Edge Computing Using Embedded Webserver with Mobile Device for Diagnosis and Prediction of Metastasis in Histopathological Images

Diagnosis of different breast cancer stages using histopathology whole slide images is the gold standard in grading the tissue metastasis. Traditional diagnosis involves labor intensive procedures and is prone to human errors. Computer aided diagnosis assists medical experts as a second opinion tool in early detection which prevents further proliferation. Computing facilities have emerged to an extent where algorithms can attain near human accuracy in prediction of diseases, offering better treatment to curb further proliferation. The work introduced in the paper provides an interface in mobile platform, which enables the user to input histopathology image and obtain the prediction results with its class probability through embedded web-server. The trained deep convolutional neural networks model is deployed into a microcomputer-based embedded system after hyper-parameter tuning, offering congruent performance. The implementation results show that the embedded platform with custom-trained CNN model is suitable for medical image classification, as it takes less execution time and mean prediction time. It is also noticed that customized CNN classifier model outperforms pre-trained models when used in embedded platforms for prediction and classification of histopathology images. This work also emphasizes the relevance of portable and flexible embedded device in real time clinical applications.

Identification S100A9 as a potential biomarker in neuroblastoma

Background More than half of Neuroblastoma (NB) patients presented with distant metastases and the relapse of metastatic patients was up to 90%. It is urgent to explore a biomarker that could facilitate the prediction of metastasis in NB patients. Methods and results In the present study, we systematically analyzed Gene Expression Omnibus datasets and focused on identifying the critical molecular networks and novel key hub genes implicated in NB metastasis. In total, 176 up-regulated and 19 down-regulated differentially expressed genes (DEGs) were identified. Based on these DEGs, a PPI network composed of 150 nodes and 452 interactions was established. Through PPI network identification combined with qRT-PCR, ELISA and IHC, S100A9 was screened as an outstanding gene. Furthermore, in vitro tumorigenesis assays demonstrated that S100A9 overexpression enhanced the proliferation, migration and invasion of NB cells. Conclusions Taken together, our findings suggested that S100A9 could participate in NB tumorigenesis and progression. In addition, S100A9 has the potential to be used as a promising clinical biomarker in the prediction of NB metastasis.

Bioinformatic Analysis Identifying S100a9 Associated With Cell Proliferation and Metastasis in Neuroblastoma

Background: More than half of neuroblastoma (NB) patients presented with distant metastases and the mortality of patients suffering from metastatic relapse was about 90%. It is urgent to find a biomarker that can facilitate the prediction of metastasis in NB patients. Methods and Results: In the present study, we systematically analyzed Gene Expression Omnibus (GEO) datasets and focused on identifying the critical molecular networks and novel key hub genes implicated in NB. We found totally, 176 up-regulated and 19 down-regulated differentially expressed genes (DEGs) were identified. Based on these DEGs, a PPI network composed of 150 nodes and 452 interactions was established. PPI network identification combined with qRT-PCR, ELISA and IHC, S100A9 as was screened as an outstanding gene. Furthermore, in vitro tumorigenesis assays demonstrated that S100A9 overexpression enhanced the proliferation, migration, and invasion of NB cells. Conclusions: Taken together, our findings suggested that S100A9 could participate in NB tumorigenesis and metastasis and that S100A9 has the potential to be used as a biomarker in the prediction of NB metastasis.

Oncogenic Functions and Clinical Significance of Circular RNAs in Colorectal Cancer

Colorectal cancer (CRC) is ranked as the second most commonly diagnosed disease in females and the third in males worldwide. Therefore, the finding of new more reliable biomarkers for early diagnosis, for prediction of metastasis, and resistance to conventional therapies is an important challenge in overcoming the disease. The current review presents circular RNAs (circRNAs) with their unique features as potential prognostic and diagnostic biomarkers in CRC. The review highlights the mechanism of action and the role of circRNAs with oncogenic functions in the CRC as well as the association between their expression and clinicopathological characteristics of CRC patients. The comprehension of the role of oncogenic circRNAs in CRC pathogenesis is growing rapidly and the next step is using them as suitable new drug targets in the personalized treatment of CRC patients.

