FIGURE 2 - uploaded by Jionglong Su
Content may be subject to copyright.
Seven types of images utilized in internal backbone extraction. (a) An example of original chromosomes; (b) an approximate central axis; (c) smoothed central axis; (d) the smoothed central axis divided into 11 parts; (e) 10-point central axis; (f) internal backbone; (g) straightened internal backbone with the same length.
Source publication
In medical imaging, chromosome straightening plays a significant role in the pathological study of chromosomes and in the development of cytogenetic maps. Whereas different approaches exist for the straightening task, they are mostly geometric algorithms whose outputs are characterized by jagged edges or fragments with discontinued banding patterns...
Contexts in source publication
Context 1
... 1. We construct the label of a curved chromosome ( Step 2. We improve the performance of deep learning models by generating more augmented chromosomes with different degrees of curvatures. We first apply random elastic deformation [25] and random rotation (from -45 to 45 degree) to the curved chromosome and its backbone simultaneously ( Fig. 2(a) and (f)) until a sizeable number of augmented chromosomes and backbones (1000 pairs in this research) are obtained for training and validation (Fig. 4). Note that the setup of the elastic deformation algorithm [25] is points = 3 and sigma = 18 for 256 × 256 images, in order to generate plausible virtual curvatures. Since we utilize ...
Context 2
... straighten the chromosome, we feed its vertical backbone ( Fig. 2(g)) as input into the optimal U-Net or optimal generator G * b , which will output the corresponding chromosome ( Fig. 5(b)). ...
Context 3
... 1. We construct the label of a curved chromosome ( Step 2. We improve the performance of deep learning models by generating more augmented chromosomes with different degrees of curvatures. We first apply random elastic deformation [25] and random rotation (from -45 to 45 degree) to the curved chromosome and its backbone simultaneously ( Fig. 2(a) and (f)) until a sizeable number of augmented chromosomes and backbones (1000 pairs in this research) are obtained for training and validation (Fig. 4). Note that the setup of the elastic deformation algorithm [25] is points = 3 and sigma = 18 for 256 × 256 images, in order to generate plausible virtual curvatures. Since we utilize ...
