AFEI: adaptive optimized vertical federated learning for heterogeneous multi-omics data integration.

Vertical federated learning has gained popularity as a means of enabling collaboration and information sharing between different entities while maintaining data privacy and security. This approach has potential applications in disease healthcare, cancer prognosis prediction, and other industries where data privacy is a major concern. Although using multi-omics data for cancer prognosis prediction provides more information for treatment selection, collecting different types of omics data can be challenging due to their production in various medical institutions. Data owners must comply with strict data protection regulations such as European Union (EU) General Data Protection Regulation. To share patient data across multiple institutions, privacy and security issues must be addressed. Therefore, we propose an adaptive optimized vertical federated-learning-based framework adaptive optimized vertical federated learning for heterogeneous multi-omics data integration (AFEI) to integrate multi-omics data collected from multiple institutions for cancer prognosis prediction. AFEI enables participating parties to build an accurate joint evaluation model for learning more information related to cancer patients from different perspectives, based on the distributed and encrypted multi-omics features shared by multiple institutions. The experimental results demonstrate that AFEI achieves higher prediction accuracy (6.5% on average) than using single omics data by utilizing the encrypted multi-omics data from different institutions, and it performs almost as well as prognosis prediction by directly integrating multi-omics data. Overall, AFEI can be seen as an efficient solution for breaking down barriers to multi-institutional collaboration and promoting the development of cancer prognosis prediction.

Artificial intelligence-based multi-omics analysis fuels cancer precision medicine.

With biotechnological advancements, innovative omics technologies are constantly emerging that have enabled researchers to access multi-layer information from the genome, epigenome, transcriptome, proteome, metabolome, and more. A wealth of omics technologies, including bulk and single-cell omics approaches, have empowered to characterize different molecular layers at unprecedented scale and resolution, providing a holistic view of tumor behavior. Multi-omics analysis allows systematic interrogation of various molecular information at each biological layer while posing tricky challenges regarding how to extract valuable insights from the exponentially increasing amount of multi-omics data. Therefore, efficient algorithms are needed to reduce the dimensionality of the data while simultaneously dissecting the mysteries behind the complex biological processes of cancer. Artificial intelligence has demonstrated the ability to analyze complementary multi-modal data streams within the oncology realm. The coincident development of multi-omics technologies and artificial intelligence algorithms has fuelled the development of cancer precision medicine. Here, we present state-of-the-art omics technologies and outline a roadmap of multi-omics integration analysis using an artificial intelligence strategy. The advances made using artificial intelligence-based multi-omics approaches are described, especially concerning early cancer screening, diagnosis, response assessment, and prognosis prediction. Finally, we discuss the challenges faced in multi-omics analysis, along with tentative future trends in this field. With the increasing application of artificial intelligence in multi-omics analysis, we anticipate a shifting paradigm in precision medicine becoming driven by artificial intelligence-based multi-omics technologies.

Development and Validation of a Nomogram for Prognosis Prediction in Patients with Synchronous Primary Thyroid and Breast Cancer Based on SEER Database.

This study aimed to develop prognostic prediction models for patients diagnosed with synchronous thyroid and breast cancer (TBC). Utilizing the SEER database, key predictive factors were identified, including T stage of thyroid cancer, T stage of breast cancer, M stage of breast cancer, patient age, thyroid cancer surgery type, and isotope therapy. A nomogram predicting 5-year and 10-year survival rates was constructed and validated, exhibiting strong performance (C-statistic: 0.79 in the development cohort (95% CI: 0.74-0.84), and 0.82 in the validation cohort (95% CI: 0.77-0.89)). The area under the Receiver Operator Characteristic (ROC) curve ranged from 0.798 to 0.883 for both cohorts. Calibration and decision curve analyses further affirmed the model's clinical utility. Stratifying patients into high-risk and low-risk groups using the nomogram revealed significant differences in survival rates (P < 0.0001). The successful development and validation of this nomogram for predicting 5-year and 10-year survival rates in patients with synchronous TBC hold promise for similar patient populations, contributing significantly to cancer research.

