Impact of aspirin use on rates of metastasis in patients with esophageal cancer: insights from the National Inpatient Sample.

Despite advancing treatment methods, esophageal cancer (EC) maintains a high mortality rate and poor prognosis. Through various mechanisms, aspirin has been suggested to have a chemopreventive effect on EC. However, the long-term impact, particularly regarding the rate of metastasis, needs to be further elucidated. NIS 2016-2020 was used to identify adult patients (age > 18 years) with EC using ICD-10 codes. Patients with missing demographics and mortality were excluded. Patients were stratified into two groups based on aspirin use. Data were collected on patient demographics, Elixhauser Comorbidity Index (ECI), and comorbidities (hypertension, chronic pulmonary disease, coronary artery disease (CAD), chronic kidney disease (CKD), congestive heart failure (CHF), coagulopathy, alcohol use, smoking, and obesity). The outcomes studied were rates of total metastasis, gastrointestinal (GI) metastasis, non-GI metastasis, and lymphoid metastasis. Multivariate logistic regression analysis was performed to evaluate the impact of aspirin use on various metastases after adjusting for patient demographics, comorbidities, and ECI. Out of 190,655 patients, 20,650 (10.8%) patients were aspirin users. Majority of the patients in the aspirin group were aged > 65 years (74.7%), males (82.1%), White race (84%), and had medicare insurance (71%). There was a higher incidence of diabetes, hypertension, chronic pulmonary disease, CAD, CKD, CHF, and smoking in aspirin users than non-aspirin users. Patients with aspirin users had a lower incidence of metastasis (28.9% vs. 38.7%, P < 0.001), GI metastasis (14.2% vs. 20.6%, P < 0.001), non-GI metastasis (15.1% vs. 22%, P < 0.001), and lymphoid metastasis (8.9% vs. 11.3%, P < 0.001) than non-aspirin users. After adjusting for confounding factors, patients with aspirin use had lower odds of having metastasis (aOR-0.73, 95% CI-0.70-0.77, P < 0.001). Our study noted that aspirin use is associated with a reduction in the rate of metastasis in patients with EC. These studies support the use of aspirin in patients with EC and suggest the need for further studies to understand the mechanism by which aspirin use reduces metastasis in patients with EC.

Cluster-based histopathology phenotype representation learning by self-supervised multi-class-token hierarchical ViT.

Developing a clinical AI model necessitates a significant amount of highly curated and carefully annotated dataset by multiple medical experts, which results in increased development time and costs. Self-supervised learning (SSL) is a method that enables AI models to leverage unlabelled data to acquire domain-specific background knowledge that can enhance their performance on various downstream tasks. In this work, we introduce CypherViT, a cluster-based histo-pathology phenotype representation learning by self-supervised multi-class-token hierarchical Vision Transformer (ViT). CypherViT is a novel backbone that can be integrated into a SSL pipeline, accommodating both coarse and fine-grained feature learning for histopathological images via a hierarchical feature agglomerative attention module with multiple classification (cls) tokens in ViT. Our qualitative analysis showcases that our approach successfully learns semantically meaningful regions of interest that align with morphological phenotypes. To validate the model, we utilize the DINO self-supervised learning (SSL) framework to train CypherViT on a substantial dataset of unlabeled breast cancer histopathological images. This trained model proves to be a generalizable and robust feature extractor for colorectal cancer images. Notably, our model demonstrates promising performance in patch-level tissue phenotyping tasks across four public datasets. The results from our quantitative experiments highlight significant advantages over existing state-of-the-art SSL models and traditional transfer learning methods, such as those relying on ImageNet pre-training.

Predictors and impact of aspiration pneumonia in patients undergoing esophagogastroduodenoscopy: national inpatient sample 2016-2020.

