Digital Pathology and Toxicogenomic proved extremely essential in today's post pandemic era. Generally, the pathology process consists of collection of body tissues (biopsy), inspection, interpretation, and storage of findings. This process needs to be fast and accurate; therefore, AI and machine learning were utilized to automate processes that take long time and effort from professional pathologists. In the case of liver pathology, Pathologists collect whole slide images (WSI), annotate liver markers then use them to inspect liver images for injury. The liver is divided histologically into lobules where diagnosis of disease is highly relative to its location. In this paper, we propose an Alternative AI System that consist of 1) WSI in-supervised per-processing and patch-of-interest annotation. 2) a semi-supervised learning Constitutional Neural Network (CNN) segmentation approach to annotate liver lobules in WSI. 3) a supervised CNN metric architecture to annotate injury presence in the liver. We introduce a new sub-dataset inspired by TG-GATEs open-source dataset named (TG-GATEs-CV). The dataset consists of images of central veins that are used in training the central lobule segmentation network. The System was tested on 120 random injured and non-injured central lobule WSIs from TG-GATEs dataset. The proposed system alleviates the time a pathologist takes to identify central veins, portal triads and liver lobules in a whole slide image. It also has the potential to lessen the time consumed to highlight injured tissue in different zones of the liver.