A comprehensive prognostic and immunological implications of PFKP in pan-cancer.

BACKGROUND: Phosphofructokinase P (PFKP) is a key rate-limiting enzyme in glycolysis, playing a crucial role in various pathophysiological processes. However, its specific function in tumors remains unclear. This study aims to evaluate the expression and specific role of PFKP across multiple tumor types (Pan-cancer) and to explore its potential clinical significance as a therapeutic target in cancer treatment. METHODS: We analyzed the expression of PFKP, immune cell infiltration, and patient prognosis across various cancers using data from the Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases. Additionally, we conducted a series of experiments in lung cancer cells, including Western blot, CCK-8 assay, colony formation assay, transwell migration assay, scratch wound healing assay, LDH release assay, and flow cytometry, to evaluate the impact of PFKP on tumor cells. RESULTS: PFKP was found to be highly expressed in most cancers and identified as a prognostic risk factor. Elevated PFKP expression is associated with poorer clinical outcomes, particularly in lung adenocarcinoma (LUAD). Receiver operating characteristic (ROC) curve analysis indicated that PFKP can effectively differentiate between cancerous and normal tissues. The expression of PFKP in most tumors showed significant correlations with tumor mutational burden (TMB), microsatellite instability (MSI), immune score, and immune cell infiltration. In vitro experiments demonstrated that PFKP overexpression promotes lung cancer cell proliferation and migration while inhibiting apoptosis, whereas PFKP deficiency results in the opposite effects. CONCLUSION: PFKP acts as an oncogene involved in tumorigenesis and may influence the immune microenvironment within the tumor. Our findings suggest that PFKP could serve as a potential biomarker for predicting prognosis and the efficacy of immunotherapy in tumors.

Serum CXCL13 level is related to treatment response and predicts disease prognosis in Waldenström macroglobulinemia.

Waldenström macroglobulinemia (WM) is a type of B-cell lymphoma that produces IgM. Our study aimed to investigate the role of CXCL13, a chemokine essential for B lymphocytes, in the evaluation of treatment response and prognosis in WM. We collected serum samples and clinical data from 72 WM patients, with 69 patients receiving systemic therapy and 3 patients opting not to receive treatment. Serum CXCL13 levels at baseline and after six months of treatments were measured by enzyme-linked immunosorbent assay. The median serum level of CXCL13 was 1 539.2 pg/ml (range 10.0-21 389.9) at baseline and significantly decreased to 123.1 pg/ml (range 0.0-6 741.5) after 6 months of treatments. At baseline, higher CXCL13 levels were associated with lower hemoglobin levels (p = 0.001), higher ß2-microglobulin levels (p = 0.001), lower albumin levels (p = 0.046), and higher IPSS-WM scores (p = 0.013). After 6 months of treatment, patients who achieved PR/VGPR had significantly lower CXCL13 levels compared to those with SD (70.2 pg/ml vs 798.6 pg/ml, p = 0.002). The median follow-up period was 40 months (range 4.2-188). Eight patients died during the follow-up period. Overall survival differed based on CXCL13 levels. When grouped by baseline CXCL13 levels, the median OS was 60.0 months in patients with serum CXCL13 > 2 000 pg/ml, while it was not reached in patients with low CXCL13 levels (p < 0.001). Based on CXCL13 levels after the treatments, the median OS was 74.0 months in patients with serum CXCL13 > 200 pg/ml, while it was not reached in patients with CXCL13 ≤ 200 pg/ml. In a subgroup of 28 patients with a series of serum samples, the increase of serum CXCL13 level was associated with disease progression or the start of next-line therapy (p < 0.001). Our study concludes that serum CXCL13 levels decrease in WM patients treated with various regimens and correlate with treatment response. Detecting serum CXCL13 at baseline or after treatment help in predicting prognosis.