OBJECTIVES: Aspiration pneumonia is a rare but feared complication among patients undergoing esophagogastroduodenoscopy (EGD). Our study aims to assess the incidence as well as risk factors for aspiration pneumonia in patients undergoing EGD. METHODS: National Inpatient Sample 2016-2020 was used to identify adult patients undergoing EGD. Patients were stratified into two groups based on the presence of aspiration pneumonia. Multivariate logistic regression analysis was performed to identify the risk factors associated with aspiration pneumonia. We adjusted for patient demographics, Elixhauser comorbidities and hospital characteristics. RESULTS: Of the 1.8 million patients undergoing EGD, 1.9% of the patients developed aspiration pneumonia. Patients with aspiration pneumonia were mostly males (59.54%), aged >65 years old (66.19%), White (72.2%), had Medicare insurance (70.5%) and were in the lowest income quartile (28.7%). On multivariate analysis, the age >65 group, White race, congestive heart failure (CHF), neurological disorders and chronic obstructive pulmonary disease were associated with higher odds of aspiration pneumonia. This complication was associated with higher in-hospital mortality (9% vs. 0.8%; P  < 0.001) and longer length of stay (10.54 days vs. 4.85 days; P  < 0.001). CONCLUSION: Our study found that rates of post-EGD aspiration pneumonia are increasing. We found a significant association between various comorbidities and aspiration pneumonia. Our data suggests that we need to optimize these patients before EGD, as the development of aspiration is associated with worsened outcomes. Further prospective studies are needed to clarify these associations.

Vaccine-Preventable Illness Leads to Adverse Outcomes in Liver Transplant Recipients.

BACKGROUND: Liver transplant recipients (LTR) and patients with chronic liver disease (CLD) are at an increased risk of infections. AIMS: The objective of our study was to assess the incidence, and impact of vaccine preventable illness (VPI) on outcomes in LTR. METHODS: National Inpatient Sample (NIS) 2016-2020 was used to identify adults (age > 18) hospitalized LTR using ICD-10 codes. Data were collected on patient demographics, hospital characteristics, etiology of liver disease, hepatic decompensations and outcomes. Patients were stratified into two groups based on the presence or absence of VPI. Multivariate logistic regression analysis was performed to identify the association between VPI and outcomes. RESULTS: Out of 170,650 hospitalized LTR, 13.5% of the patients had VPI. The most common VPI was noted to be influenza (10.7%), followed by pneumococcal infection (2.7%). Incidence of mortality (6.9% vs. 1.6%, p < 0.001), ICU admissions (14.3% vs. 3.4%, p < 0.001), and acute kidney injury (AKI) (43.7% vs 37.35%, p < 0.001) was higher in the VPI group. CONCLUSION: More than 13% of the LT hospitalizations had concomitant VPI. VPI in LTR was associated with worse outcomes. Our data suggests the need to identify factors associated with reduced vaccination rates and identify strategies to improve vaccination rates and responses in these patients.

Tofacitinib in Acute Severe Ulcerative Colitis (TACOS): A Randomized Controlled Trial.

INTRODUCTION: Intravenous corticosteroids are the mainstay of treatment of patients hospitalized with acute severe ulcerative colitis (ASUC). However, 30%-40% of the patients are refractory to corticosteroids. We investigated whether addition of tofacitinib to corticosteroids improved the treatment responsiveness in patients with ASUC. METHODS: This single-center, double-blind, placebo-controlled trial randomized adult patients with ASUC (defined by the Truelove Witts severity criteria) to receive either tofacitinib (10 mg thrice daily) or a matching placebo for 7 days while continuing intravenous corticosteroids (hydrocortisone 100 mg every 6 hours). The primary end point was response to treatment (decline in the Lichtiger index by >3 points and an absolute score <10 for 2 consecutive days without the need for rescue therapy) by day 7. The key secondary outcome was the cumulative probability of requiring initiation of infliximab or undergoing colectomy within 90 days following randomization. All analyses were performed in the intention-to-treat population. RESULTS: A total of 104 patients were randomly assigned to a treatment group (53 to tofacitinib and 51 to placebo). At day 7, response to treatment was achieved in 44/53 (83.01%) patients receiving tofacitinib vs 30/51 (58.82%) patients receiving placebo (odds ratio 3.42, 95% confidence interval 1.37-8.48, P = 0.007). The need for rescue therapy by day 7 was lower in the tofacitinib arm (odds ratio 0.27, 95% confidence interval 0.09-0.78, P = 0.01). The cumulative probability of need for rescue therapy at day 90 was 0.13 in patients who received tofacitinib vs 0.38 in patients receiving placebo (log-rank P = 0.003). Most of the treatment-related adverse effects were mild. One patient, receiving tofacitinib, developed dural venous sinus thrombosis. DISCUSSION: In patients with ASUC, combination of tofacitinib and corticosteroids improved treatment responsiveness and decreased the need for rescue therapy.