Construction and Validation of Nomogram for Prediction of Metastasis in Osteosarcoma Patients

Background: Osteosarcoma is one of the most common malignant tumors of the bone with poor prognosis. The present study was aimed to establish and validate nomogram to predict the risk of metastasis for individual osteosarcoma patients.Methods: 2195 osteosarcoma patients were retrospectively collected from Surveillance, Epidemiology, and End Results (SEER) database. The least absolute shrinkage and selection operator regression model was applied to promote feature selection for the metastasis signature model. Univariate and multivariate logistic analysis were used to discern independent predictive factors. These predictive factors were included in the nomogram to evaluate metastasis probability in osteosarcoma patients. Nomogram was validated internally and externally. Results: We randomly assigned 2195 osteosarcoma patients into the training (n = 1464) and validation (n = 731) subgroups. Age, sex, tumor grade, tumor size and use of surgery were identified as the predictive factors for the risk of metastasis (all p<0.05) and further incorporated to develop the nomogram. Predictive accuracy (AUC) for metastasis prediction were 0.7427067 in the training subgroup and 0.6798374 in the external validation subgroup, respectively. And the decision curve analysis also proved the value of the model.Conclusions: We have constructed new high-performance nomogram to assist clinicians in identifying osteosarcoma patients with high risk of metastasis.

Clinical correlation of plasma amino acid profile of post-surgical gastric cancer patients and metastatic gastric cancer patients for establishing biomarker of metastasis

Objectives: This research is aimed to evaluate plasma free amino acid in gastric cancer patients without metastasis (early gastric cancer post gastrectomy) and with metastasis (advanced gastric cancer). Amino acids level of postoperative gastric cancer (M0) patients are compared with metastatic gastric cancer (M1) patients in search of biomarker which can predict the metastasis of gastric cancer. We have made clinical correlation of patients’ vital signs, respiratory rate, pulse rate, blood pressure, body temperature, disease stages, chief complaints, complications and survival curve within light of metastatic and nonmetastatic domain. Background: Majority of cancer patients are diagnosed after seeding of metastatic cells to adjacent organs and distant sites. At this point, treatment is palliative and supportive. The cellular propagation of cancer cells and tumor micro-environment plays vital role in genesis of gastric cancer. Genetic alteration leading to faulty nucleotides to amino acids, then to protein, and finally formation of tumor is the natural sequence of pathogenesis of gastric cancer. Prediction of metastasis by use of plasma free amino acid profile may be of great significance because it will help to tailor the patient specific cancer treatment. Plasma Amino acids are ideal for being developed as tool for prediction of metastasis as they are affordable, less expensive and convenient. Method: This study includes total 54 patients, among which 27 had metastasis of Gastric cancer and rest 27 had undergone gastric surgery at early stage with no recurrence at the time of the study. Twenty-three amino acids were studied. Student’s t test was performed to find out statistically significant values of amino acids. The p value of ≤ 0.05 was considered statistically significant. Amino acids with significant p values were investigated with multivariate logistic regression. Partial Least Squares Discriminant Analysis (PLS DA) was done using Microsoft SPSS 23 version software®. Variable Importance of Projection (VIP) was estimated, values ≥ 1 was considered statistically significant. Result: Performance Score (PS) (p= 0.004) and Body Mass Index (BMI) (p= 0.035) were statistically significant between M0 and M1 groups. Staging (I, II vs. III, IV) (p< 0.001) was significant. Seven amino acids, Asp, Cys, Hcy, His, Leu, Orn and Ser were significant between M0 and M1 in first month evaluation. Eight amino acids, Cys, Hcy, His, Leu, Met, Thr, Trp and Tyr were significant between M0 and M1 in sixth month evaluation. PLS DA regression analysis, VIP test showed Cys, Ser, Hcy, Thr, His, Met, Tyr, Trp to be more important amino acids of significance. Kaplan Meier Overall Survival (OS) = 34.979 months. Mean survival time in M0 was 43.53± 1.741 months. Mean survival in M1 was 26.29± 2.635 months. Conclusion: We found BMI and PS as most important variables in defining and determining the disease status of gastric cancer patients. Nutrition and physical activity is very much characteristic of disease outcome from a physician’s perspective. This study propounds amino acids can be valuable biomarkers of predictive and prognostic importance in metastasis in gastric cancer patients.