Psychosis Prognosis Predictor: A continuous and uncertainty-aware prediction of treatment outcome in first-episode psychosis.

INTRODUCTION: Machine learning models have shown promising potential in individual-level outcome prediction for patients with psychosis, but also have several limitations. To address some of these limitations, we present a model that predicts multiple outcomes, based on longitudinal patient data, while integrating prediction uncertainty to facilitate more reliable clinical decision-making. MATERIAL AND METHODS: We devised a recurrent neural network architecture incorporating long short-term memory (LSTM) units to facilitate outcome prediction by leveraging multimodal baseline variables and clinical data collected at multiple time points. To account for model uncertainty, we employed a novel fuzzy logic approach to integrate the level of uncertainty into individual predictions. We predicted antipsychotic treatment outcomes in 446 first-episode psychosis patients in the OPTiMiSE study, for six different clinical scenarios. The treatment outcome measures assessed at both week 4 and week 10 encompassed symptomatic remission, clinical global remission, and functional remission. RESULTS: Using only baseline predictors to predict different outcomes at week 4, leave-one-site-out validation AUC ranged from 0.62 to 0.66; performance improved when clinical data from week 1 was added (AUC = 0.66-0.71). For outcome at week 10, using only baseline variables, the models achieved AUC = 0.56-0.64; using data from more time points (weeks 1, 4, and 6) improved the performance to AUC = 0.72-0.74. After incorporating prediction uncertainties and stratifying the model decisions based on model confidence, we could achieve accuracies above 0.8 for ~50% of patients in five out of the six clinical scenarios. CONCLUSION: We constructed prediction models utilizing a recurrent neural network architecture tailored to clinical scenarios derived from a time series dataset. One crucial aspect we incorporated was the consideration of uncertainty in individual predictions, which enhances the reliability of decision-making based on the model's output. We provided evidence showcasing the significance of leveraging time series data for achieving more accurate treatment outcome prediction in the field of psychiatry.

M1 macrophage-related gene model for NSCLC immunotherapy response prediction.

Patients diagnosed with non-small cell lung cancer (NSCLC) have a limited lifespan and exhibit poor immunotherapy outcomes. M1 macrophages have been found to be essential for antitumor immunity. This study aims to develop an immunotherapy response evaluation model for NSCLC patients based on transcription. RNA sequencing profiles of 254 advanced-stage NSCLC patients treated with immunotherapy are downloaded from the POPLAR and OAK projects. Immune cell infiltration in NSCLC patients is examined, and thereafter, different coexpressed genes are identified. Next, the impact of M1 macrophage-related genes on the prognosis of NSCLC patients is investigated. Six M1 macrophage coexpressed genes, namely, NKX2-1, CD8A , SFTA3, IL2RB, IDO1, and CXCL9, exhibit a strong association with the prognosis of NSCLC and serve as effective predictors for immunotherapy response. A response model is constructed using a Cox regression model and Lasso Cox regression analysis. The M1 genes are validated in our TD-FOREKNOW NSCLC clinical trial by RT-qPCR. The response model shows excellent immunotherapy response prediction and prognosis evaluation value in advanced-stage NSCLC. This model can effectively predict advanced NSCLC prognosis and aid in identifying patients who could benefit from customized immunotherapy as well as sensitive drugs.

A novel nomogram for predicting overall survival in patients with tongue squamous cell carcinoma using clinical features and MRI radiomics data: a pilot study.

