Background: Accurate and robust pathological image analysis for colorectal cancer (CRC) diagnosis is time-consuming and knowledge-intensive, but is essential for CRC patients’ treatment. The current heavy workload of pathologists in clinics/hospitals may easily lead to unconscious misdiagnosis of CRC based on their daily image analyses. Methods: Based on a state-of-the-art transfer-learned deep convolutional neural network in artificial intelligence (AI), we proposed a novel patch aggregation strategy for clinic CRC prediction/diagnosis using weakly labeled pathological whole slide image (WSI) patches. This approach was trained and validated using an unprecedented and enormously large number of 170,099 patches, >14,680 WSIs, from >9,631 subjects that covered diverse and representative clinical cases from multi-independent-sources across China, U.S., and Germany. Results: Our innovative AI tool was consistently nearly perfectly agreed with (average Kappa-statistic 0.896) and even often better than most of the experienced expert pathologists when tested in diagnosing CRC WSIs from multi-centers. The average area under the receiver operating characteristics curve (AUC) of AI was greater than that of the pathologists (0.981 vs 0.970) and achieved the best performance among the application of other AI methods to CRC diagnosis. Our AI-generated heatmap highlights the image regions of cancer tissue/cells. Conclusions: This first-ever generalizable AI system can handle large amounts of WSIs consistently and robustly without potential bias due to fatigue commonly experienced by clinical pathologists. Hence, it will drastically alleviate the heavy clinical burden of daily pathology diagnosis, and improve the treatment for CRC patients. This tool is generalizable to other cancer diagnosis based on image recognition.

Background: Hailey–Hailey disease (HHD) is an autosomal dominant cutaneous disorder that manifests as repeated blisters and erosions on flexural or intertriginous skin areas. The calcium-transporting ATPase type 2C member 1 gene (ATP2C1) encodes the secretory pathway Ca²⁺/Mn²⁺-ATPase 1 (SPCA1), whose deficiency is responsible for HHD. An ATP2C1 splice-site mutation (c.325-2A>G, p.Ala109_Gln120del) was previously identified in a Han Chinese family with HHD. Methods: In this study, the identified ATP2C1 splice-site mutation (c.325-2A>G, p.Ala109_Gln120del) was investigated in transfected human embryonic kidney 293 cells to analyze its pathogenic mechanism in HHD patients by using cycloheximide chase assay, CCK8 assay and in silico modeling of SPCA1 mutant. Results: Cycloheximide chase assay showed that the degradation rate of the SPCA1 mutant was not obviously faster than that of the normal SPCA1. CCK8 assay showed that cell proliferation rates in the wild-type, A109_Q120del, and empty vector control groups all decreased in the gradient Mn²⁺ solutions in a dose-dependent manner. The cell proliferation rate in the wild-type was lower than that in the A109_Q120del and empty vector control (both P < 0.01), indicating overexpression of normal SPCA1 may rather induce Golgi stress, and even cell death. The cell proliferation rate in the A109_Q120del was lower than that in the empty vector control (P < 0.01), indicating that overexpression of the mutated SPCA1 may decrease its detoxification capability. Three-dimensional (3D) structure model of SPCA1 built by SWISS-MODEL and PyMOL showed that absence of the 12 amino acids from p.Ala109 to p.Gln120 in the SPCA1 mutant can cause obviously shortened transmembrane 2, which may affect correct localization of SPCA1 on the Golgi. Conclusion: These results demonstrate that the ATP2C1 mutation (c.325-2A>G, p.Ala109_Gln120del) may cause impaired SPCA1 capability to detoxify Mn²⁺ and abnormal SPCA1 structure, which reveals a new side for the pathogenesis of HHD.

Female breast cancer is the most common and the fifth deadliest cancer worldwide. It is influenced by a combination of genetic, hormonal, and environmental factors. The excision repair cross‐complementation group 3 gene (ERCC3) has recently been identified as a breast cancer susceptibility gene in various cohorts of different geographical and ethnic origin. To explore the role of ERCC3 mutations in breast cancer development and pathological diagnosis, genetic analysis was conducted in 291 patients and 291 controls from mainland China. Bioinformatic analysis and immunohistochemistry (IHC) were performed. A novel ERCC3 mutation p.Y116X was identified in a breast cancer family, while no frequency bias for the genotype and allele of rs754010782 and rs371627165 was observed (all P > 0.05). Bioinformatic analysis revealed that ERCC3 expression was negatively associated with estrogen receptor (ER), progesterone receptor (PR), nontriple‐negative status, and nodal status of breast cancers. ERCC3 amplifications and deep deletions primarily occurred in breast invasive cancer not otherwise specified (NOS) and metaplastic breast cancer, respectively. The decreased ERCC3 expression in tumor tissues of patient with p.Y116X mutation was found by IHC. The ERCC3 mutation p.Y116X may increase breast cancer risk in the Han‐Chinese population. ERCC3 exhibits potential as a biomarker for the pathological diagnosis of breast cancer.