Inflammatory Bowel Disease and X (Formerly Twitter) Influencers: Who Are They and What Do They Say?

Background and objective More than half of the population suffering from inflammatory bowel disease (IBD) use the internet as a primary source of information on their condition. X (formerly Twitter) has been increasingly used to disseminate healthcare-related information. In this study, we aimed to identify top influencers on the topic of IBD on X and correlate the relevance of their social media engagements with their professional expertise or academic productivity. Methods X (formerly Twitter) influence scores for the search topic IBD were obtained using Cronycle API, a proprietary software employing multiple algorithms to rank influencers. Data regarding gender, profession, location, and research productivity represented as h-index was collected. Results We collected information on the top 100 IBD influencers on X. The majority of influencers were gastroenterologists, followed by IBD advocates. Of note, 62% of the IBD influencers were from the US followed by the UK and Canada. A positive correlation was observed between the X topic score and the h-index of the influencer (r=+0.488, p<0.001) Conclusions The strong correlation observed between the X topic score and h-index suggests that social media is a viable platform for gaining information regarding IBD. Further research aimed at counteracting misleading information by providing facts and data in a succinct manner about IBD on social media is required to improve disease awareness.

The Effects of Malnutrition on Inpatient Outcomes in Patients With Gastroparesis: A Nationwide Analysis.

Introduction Gastroparesis (GP) is a chronic debilitating gastric motility disorder defined as delayed emptying of the stomach content without mechanical obstruction. It can result in nutritional deficiencies, leading to poor overall outcomes. We assessed the impact of malnutrition on in-hospital outcomes in patients with gastroparesis. Methods Patients with a primary discharge diagnosis of GP between January 2016 and December 2019 were included in the National Inpatient Sample (NIS) database. Data on patient demographics, hospital characteristics, the Charlson Comorbidity Index (CCI), and the etiology of gastroparesis were collected. The association between malnutrition and outcomes, including mortality, deep vein thrombosis (DVT), pulmonary embolism (PE), sepsis, acute kidney injury (AKI), length of stay (LOS), and total hospitalization charges (THC), were analyzed using the multivariate regression model. Results A total of 182,580 patients with gastroparesis were included in the analysis. Patients with gastroparesis and malnutrition had a higher risk of mortality (adjusted odds ratio {aOR}, 3.29; p<0.001), sepsis (aOR, 0.43; p<0.001), DVT (aOR, 2.34; p<0.001), and PE (aOR, 2.68; p<0.001) compared to patients with gastroparesis without malnutrition. No significant difference was noted in the rates of AKI. Patients with malnutrition also had a prolonged LOS (2.96 days; p<0.001) and higher THC ($22,890; p<0.001) compared to patients without malnutrition. Conclusion Gastroparesis patients with malnutrition are at a greater risk of worse outcomes than those without malnutrition. The early identification of malnutrition in gastroparesis patients can predict morbidity and mortality and assist in risk stratification to enhance outcomes. Further studies are encouraged to identify factors associated with malnutrition in gastroparesis and the impact of interventions to prevent and treat malnutrition.

Hyponatremia and Outcome: Is Severity More Important Than Etiology?