OBJECTIVE: Tongue squamous cell carcinoma (TSCC) accounts for 43.4% of oral cancers in China and has a poor prognosis. This study aimed to explore whether radiomics features extracted from preoperative magnetic resonance imaging (MRI) could predict overall survival (OS) in patients with TSCC. METHODS: The clinical imaging data of 232 patients with pathologically confirmed TSCC at Xiangyang No. 1 People's Hospital were retrospectively analyzed from February 2010 to October 2022. Based on 2-10 years of follow-up, patients were categorized into two groups: control (healthy survival, n = 148) and research (adverse events: recurrence or metastasis-related death, n = 84). A training and a test set were established using a 7:3 ratio and a time node. Radiomics features were extracted from axial T2-weighted imaging, contrast-enhanced T1-weighted imaging, and diffusion-weighted imaging (DWI) sequences. The corresponding radiomics scores were generated using the least absolute shrinkage and selection operator algorithm. Kaplan-Meier and multivariate Cox regression analyses were used to screen for independent factors affecting adverse events in patients with TSCC using clinical and pathological results. A novel nomogram was established to predict the probability of adverse events and OS in patients with TSCC. RESULTS: The incidence of adverse events within 2-10 years after surgery was 36.21%. Kaplan-Meier analysis revealed that hot pot consumption, betel nut chewing, platelet-lymphocyte ratio, drug use, neutrophil-lymphocyte ratio, Radscore, and other factors impacted TSCC survival. Multivariate Cox regression analysis revealed that the clinical stage (P < 0.001), hot pot consumption (P < 0.001), Radscore 1 (P = 0.01), and Radscore 2 (P < 0.001) were independent factors affecting TSCC-OS. The same result was validated by the XGBoost algorithm. The nomogram based on the aforementioned factors exhibited good discrimination (C-index 0.86/0.81) and calibration (P > 0.05) in the training and test sets, accurately predicting the risk of adverse events and survival. CONCLUSION: The nomogram constructed using clinical data and MRI radiomics parameters may accurately predict TSCC-OS noninvasively, thereby assisting clinicians in promptly modifying treatment strategies to improve patient prognosis.

A predictive study of genes related to lactic acid metabolism in cervical carcinoma.

OBJECTIVES: Lactic acid metabolism, a hallmark of carcinogenesis, may play potential roles in cervical carcinoma, assisting the prognosis prediction. MATERIALS AND METHODS: A regression analysis was conducted to identify the ones with the most frequent variation in mutations and CNV changes in lactate metabolism-related (L-related) genes, after which a prognostic nomogram was built based on selected genes and clinical features by machine learning methods. RESULTS: EGLN1, IL1, IL12RB1, ENO1, and 10 other genes had the most frequent changes and prognostic differences in overall survival (OS). The lactated associated risk (LAR) score model can distinguish the patients in OS (p = 0.046, HR = 101.9, 95%CI 1.1-9447.6), and together with clinical features has a higher AUC (AUC = 0.839). Furthermore, CD8+ T, activated CD4+ memory T and resting mast cells were significantly negatively associated with the LAR score. CONCLUSIONS: Lactic acid metabolism is closely related to the prognosis of cervical carcinoma, where the immune microenvironment may play an important role.

It is time to implement molecular classification in endometrial cancer.

A huge effort has been done in redefining endometrial cancer (EC) risk classes in the last decade. However, known prognostic factors (FIGO staging and grading, biomolecular classification and ESMO-ESGO-ESTRO risk classes stratification) are not able to predict outcomes and especially recurrences. Biomolecular classification has helped in re-classifying patients for a more appropriate adjuvant treatment and clinical studies suggest that currently used molecular classification improves the risk assessment of women with EC, however, it does not clearly explain differences in recurrence profiles. Furthermore, a lack of evidence appears in EC guidelines. Here, we summarize the main concepts why molecular classification is not enough in the management of endometrial cancer, by highlighting some promising innovative examples in scientific literature studies with a clinical potential significant impact.

The prognostic role of MRI-based radiomics in tongue carcinoma: a multicentric validation study.

