Learning based automatic detection of myonuclei in isolated single skeletal muscle fibers using multi-focus image fusion

Abstract

Accurate and robust detection of myonuclei in single muscle fiber is required to calculate myonuclear domain size. However, this task is challenging because: 1) The myonuclei have a variety of sizes and shapes. 2) Imaging techniques exhibit myonuclei that are often overlapping. 3) Inhomogeneous intensity due to DAPI concentration in heterochromatin, abundant in mouse nuclei, results in a speckled appearance inside each myonucleus. In this paper, we propose a novel automatic approach to robustly detect the myonuclei in isolated single skeletal muscle fibers. The original z-stack images are first fused into one all-in-focus image. A sufficient number of ellipse fitting hypotheses are then generated using the myonuclei contour segments. A set of morphological features are calculated from the ellipses and utilized to train a support vector machine (SVM) classifier to choose the best candidates. A modified inner geodesic distance based clustering algorithm is used to produce the final results. The proposed method was extensively tested using 42 sets of z-stack images containing about 1500 myonuclei. The method demonstrates excellent results outperforming current state-of-the-arts.

Robust Nucleus/Cell Detection and Segmentation in Digital Pathology and Microscopy Images: A Comprehensive Review

“…Su et al [122], [132] have applied a binary SVM classifier to automatic myonuclear detection in isolated single muscle fiber fluorescence images. It consists of four major steps: 1) extract foreground edges by applying the Otsu's method [63] to the fused z-stack image, 2) fit a sufficient number of ellipse hypotheses using heteroscedastic Errors-in-variable (HEIV) regression [133], 3) use SVM with a set of specifically designed ellipse features to select the good candidate fittings, and 4) apply inner geodesic distance based clustering to ellipse refinement, which determines the locations of myonuclear centers.…”

Combined flat-field and frequency filter approach to correcting artifacts of multichannel two-photon microscopy

“…Early WSI systems incorporated linear array detectors to mitigate the effect of uneven sample illumination that would cause vignetting of scanned images 6 . For MPM and confocal microscopy, z-stacking, or the collection of image tiles at increment points of focus through a sample, and the creation of a composite image from the combined z-scans can have a significant impact on data quality by reducing the effect of imperfect sectioning 7 . This process can correct for brightness nonuniformity over large acquisition areas due to the uneven texture of the sample placing areas outside of the initial focal plane.…”

“… 6 For MPM and confocal microscopy, -stacking, or the collection of image tiles at increment points of focus through a sample, and the creation of a composite image from the combined -scans can have a significant impact on data quality by reducing the effect of imperfect sectioning. 7 This process can correct for brightness nonuniformity over large acquisition areas due to the uneven texture of the sample placing areas outside of the initial focal plane. In addition, postprocessing techniques for reducing stripe, grid, or tiling artifacts have been a wide area of research in the field of aerial imaging and microscopy.…”

Using Deep Learning to Automate Eosinophil Counting in Pediatric Ulcerative Colitis Histopathological Images

“…Cell counting algorithms have existed for some time as reviewed by Xing et al [8]. Machine learning algorithms for cell counting include Distance Transform [9], morphology operation [10], H-Minima [10] and H-Maxima Transform [11], Laplacian of Gaussian [12], Maximally Stable Extremal Region [13], Hough Transform [14,15], Radial Symmetry-Based Voting [16,17], and Supervised Learning (Support Vector Machine [18,19], Random Forest [20], and Deep Learning [21][22][23]). Our method builds on the success of deep learning in a variety of image processing tasks.…”