BACKGROUND AND OBJECTIVE: Hyponatremia is the most common electrolyte abnormality encountered in a hospital setting, and the data regarding the contribution of hyponatremia to overall mortality are conflicting. The study objective was to determine patients' clinical profiles and outcomes with hyponatremia. METHODS:  This prospective cross-sectional study was conducted at Dayanand Medical College and Hospital, Ludhiana, and included 375 adult patients aged more than 18 years with a confirmed diagnosis of hyponatremia. Patients were subdivided into three groups based on the severity of hyponatremia: mild (130-135 mmol/L), moderate (125-129 mmol/L), and profound (<125 mmol/L). RESULTS: The most common symptom was confusion (57.3%) followed by deep somnolence (40%) and nausea (36.8%). The most common cause of hyponatremia was diuretics (30.7%), followed by the syndrome of inappropriate antidiuretic hormone secretion (SIADH) (17.8%) and chronic liver disease (CLD) (14.1%). The severity of hyponatremia did not significantly influence the outcome. Patients with CLD and chronic kidney disease (CKD) as the etiology of hyponatremia had significantly worse outcomes compared to other causes of hyponatremia. The most common type was hypovolemic hypotonic followed by euvolemic hypotonic and hypervolemia hypotonic hyponatremia. Nearly half of the total deaths were observed in the hypervolemic hyponatremia group and were significantly higher compared to the other two groups (p=0.001). Correction of hyponatremia (i.e., serum sodium >135 mmol/L) was significantly linked with good outcomes (p=0.003). CONCLUSION: Our study showed that the etiology of hyponatremia was a more important prognostic indicator rather than the severity of hyponatremia. Normalization of serum sodium was associated with improved survival.

Prevalence and Impact of Malnutrition in Hospitalizations Among Celiac Diseases: A Nationwide Analysis.

BACKGROUND/AIMS: Celiac disease (CD) is a T-cell-mediated gluten sensitivity that results in villous atrophy in the small intestine, leading to chronic malabsorption. Patients with celiac disease are prone to malnutrition. We assessed the impact of malnutrition on in-hospital outcomes in patients with CD. MATERIALS AND METHODS: Patients with a primary discharge diagnosis of CD between January 2016 and December 2019 were included in the National Inpatient Sample Database. Data were collected on patient demographics, hospital characteristics, the Charlson Comorbidity Index (CCI), and concomitant comorbidities. The association between malnutrition and outcomes, including mortality, deep vein thrombosis (DVT), pulmonary embolism (PE), sepsis, acute kidney injury (AKI), length of stay (LOS), and total hospitalization charges (THC), was analyzed using the multivariate regression model. RESULTS: A total of 187310 patients with CD were included in the analysis. Patients with CD and malnutrition had a higher risk of mortality (adjusted odds ratio [aOR], 2.08; p<0.001), AKI (aOR=1.18, p=0.003), and DVT (aOR=1.53; p<0.001) compared to patients with CD without malnutrition. No significant difference was noted in the rates of sepsis and PE. Patients with malnutrition also had a prolonged LOS (2.89 days; p<0.001) and higher THC ($22252.18; p<0.001) compared to patients without malnutrition. DISCUSSION: Patients with CD and malnutrition are at high risk of worse outcomes. Early identification of malnutrition in CD can help prevent morbidity and mortality. Even strict adherence to a gluten-free diet has been associated with malnutrition. Further studies identifying factors associated with malnutrition in CD and the impact of interventions to prevent and treat malnutrition are encouraged.

The Impact of Type 2 Myocardial Infarction in Acute Pancreatitis: Analysis of 1.1 Million Hospitalizations and Review of the Literature.