PURPOSE: Radiomics is an emerging field that utilizes quantitative features extracted from medical images to predict clinically meaningful outcomes. Validating findings is crucial to assess radiomics applicability. We aimed to validate previously published magnetic resonance imaging (MRI) radiomics models to predict oncological outcomes in oral tongue squamous cell carcinoma (OTSCC). MATERIALS AND METHODS: Retrospective multicentric study on OTSCC surgically treated from 2010 to 2019. All patients performed preoperative MRI, including contrast-enhanced T1-weighted (CE-T1), diffusion-weighted sequences and apparent diffusion coefficient map. We evaluated overall survival (OS), locoregional recurrence-free survival (LRRFS), cause-specific mortality (CSM). We elaborated different models based on clinical and radiomic data. C-indexes assessed the prediction accuracy of the models. RESULTS: We collected 112 consecutive independent patients from three Italian Institutions to validate the previously published MRI radiomic models based on 79 different patients. The C-indexes for the hybrid clinical-radiomic models in the validation cohort were lower than those in the training cohort but remained > 0.5 in most cases. CE-T1 sequence provided the best fit to the models: the C-indexes obtained were 0.61, 0.59, 0.64 (pretreatment model) and 0.65, 0.69, 0.70 (posttreatment model) for OS, LRRFS and CSM, respectively. CONCLUSION: Our clinical-radiomic models retain a potential to predict OS, LRRFS and CSM in heterogeneous cohorts across different centers. These findings encourage further research, aimed at overcoming current limitations, due to the variability of imaging acquisition, processing and tumor volume delineation.

Multimodality Risk Assessment of Patients with Ischemic Heart Disease Using Deep Learning Models Applied to Electrocardiograms and Chest X-rays.

Comprehensive management approaches for patients with ischemic heart disease (IHD) are important aids for prognostication and treatment planning. While single-modality deep neural networks (DNNs) have shown promising performance for detecting cardiac abnormalities, the potential benefits of using DNNs for multimodality risk assessment in patients with IHD have not been reported. The purpose of this study was to investigate the effectiveness of multimodality risk assessment in patients with IHD using a DNN that utilizes 12-lead electrocardiograms (ECGs) and chest X-rays (CXRs), with the prediction of major adverse cardiovascular events (MACEs) being of particular concern.DNN models were applied to detection of left ventricular systolic dysfunction (LVSD) on ECGs and identification of cardiomegaly findings on CXRs. A total of 2107 patients who underwent elective percutaneous coronary intervention were categorized into 4 groups according to the models' outputs: Dual-modality high-risk (n = 105), ECG high-risk (n = 181), CXR high-risk (n = 392), and No-risk (n = 1,429).A total of 342 MACEs were observed. The incidence of a MACE was the highest in the Dual-modality high-risk group (P < 0.001). Multivariate Cox hazards analysis for predicting MACE revealed that the Dual-modality high-risk group had a significantly higher risk of MACE than the No-risk group (hazard ratio (HR): 2.370, P < 0.001), the ECG high-risk group (HR: 1.906, P = 0.010), and the CXR high-risk group (HR: 1.624, P = 0.018), after controlling for confounding factors.The results suggest the usefulness of multimodality risk assessment using DNN models applied to 12-lead ECG and CXR data from patients with IHD.

Integrated bioinformatics analysis for conducting a prognostic model and identifying immunotherapeutic targets in gastric cancer.

BACKGROUND: Gastric cancer is the third leading cause of death from cancer worldwide and has a poor prognosis. Practical risk scores and prognostic models for gastric cancer are lacking. While immunotherapy has succeeded in some cancers, few gastric cancer patients benefit from immunotherapy. Immune genes and the tumor microenvironment (TME) are essential for cancer progression and immunotherapy response. However, the roles of immune genes and the tumor microenvironment in immunotherapy remain unclear. The study aimed to construct a prognostic prediction model and identify immunotherapeutic targets for gastric cancer (GC) patients by exploring immune genes and the tumor microenvironment. RESULTS: An immune-related risk score (IRRS) model, including APOH, RNASE2, F2R, DEFB126, CXCL6, and CXCL3 genes, was constructed for risk stratification. Patients in the low-risk group, which was characterized by elevated tumor mutation burden (TMB) have higher survival rate. The risk level was remarkably correlated with tumor-infiltrating immune cells (TIICs), the immune checkpoint molecule expression, and immunophenoscore (IPS). CXCL3 and CXCL6 were significantly upregulated in gastric cancer tissues compared with normal tissues using the UALCAN database and RT-qPCR. The nomogram showed good calibration and moderate discrimination in predicting overall survival (OS) at 1-, 3-, and 5- year for gastric cancer patients using risk-level and clinical characteristics. CONCLUSION: Our findings provided a risk stratification and prognosis prediction tool for gastric cancer patients and further the research into immunotherapy in gastric cancer.

