, we queried the clinical, histopathologic and transcriptomic features of ING and nodular diabetic nephropathy (DN). Methods: All renal biopsy reports accessioned at Indiana University Health from 2001 to 2016 were reviewed to identify 48 ING cases. Clinical and histopathologic features were compared between individuals with ING and DN (n = 751). Glomeruli of ING (n = 5), DN (n = 18) and reference (REF) nephrectomy (n = 9) samples were isolated by laser microdissection and RNA was sequenced. Immunohistochemistry of proline-rich 36 (PRR36) protein was performed. Results: ING subjects were frequently hypertensive (95.8%) with a smoking history (66.7%). ING subjects were older, had lower proteinuria and had less hyaline arteriolosclerosis than DN subjects. Butanoate metabolism was an enriched pathway in ING samples compared with either REF or DN samples. The top differentially expressed gene, PRR36, had increased expression in glomeruli 248-fold [false discovery rate (FDR) P = 5.93 × 10-6] compared with the REF and increased 109-fold (FDR P = 1.85 × 10-6) compared with DN samples. Immunohistochemistry revealed a reduced proportion of cells with perinuclear reaction in ING samples as compared to DN. Conclusions: Despite similar clinical and histopathologic characteristics in ING and DN, the uncovered transcriptomic signature suggests that ING has distinct molecular features from nodular DN. Further study is warranted to understand these relationships.
Convolutional neural network denoising in fluorescence lifetime imaging microscopy (FLIM)
(SPIE, 2021) Mannam, Varun; Zhang, Yide; Yuan, Xiaotong; Hato, Takashi; Dagher, Pierre C.; Nichols, Evan L.; Smith, Cody J.; Dunn, Kenneth W.; Howard, Scott; Medicine, School of Medicine
Fluorescence lifetime imaging microscopy (FLIM) systems are limited by their slow processing speed, low signal- to-noise ratio (SNR), and expensive and challenging hardware setups. In this work, we demonstrate applying a denoising convolutional network to improve FLIM SNR. The network will integrated with an instant FLIM system with fast data acquisition based on analog signal processing, high SNR using high-efficiency pulse-modulation, and cost-effective implementation utilizing off-the-shelf radio-frequency components. Our instant FLIM system simultaneously provides the intensity, lifetime, and phasor plots in vivo and ex vivo. By integrating image de- noising using the trained deep learning model on the FLIM data, provide accurate FLIM phasor measurements are obtained. The enhanced phasor is then passed through the K-means clustering segmentation method, an unbiased and unsupervised machine learning technique to separate different fluorophores accurately. Our experimental in vivo mouse kidney results indicate that introducing the deep learning image denoising model before the segmentation effectively removes the noise in the phasor compared to existing methods and provides clearer segments. Hence, the proposed deep learning-based workflow provides fast and accurate automatic segmentation of fluorescence images using instant FLIM. The denoising operation is effective for the segmentation if the FLIM measurements are noisy. The clustering can effectively enhance the detection of biological structures of interest in biomedical imaging applications.

Browsing by Author "Dunn, Kenneth W."

Results Per Page

Sort Options