Introduction  Acute pancreatitis (AP) is a common inflammatory disorder with acute onset and rapid progression. Studies have reported cardiac injury in patients with AP. It is often thought that stress cardiomyopathy can induce these changes leading to type 2 myocardial infarction (type 2 MI) in AP. Our study aims to assess the prevalence as well as the impact of type 2 MI on outcomes in patients with AP.  Methods National Inpatient Sample (NIS) 2016-2020 was used to identify adult patients (age>18) with acute pancreatitis. We excluded patients with STEMI, NSTEMI, pancreatic cancer, or chronic pancreatitis. Patients with missing demographics and mortality were also excluded. Patients were stratified into two groups, based on the presence of type 2 MI. Multivariate logistic regression analysis was performed to assess the impact of concomitant type 2 MI on mortality, sepsis, acute kidney injury (AKI), ICU admission, deep venous thrombosis (DVT), and pulmonary embolism (PE) after adjusting for patient demographics, hospital characteristics, etiology of AP and the Elixhauser comorbidities.  Results Of the 1.1 million patients in the study population, only 2315 patients had type 2 MI. The majority of the patients in the type 2 MI group were aged >65 years (49.2%, p<0.001), males (54.6%, p=0.63), White (67.6%, p=0.19), had Medicare insurance (55.5%, p<0.001), and were in the lowest income quartile (34.8%, p=0.12). Patients in the type 2 MI group had a higher incidence of mortality (5.4% vs 0.6%, p<0.001), sepsis (7.1% vs 3.7%, p<0.001), shock (9.3% vs 0.9%, p<0.001), AKI (42.9% vs. 11.8%, p<0.001) and ICU admission (12.1% vs 1.4%, p<0.001). After adjusting for confounding factors, patients in the type 2 MI group were noted to be at higher odds of mortality (aOR=2.4; 95% CI 1.5-3.8, p<0.001). Patients in the type 2 MI group had a longer length of stay (adjusted coefficient=2.1 days; 95% CI 1.4-2.8; p<0.001) and higher total hospitalization charges (adjusted coefficient=$45,088; 95% CI $30,224-$59,952; p<0.001).  Conclusion Although the prevalence of type 2 MI in AP is low, the presence of type 2 MI is associated with increased mortality and worse outcomes. Physicians should be aware of this association and these patients should be monitored carefully to prevent worse outcomes.

Ectopic Liver Tissue Encountered Incidentally During a Laparoscopic Cholecystectomy: A Case Report.

A 60-year-old male patient who presented with right upper quadrant (RUQ) pain was diagnosed with acute cholecystitis after an ultrasound of the abdomen revealed multiple gallstones, gallbladder wall thickening, pericholecystic fluid, and a positive sonographic Murphy sign. The patient was admitted, administered IV fluids, antibiotics, and pain relief, and scheduled for laparoscopic cholecystectomy. During surgery, an incidental finding of ectopic liver tissue attached to the gallbladder was noted. Histopathology confirmed the presence of chronic cholecystitis and multifaceted cholesterol stones. Normal liver tissue was noted in the ectopic mass. Ectopic liver tissue is defined as liver tissue located outside the main liver parenchyma and is usually asymptomatic. They are usually detected at the time of autopsies, incidentally during surgeries, or during imaging done for other etiologies. They can occur at various sites in the body. Ectopic liver tissue can cause potential complications such as hepatocellular carcinoma and torsion, and in the event that they are incidentally detected, it is advised to remove them. The case report highlights the importance of dealing with incidental findings during laparoscopic cholecystectomy and creating awareness about it.

Learning binary and sparse permutation-invariant representations for fast and memory efficient whole slide image search.

Considering their gigapixel sizes, the representation of whole slide images (WSIs) for classification and retrieval systems is a non-trivial task. Patch processing and multi-Instance Learning (MIL) are common approaches to analyze WSIs. However, in end-to-end training, these methods require high GPU memory consumption due to the simultaneous processing of multiple sets of patches. Furthermore, compact WSI representations through binary and/or sparse representations are urgently needed for real-time image retrieval within large medical archives. To address these challenges, we propose a novel framework for learning compact WSI representations utilizing deep conditional generative modeling and the Fisher Vector Theory. The training of our method is instance-based, achieving better memory and computational efficiency during the training. To achieve efficient large-scale WSI search, we introduce new loss functions, namely gradient sparsity and gradient quantization losses, for learning sparse and binary permutation-invariant WSI representations called Conditioned Sparse Fisher Vector (C-Deep-SFV), and Conditioned Binary Fisher Vector (C-Deep-BFV). The learned WSI representations are validated on the largest public WSI archive, The Cancer Genomic Atlas (TCGA) and also Liver-Kidney-Stomach (LKS) dataset. For WSI search, the proposed method outperforms Yottixel and Gaussian Mixture Model (GMM)-based Fisher Vector both in terms of retrieval accuracy and speed. For WSI classification, we achieve competitive performance against state-of-art on lung cancer data from TCGA and the public benchmark LKS dataset.

Decentralized federated learning through proxy model sharing.