Clinical plasma cells-related genes to aid therapy in colon cancer.

The tumor immune microenvironment (TIME) of colon cancer (CC) has been associated with extensive immune cell infiltration (IMI). Increasing evidence demonstrated that plasma cells (PC) have an extremely important role in advance of antitumor immunity. Nonetheless, there is a lack of comprehensive analyses of PC infiltration in clinical prognosis and immunotherapy in CC. This study systematically addressed the gene expression model and clinical information of CC patients. Clinical samples were obtained from the TCGA (The Cancer Genome Atlas) databases. Gene ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), GSVA, and the MAlignant Tumors using Expression data (ESTIMATE) algorithm were employed to research the potential mechanism and pathways. Immunophenoscore (IPS) was obtained to evaluate the immunotherapeutic significance of risk score. Half maximal inhibitory concentration (IC50) of chemotherapeutic medicine was predicted by employing the pRRophetic algorithm. A total of 513 CC samples (including 472 tumor samples and 41 normal samples) were collected from the TCGA-GDC database. Significant black modules and 313 candidate genes were considered PC-related genes by accessing WGCNA. Five pivotal genes were established through multiple analyses, which revealed excellent prognostic. The underlying correlation between risk score with tumor mutation burden (TMB) was further explored. In addition, the risk score was obviously correlated with various tumor immune microenvironment (TIME). Also, risk CC samples showed various signaling pathways activity and different pivotal sensitivities to administering chemotherapy. Finally, the biological roles of the CD177 gene were uncovered in CC.

Hedgehog signaling-related genomics signature for the accurate progress and prognosis prediction in gastric cancer.

The Hedgehog pathway is thought to be closely associated with the progression of GC; however, a specific link between the Hedgehog pathway on the prognosis and immune infiltration of gastric cancer is still lacking. This study collected Hedgehog pathway-related genes. The Hedgehog pathway-related pattern were identified by consensus cluster analysis. Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), and gene set enrichment analysis (GSEA) were used to identify the biological functions which were significantly altered between predefined Cluster1 and Cluster2 in consensus clustering. The risk model of gastric cancer based on Hedgehog signaling pathway was constructed by univariate and multivariate COX regression, and the nomogram was constructed. The results showed that there were significant differences in the expression of Hedgehog pathway-related genes between the two groups. In addition, the constructed risk model was significantly correlated with the clinical prognosis and immune cell infiltration level of patients with gastric cancer. The model effectively predicted the efficacy of chemotherapy in GC patients and the sensitivity of drug treatment between groups. We systematically revealed the mechanism of Hedgehog pathway in gastric cancer and selected biomarkers with biological significance from a new perspective, providing potential direction for the treatment of gastric cancer.

Outcome-Supervised Deep Learning on Pathologic Whole Slide Images for Survival Prediction of Immunotherapy in Patients with Non-Small Cell Lung Cancer.

