Exploring pathological signatures for predicting the recurrence of early-stage hepatocellular carcinoma based on deep learning.

Background: Postoperative recurrence impedes the curability of early-stage hepatocellular carcinoma (E-HCC). We aimed to establish a novel recurrence-related pathological prognosticator with artificial intelligence, and investigate the relationship between pathological features and the local immunological microenvironment. Methods: A total of 576 whole-slide images (WSIs) were collected from 547 patients with E-HCC in the Zhongshan cohort, which was randomly divided into a training cohort and a validation cohort. The external validation cohort comprised 147 Tumor Node Metastasis (TNM) stage I patients from The Cancer Genome Atlas (TCGA) database. Six types of HCC tissues were identified by a weakly supervised convolutional neural network. A recurrence-related histological score (HS) was constructed and validated. The correlation between immune microenvironment and HS was evaluated through extensive immunohistochemical data. Results: The overall classification accuracy of HCC tissues was 94.17%. The C-indexes of HS in the training, validation and TCGA cohorts were 0.804, 0.739 and 0.708, respectively. Multivariate analysis showed that the HS (HR= 4.05, 95% CI: 3.40-4.84) was an independent predictor for recurrence-free survival. Patients in HS high-risk group had elevated preoperative alpha-fetoprotein levels, poorer tumor differentiation and a higher proportion of microvascular invasion. The immunohistochemistry data linked the HS to local immune cell infiltration. HS was positively correlated with the expression level of peritumoral CD14+ cells (p= 0.013), and negatively with the intratumoral CD8+ cells (p< 0.001). Conclusions: The study established a novel histological score that predicted short-term and long-term recurrence for E-HCCs using deep learning, which could facilitate clinical decision making in recurrence prediction and management.

Artificial Intelligence-Based Multiclass Classification of Benign or Malignant Mucosal Lesions of the Stomach.

Gastric cancer (GC) is one of the leading causes of cancer-related death worldwide. It takes some time from chronic gastritis to develop in GC. Early detection of GC will help patients obtain timely treatment. Understanding disease evolution is crucial for the prevention and treatment of GC. Here, we present a convolutional neural network (CNN)-based system to detect abnormalities in the gastric mucosa. We identified normal mucosa, chronic gastritis, and intestinal-type GC: this is the most common route of gastric carcinogenesis. We integrated digitalizing histopathology of whole-slide images (WSIs), stain normalization, a deep CNN, and a random forest classifier. The staining variability of WSIs was reduced significantly through stain normalization, and saved the cost and time of preparing new slides. Stain normalization improved the effect of the CNN model. The accuracy rate at the patch-level reached 98.4%, and 94.5% for discriminating normal → chronic gastritis → GC. The accuracy rate at the WSIs-level for discriminating normal tissue and cancerous tissue reached 96.0%, which is a state-of-the-art result. Survival analyses indicated that the features extracted from the CNN exerted a significant impact on predicting the survival of cancer patients. Our CNN model disclosed significant potential for adjuvant diagnosis of gastric diseases, especially GC, and usefulness for predicting the prognosis.

[Algicidal activity against red-tide algaes by marine bacterial strain N3 isolated from a HABs area, southern China].

A marine algicidal bacterium N3 was isolated from a HABs area in Mirs Bay, a subtropical bay, in southern China. Algicidal activity and algicidal mode against Phaeodactylum tricornutum, Scrippsiella trochoidea, Prorocentrum micans and Skeletonema costatum were observed by the liquid infection method. The results showed that there were no algicidal activities against P. tricornutum and S. costatum. However, when the bacterial volume fractions were 2% and 10% , S. trochoidea and P. micans could be killed, respectively. S. trochoidea cells which were exposed to strain N3 became irregular in shape and the cellular components lost their integrity and were decomposed. While, the P. micans cells became inflated and the cellular components aggregated, followed by cell lysis. Strain N3 killed S. trochoidea and P. micans directly, and the algicidal activities of the bacterial strain N3 was concentration-dependent. To S. trochoidea, 2% (V/V) of bacteria in algae showed the strongest algicidal activity, all of the S. trochoidea cells were killed within 120 h. But the growth rates of cells, in the 1% and 0. 1% treatment groups, were only slightly lower than that in the control group. In all treatment groups, the densities of strain N3 were in declining trends. While, to P. micans, 10% and 5% of bacteria in algae showed strong algicidal activities, 78% and 70% of the S. trochoidea were killed within 120 h, respectively. However, the number of S. trochoidea after exposure to 1% of bacterial cultures still increased up to 5 incubation days. And in the three treatment groups, the densities of strain N3 experienced a decrease process. The isolated strain N3 was identified as Bacillus sp. by morphological observation, physiological and biochemical characterization, and homology comparisons based on 16S rRNA sequences.

Algicidal activity against Skeletonema costatum by marine bacteria isolated from a high frequency harmful algal blooms area in southern Chinese coast.

Four marine bacterial strains P1, P5, N5 and N21 were isolated from the surface water and sediment of Mirs Bay in southern Chinese coast using the liquid infection method with 48-well plates. These bacteria were all shown to have algicidal activities against Skeletonema costatum. Based on morphological observations, biochemical tests and homology comparisons by 16S rDNA sequences, the isolated strains P1, P5, N5 and N21 were identified as Halobacillus sp., Muricauda sp., Kangiella sp. and Roseivirga sp., respectively. Our results showed that bacterial strain P1 killed S. costatum by release of heat labile algicide, while strains P5, N5 and N21 killed them directly. The algicidal processes of four bacterial strains were different. Strains P1, N5 and N21 disrupted the chain structure and S. costatum appeared as single cells, in which the cellular components were aggregated and the individual cells were inflated and finally lysed, while strain P5 decomposed the algal chains directly. We also showed that the algicidal activities of the bacterial strains were concentration-dependent. More specifically, 10 % (v/v) of bacteria in algae showed the strongest algicidal activities, as all S. costatum cells were killed by strains N5 and N21 within 72 h and by strains P1 and P5 within 96 h. 5 % of bacteria in algae also showed significant algicidal activities, as all S. costatum were killed by strains N5, P5 and N21 within 72, 96 and 120 h, respectively, whereas at this concentration, only 73.4 % of S. costatum cells exposed to strain P1 were killed within 120 h. At the concentration of 1 % bacteria in algae, the number of S. costatum cells continued to increase and the growth rate of algae upon exposure to strain N5 was significantly inhibited.