Search

Background Risk stratification tools for patients with advanced melanoma (AM) treated with immune checkpoint inhibitors (ICI) are lacking. We identified a new prognostic model associated with overall survival (OS). Patients and Methods A total of 318 treatment naïve patients with AM receiving ICI were collected from a multi-centre retrospective cohort study. LASSO Cox regression identified independent prognostic factors associated with OS. Model validation was carried out on 500 iterations of bootstrapped samples. Harrel’s C-index was calculated and internally validated to outline the model’s discriminatory performance. External validation was carried out in 142 advanced melanoma patients receiving ICI in later lines. Results High white blood cell count (WBC), high lactate dehydrogenase (LDH), low albumin, Eastern Cooperative Oncology Group (ECOG) performance status ≥1, and the presence of liver metastases were included in the model. Patients were parsed into 3 risk groups: favorable (0-1 factors) OS of 52.9 months, intermediate (2-3 factors) OS 13.0 months, and poor (≥4 factors) OS 2.7 months. The C-index of the model from the discovery cohort was 0.69. External validation in later-lines (N = 142) of therapy demonstrated a c-index of 0.65. Conclusions Liver metastases, low albumin, high LDH, high WBC, and ECOG≥1 can be combined into a prognostic model for AM patients treated with ICI.

One anastomosis gastric bypass (OAGB) is now one of the mainstream bariatric surgical procedures with proven safety and efficacy. However, data on the gastrointestinal quality of life following OAGB is lacking.This is a retrospective analysis of a prospectively collected database, performed at a single tertiary care teaching hospital from January 2016 until March 2021. All patients undertook the Gastrointestinal Quality of Life Index (GIQLI) questionnaire. GIQLI was correlated with various parameters. Principal component analysis (PCA) was used to assess the importance of each question in the questionnaire and devise a "Mini GIQLI score".A total of 60 patients were included. The %TWL at 3 and 5 years was 26.2 ± 11.9%, 31.7 ± 11 respectively. The mean weight regain was 5.6 ± 8.5 kg. The mean GIQLI score was 125 ± 13.1. The mean scores for questions pertaining to gastrointestinal, social, psychological, and physical domains were 3.49, 3.7, 3.45, and 3.27 respectively. Scree plot of principal component analysis showed that a new score ("Mini GIQLI") combining only 5 questions had good correlation with the overall GIQOL score (r = 0.842). The five questions related to anxiety, fatigue, feeling unwell, loss of endurance, and feeling unfit.Patients report a good score on GIQLI assessment following OAGB. The Mini GIQLI score is a quicker tool with good correlation to the full-length GIQLI score.