Although programmed death-(ligand) 1 (PD-(L)1) inhibitors are marked by durable efficacy in patients with non-small cell lung cancer (NSCLC), approximately 60% of the patients still suffer from recurrence and metastasis after PD-(L)1 inhibitor treatment. To accurately predict the response to PD-(L)1 inhibitors, we presented a deep learning model using a Vision Transformer (ViT) network based on hematoxylin and eosin (H&E)-stained specimens of patients with NSCLC. Two independent cohorts of patients with NSCLC receiving PD-(L)1 inhibitors from Shandong Cancer Hospital and Institute and Shandong Provincial Hospital were enrolled for model training and external validation, respectively. Whole slide images (WSIs) of H&E-stained histologic specimens were obtained from these patients and patched into 1024 × 1024 pixels. The patch-level model was trained based on ViT to identify the predictive patches, and patch-level probability distribution was performed. Then, we trained a patient-level survival model based on the ViT-Recursive Neural Network framework and externally validated it in the Shandong Provincial Hospital cohort. A total of 291 WSIs of H&E-stained histologic specimens from 198 patients with NSCLC in Shandong Cancer Hospital and 62 WSIs from 30 patients with NSCLC in Shandong Provincial Hospital were included in the model training and validation. The model achieved an accuracy of 88.6% in the internal validation cohort and 81% in the external validation cohort. The survival model also remained a statistically independent predictor of survival from PD-(L)1 inhibitors. In conclusion, the outcome-supervised ViT-Recursive Neural Network survival model based on pathologic WSIs could be used to predict immunotherapy efficacy in patients with NSCLC.

Prognosis prediction in esophageal signet-ring-cell carcinoma: a competing risk analysis.

OBJECTIVE: This study aims to construct and validate a competing risk nomogram model to predict 1-year, 3-year, and 5-year cancer-specific survival (CSS) for patients with esophageal signet-ring-cell carcinoma. METHODS: Patients diagnosed with esophageal signet-ring-cell carcinoma (ESRCC) between 2010 and 2015 were abstracted from the Surveillance, Epidemiology, and End Results (SEER) database. We performed the competing risk model to select significant variables to build a competing risk nomogram, which was used to estimate 1-year, 3-year, and 5-year CSS probability. The C-index, receiver operating characteristic (ROC) curve, calibration plot, Brier score, and decision curve analysis were performed in the internal validation. RESULTS: A total of 564 patients with esophageal signet-ring-cell carcinoma fulfilled the eligibility criteria. The competing risk nomogram identified 4 prognostic variables, involving the gender, lung metastases, liver metastases, and receiving surgery. The C indexes of nomogram were 0.61, 0.75, and 0.70, respectively for 5-year, 3-year, and 1-year CSS prediction. The calibration plots displayed high consistency. The Brier scores and decision curve analysis respectively favored good prediction ability and clinical utility of the nomogram. CONCLUSIONS: A competing risk nomogram for esophageal signet-ring-cell carcinoma was successfully constructed and internally validated. This model is expected to predict 1-year, 3-year, and 5-year CSS, and help oncologists and pathologists in clinical decision making and health care management for esophageal signet-ring-cell carcinoma patients.

The clinic-based predictive modeling for prognosis of patients with cryptococcal meningitis.

BACKGROUND: Cryptococcal meningitis (CM) is the most common fungal infection of the central nervous system that can cause significant morbidity and mortality. Although several prognostic factors have been identified, their clinical efficacy and use in combination to predict outcomes in immunocompetent patients with CM are not clear. Therefore, we aimed to determine the utility of those prognostic factors alone or in combination in predicting outcomes of immunocompetent patients with CM. METHODS: The demographic and clinical data of patients with CM were collected and analyzed. The clinical outcome was graded by the Glasgow outcome scale (GOS) at discharge, and patients were divided into good (score of 5) and unfavorable (score of 1-4) outcome groups. Prognostic model was created and receiver-operating characteristic curve analyses were conducted. RESULTS: A total of 156 patients were included in our study. Patients with higher age at onset (p = 0.021), ventriculoperitoneal shunt placement (p = 0.010), Glasgow Coma Scale (GCS) score of less than 15(p< 0.001), lower CSF glucose concentration (p = 0.037) and immunocompromised condition (p = 0.002) tended to have worse outcomes. Logistic regression analysis was used to create a combined score which had a higher AUC (0.815) than those factors used alone for predicting outcome. CONCLUSIONS: Our study shows that a prediction model based on clinical characteristics had satisfactory accuracy in prognostic prediction. Early recognition of CM patients at risk of poor prognosis using this model would be helpful in providing timely management and therapy to improve outcomes and to identify individuals who warrant early follow-up and intervention.