Institutions in highly regulated domains such as finance and healthcare often have restrictive rules around data sharing. Federated learning is a distributed learning framework that enables multi-institutional collaborations on decentralized data with improved protection for each collaborator's data privacy. In this paper, we propose a communication-efficient scheme for decentralized federated learning called ProxyFL, or proxy-based federated learning. Each participant in ProxyFL maintains two models, a private model, and a publicly shared proxy model designed to protect the participant's privacy. Proxy models allow efficient information exchange among participants without the need of a centralized server. The proposed method eliminates a significant limitation of canonical federated learning by allowing model heterogeneity; each participant can have a private model with any architecture. Furthermore, our protocol for communication by proxy leads to stronger privacy guarantees using differential privacy analysis. Experiments on popular image datasets, and a cancer diagnostic problem using high-quality gigapixel histology whole slide images, show that ProxyFL can outperform existing alternatives with much less communication overhead and stronger privacy.

Low prevalence of primary sclerosing cholangitis in patients with inflammatory bowel disease in India.

BACKGROUND/AIMS: Primary sclerosing cholangitis (PSC) represents the most common hepatobiliary extraintestinal manifestation of inflammatory bowel disease (IBD), including ulcerative colitis (UC) and Crohn's disease (CD). Limited data exist on PSC in patients with IBD from India. We aimed to assess the prevalence and disease spectrum of PSC in Indian patients with IBD. METHODS: Database of IBD patients at 5 tertiary care IBD centers in India were analyzed retrospectively. Data were extracted and the prevalence of PSC-IBD was calculated. RESULTS: Forty-eight patients out of 12,216 patients with IBD (9,231 UC, 2,939 CD, and 46 IBD unclassified) were identified to have PSC, resulting in a prevalence of 0.39%. The UC to CD ratio was 7:1. Male sex and pancolitis (UC) or colonic CD were more commonly associated with PSC-IBD. The diagnosis of IBD preceded the diagnosis of PSC in most of the patients. Majority of the patients were symptomatic for liver disease at diagnosis. Eight patients (16.66%) developed cirrhosis, 5 patients (10.41%), all UC, developed malignancies (3 colorectal cancer [6.25%] and 2 cholangiocarcinoma [4.16%]), and 3 patients died (2 decompensated liver disease [4.16%] and 1 cholangiocarcinoma [2.08%]) on follow-up. None of the patients mandated surgical therapy for IBD. CONCLUSIONS: Concomitant PSC in patients with IBD is uncommon in India and is associated with lower rates of development of malignancies.

Evolutionary deep feature selection for compact representation of gigapixel images in digital pathology.

Despite the recent progress in Deep Neural Networks (DNNs) to characterize histopathology images, compactly representing a gigapixel whole-slide image (WSI) via salient features to enable computational pathology is still an urgent need and a significant challenge. In this paper, we propose a novel WSI characterization approach to represent, search and classify biopsy specimens using a compact feature vector (CFV) extracted from a multitude of deep feature vectors. Since the non-optimal design and training of deep networks may result in many irrelevant and redundant features and also cause computational bottlenecks, we proposed a low-cost stochastic method to optimize the output of pre-trained deep networks using evolutionary algorithms to generate a very small set of features to accurately represent each tissue/biopsy. The performance of the proposed method has been assessed using WSIs from the publicly available TCGA image data. In addition to acquiring a very compact representation (i.e., 11,000 times smaller than the initial set of features), the optimized features achieved 93% classification accuracy resulting in 11% improvement compared to the published benchmarks. The experimental results reveal that the proposed method can reliably select salient features of the biopsy sample. Furthermore, the proposed approach holds the potential to immensely facilitate the adoption of digital pathology by enabling a new generation of WSI representation for efficient storage and more user-friendly visualization.

Federated learning and differential privacy for medical image analysis.

