Semantic Segmentation of Intralobular and Extralobular Tissue from Liver Scaffold H&E Images

Abstract

Decellularized tissue is an important source for biological tissue engineering. Evaluation of the quality of decellularized tissue is performed using scanned images of hematoxylin-eosin stained (H&E) tissue sections and is usually dependent on the observer. The first step in creating a tool for the assessment of the quality of the liver scaffold without observer bias is the automatic segmentation of the whole slide image into three classes: the background, intralobular area, and extralobular area. Such segmentation enables to perform the texture analysis in the intralobular area of the liver scaffold, which is crucial part in the recellularization procedure. Existing semi-automatic methods for general segmentation (i.e., thresholding, watershed, etc.) do not meet the quality requirements. Moreover, there are no methods available to solve this task automatically. Given the low amount of training data, we proposed a two-stage method. The first stage is based on classification of simple hand-crafted descriptors of the pixels and their neighborhoods. This method is trained on partially annotated data. Its outputs are used for training of the second-stage approach, which is based on a convolutional neural network (CNN). Our architecture inspired by U-Net reaches very promising results, despite a very low amount of the training data. We provide qualitative and quantitative data for both stages. With the best training setup, we reach 90.70% recognition accuracy.

Using virtual microscopy for the development of sampling strategies in quantitative histology and design‐based stereology

“…The resulting parameters describe the level of ECM structure preservation within individual lobules. It is a two‐step procedure: (b) The first step is a whole scan analysis separating the lobules (segmentation to distinguish the sinusoidal ECM from more dense interlobular septa, triads and vessels) (Jirik et al, 2020, 2021). Typically, 5 lobules are then selected either manually by the end user or automatically (marked with cyan dots in b).…”

“…The importance of proper sampling of sections is elaborated in Figure S3. We found that the tissue composition of the canine soft palate was so heterogeneous that a whole half of the velum had to be sampled using a minimum of 10-15 sections distributed approximately uniformly throughout the complete tissue block to obtain The first step is a whole scan analysis separating the lobules (segmentation to distinguish the sinusoidal ECM from more dense interlobular septa, triads and vessels) (Jirik et al, 2020(Jirik et al, , 2021. Typically, 5 lobules are then selected either manually by the end user or automatically (marked with cyan dots in b).…”

Artificial Intelligence-Based Opportunities in Liver Pathology—A Systematic Review

“…Twenty articles were reviewed through full text screening (Table 3) [34][35][36][37][38][39][40][41][42][43][44][45][46][47][48][49][50][51][52][53]. As for tumoral studies, none used prospectively collected data.…”

Identification and location method of strip ingot for autonomous robot system using kmeans clustering and color segmentation

“…Considering that simple line detection can only determine the longest edge of the ingot, but cannot determine the position of the short edge of the ingot, image segmentation method can be used to extract the ingot distribution area. In the image segmentation algorithm, the commonly used image segmentation algorithms include Otsu threshold segmentation [33][34][35], Fully Convolutional Networks (FCN) [36] and other semantic segmentation [37]. Otsu segmentation algorithm is simple, fast, and not affected by the brightness and contrast of the image [34].…”