Short-term prognostic models for severe acute kidney injury patients receiving prolonged intermittent renal replacement therapy based on machine learning.

BACKGROUND: As an effective measurement for severe acute kidney injury (AKI), the prolonged intermittent renal replacement therapy (PIRRT) received attention. Also, machine learning has advanced and been applied to medicine. This study aimed to establish short-term prognosis prediction models for severe AKI patients who received PIRRT by machine learning. METHODS: The hospitalized AKI patients who received PIRRT were assigned to this retrospective case-control study. They were grouped based on survival situation and renal recovery status. To screen the correlation, Pearson's correlation coefficient, partial ETA square, and chi-square test were applied, eight machine learning models were used for training. RESULTS: Among 493 subjects, the mortality rate was 51.93% and the kidney recovery rate was 30.43% at 30 days post-discharge, respectively. The indices related to survival were Sodium, Total protein, Lactate dehydrogenase (LDH), Phosphorus, Thrombin time, Liver cirrhosis, chronic kidney disease stage, number of vital organ injuries, and AKI stage, while Sodium, Total protein, LDH, Phosphorus, Thrombin time, Diabetes, peripherally inserted central catheter and AKI stage were selected to predict the 30-day renal recovery. Naive Bayes has a good performance in the prediction model for survival, Random Forest has a good performance in 30-day renal recovery prediction model, while for 90-day renal recovery prediction model, it's K-Nearest Neighbor. CONCLUSIONS: Machine learning can not only screen out indicators influencing prognosis of AKI patients receiving PIRRT, but also establish prediction models to optimize the risk assessment of these people. Moreover, attention should be paid to serum electrolytes to improve prognosis.

Machine Learning Quantified Tumor-Stroma Ratio Is an Independent Prognosticator in Muscle-Invasive Bladder Cancer.

Although the tumor-stroma ratio (TSR) has prognostic value in many cancers, the traditional semi-quantitative visual assessment method has inter-observer variability, making it impossible for clinical practice. We aimed to develop a machine learning (ML) algorithm for accurately quantifying TSR in hematoxylin-and-eosin (H&E)-stained whole slide images (WSI) and further investigate its prognostic effect in patients with muscle-invasive bladder cancer (MIBC). We used an optimal cell classifier previously built based on QuPath open-source software and ML algorithm for quantitative calculation of TSR. We retrospectively analyzed data from two independent cohorts to verify the prognostic significance of ML-based TSR in MIBC patients. WSIs from 133 MIBC patients were used as the discovery set to identify the optimal association of TSR with patient survival outcomes. Furthermore, we performed validation in an independent external cohort consisting of 261 MIBC patients. We demonstrated a significant prognostic association of ML-based TSR with survival outcomes in MIBC patients (p < 0.001 for all comparisons), with higher TSR associated with better prognosis. Uni- and multivariate Cox regression analyses showed that TSR was independently associated with overall survival (p < 0.001 for all analyses) after adjusting for clinicopathological factors including age, gender, and pathologic stage. TSR was found to be a strong prognostic factor that was not redundant with the existing staging system in different subgroup analyses (p < 0.05 for all analyses). Finally, the expression of six genes (DACH1, DEEND2A, NOTCH4, DTWD1, TAF6L, and MARCHF5) were significantly associated with TSR, revealing possible potential biological relevance. In conclusion, we developed an ML algorithm based on WSIs of MIBC patients to accurately quantify TSR and demonstrated its prognostic validity for MIBC patients in two independent cohorts. This objective quantitative method allows application in clinical practice while reducing the workload of pathologists. Thus, it might be of significant aid in promoting precise pathology services in MIBC.