The artificial intelligence revolution has been spurred forward by the availability of large-scale datasets. In contrast, the paucity of large-scale medical datasets hinders the application of machine learning in healthcare. The lack of publicly available multi-centric and diverse datasets mainly stems from confidentiality and privacy concerns around sharing medical data. To demonstrate a feasible path forward in medical image imaging, we conduct a case study of applying a differentially private federated learning framework for analysis of histopathology images, the largest and perhaps most complex medical images. We study the effects of IID and non-IID distributions along with the number of healthcare providers, i.e., hospitals and clinics, and the individual dataset sizes, using The Cancer Genome Atlas (TCGA) dataset, a public repository, to simulate a distributed environment. We empirically compare the performance of private, distributed training to conventional training and demonstrate that distributed training can achieve similar performance with strong privacy guarantees. We also study the effect of different source domains for histopathology images by evaluating the performance using external validation. Our work indicates that differentially private federated learning is a viable and reliable framework for the collaborative development of machine learning models in medical image analysis.

Social media data analysis to predict mental state of users using machine learning techniques.

BACKGROUND: Social media platforms such as Facebook, WhatsApp, and Instagram etc., are becoming very popular now not only for youth but for all walks of life. People are more often seen in busy in tweeting, chatting, or putting selfies. No one actually knows the mental state of a person in the online platform. In this article, we will be focusing on how social media is affecting issues such as road accident, murder, and suicide. The research is done by three parts. MATERIALS AND METHODS: Google Form analysis, machine learning used for prediction, and by sentimental analysis of what people think in twitter. All the datasets are based in India. From these datasets, the different machine learning algorithm is used to do the analysis. The project strives to bring the real-world solution in the matter of advancement. RESULTS: The static data analysis and dynamic data analysis shows the various sentimental analysis and predictions and the technique to predict different mental states. Thus we get clearly about the current world is getting into social issues. This research findings helps to bring social awareness among the current generation by understanding the sensitivity of the youths. CONCLUSION: Thus through this paper we get known clearly how the current world is getting into social issues like victim of murders or road accidents or committing suicide. The paper clearly helps us to understand the sensitivity of the youths. Therefore brings a social awareness among the current generation.

Selection, Visualization, and Interpretation of Deep Features in Lung Adenocarcinoma and Squamous Cell Carcinoma.

Although deep learning networks applied to digital images have shown impressive results for many pathology-related tasks, their black-box approach and limitation in terms of interpretability are significant obstacles for their widespread clinical utility. This study investigates the visualization of deep features (DFs) to characterize two lung cancer subtypes, adenocarcinoma and squamous cell carcinoma. It demonstrates that a subset of DFs, called prominent DFs, can accurately distinguish these two cancer subtypes. Visualization of such individual DFs allows for a better understanding of histopathologic patterns at both the whole-slide and patch levels, and discrimination of these cancer types. These DFs were visualized at the whole slide image level through DF-specific heatmaps and at tissue patch level through the generation of activation maps. In addition, these prominent DFs can distinguish carcinomas of organs other than the lung. This framework may serve as a platform for evaluating the interpretability of any deep network for diagnostic decision making.

Fine-Tuning and training of densenet for histopathology image representation using TCGA diagnostic slides.

Feature vectors provided by pre-trained deep artificial neural networks have become a dominant source for image representation in recent literature. Their contribution to the performance of image analysis can be improved through fine-tuning. As an ultimate solution, one might even train a deep network from scratch with the domain-relevant images, a highly desirable option which is generally impeded in pathology by lack of labeled images and the computational expense. In this study, we propose a new network, namely KimiaNet, that employs the topology of the DenseNet with four dense blocks, fine-tuned and trained with histopathology images in different configurations. We used more than 240,000 image patches with 1000×1000 pixels acquired at 20× magnification through our proposed "high-cellularity mosaic" approach to enable the usage of weak labels of 7126 whole slide images of formalin-fixed paraffin-embedded human pathology samples publicly available through The Cancer Genome Atlas (TCGA) repository. We tested KimiaNet using three public datasets, namely TCGA, endometrial cancer images, and colorectal cancer images by evaluating the performance of search and classification when corresponding features of different networks are used for image representation. As well, we designed and trained multiple convolutional batch-normalized ReLU (CBR) networks. The results show that KimiaNet provides superior results compared to the original DenseNet and smaller CBR networks when used as feature extractor to represent histopathology images.