SummaryBackgroundThe global launch of ChatGPT on November 30, 2022 has sparked widespread public interest in large language models (LLMs), and interest in the medical community is growing. Indeed, recent preprints on medRxiv have examined ChatGPT and GPT-3 in the context of standardized exams, such as the United States Medical Licensing Examination. These studies demonstrate modest performance relative to national averages. In this work, we enhance OpenAI’s GPT-3 model through zero-shot learning, anticipating that it outperforms experienced neurosurgeons in written question-answer tasks for common clinical and surgical questions on vestibular schwannoma. We aimed to address LLM accountability by including in-text citations and references to the responses provided by GPT-3.MethodsThe analysis involved (i) creating a dataset through web scraping, (ii) developing a chat-based platform called neuroGPT-X, (iii) enlisting expert neurosurgeons across international centers to create and answer questions and evaluate responses, and (iv) analyzing the evaluation results on the management of vestibular schwannoma. The survey had a blinded and unblinded phase. In the blinded phase, a neurosurgeon with 30+ years of experience curated 15 questions regarding common clinical and surgical contexts of vestibular schwannoma. Then, four neurosurgeons, ChatGPT (January 30, 2023 model, akanaiveGPT), and a context-enriched GPT model independently provided their responses. Three experienced neurosurgeons blindly evaluated the responses for accuracy, coherence, relevance, thoroughness, speed, and overall rating. Then, all seven neurosurgeons were unblinded to all responses and provided their thoughts on the potential of expert LLMs in the clinical setting.FindingsBoth the naive and content-enriched GPT models provided faster responses to the standardized question set (p0.999) or performance across the aforementioned domains (p>0.999). Of interest, all expert surgeons expressed concerns about the reliability of GPT in accurately addressing the nuances and controversies surrounding the management of vestibular schwannoma. Further, we developed neuroGPT-X, a chat-based platform designed to provide point-of-care clinical support and mitigate limitations of human memory. neuroGPT-X incorporates features such as in-text citations and references to enable accurate, relevant, and reliable information in real-time.InterpretationA context-enriched GPT model provided non-inferior responses compared to experienced neurosurgeons in generating written responses to a complex neurosurgical problem for which evidence-based consensus for management is lacking. We show that context enrichment of LLMs is well-suited to transform clinical practice by providing subspecialty-level answers to clinical questions in an accountable manner.Research in ContextEvidence before this studyWe searched PubMed for “(vestibular schwannoma OR acoustic schwannoma) AND (GPT-3 OR Generative Pretrained Transformer OR large language model)” with no filters and identified no relevant articles. We then searched PubMed using the string “(subspecialty OR neurosurgery OR physician) AND (GPT-3 OR Generative Pretrained Transformer OR large language model) AND (fine-tuning OR context enrichment)” with no filters and identified three studies. One study noted that domain-specific knowledge enhanced pre-trained language models.Added value of this studyTo our knowledge, this is the first study to show the non-inferiority of a context-enriched LLM in a question-answer task on common clinical and surgical questions compared to experienced neurosurgeons worldwide, determined by their neurosurgical colleagues. Furthermore, we developed the first online platform incorporating an LLM, chat memory, in-text citations, and references regarding comprehensive vestibular schwannoma management. To assess the model’s performance, a neurosurgeon with 30+ years of experience managing patients with vestibular schwannoma curated 15 questions to the model, ChatGPT, and four international expert neurosurgeons. A separate, blinded group of three expert neurosurgeons assessed these answers for accuracy, coherence, relevance, thoroughness, speed, and overall rating. This study demonstrated the capability of context-enriched LLMs as point-of-care informational aids. Importantly, all expert surgeons raised questions regarding the nuances and role of human experience and intuition that GPT may not capture in generating opinions or recommendations.Implications of all the available evidenceThe present study, with its subspecialist-level performance and interpretable results, suggests that context-enriched LLMs show promise as a point-of-care medical resource. Evaluations from experienced neurosurgeons showed that a context-enriched GPT model was rated similarly to neurosurgeon responses across evaluation domains in this study. This work serves as a springboard for expanding this tool into more medical specialties, incorporating evidence-based clinical information, and developing expert-level dialogue surrounding LLMs in healthcare.

Authors: Ye-Jean Park, Abhinav Pillai, Jiawen Deng, Mehul Gupta, Edward Guo, Mike Paget, and Christopher Naugler Large language models (LLMs) – artificial intelligence (AI) systems based on deep learning algorithms– are capable of processing and summarizing vast volumes of text and subsequently generating human-like responses when prompted (1). Through 1) language model pretraining, where the model learns a large corpus of raw, unlabeled text in an unsupervised manner, and 2) fine-tuning, where the model learns to apply knowledge from one task to another particular task using labelled training data, LLMs are ultimately able to understand and create a countless number of text styles and syntaxes (1). Techniques such as transfer learning have enabled LLMs to expand their capabilities beyond pretraining and fine-tuning, and now include techniques such as zero-shot and few-shot learning, allowing them to perform new tasks with little to no additional training data (2, 3). Therefore, LLMs have broad applicability to systems ranging from natural language processors (NLPs) to language translators to virtual assistant chatbots. Several types of LLMs exist, including transformer models, such as Bidirectional Encoder Representations from Transformers (BERT) or Generative Pre-trained Transformer (GPT) (4). Such models are based on what is known as a transformer architecture, a type of neural network model considered the state-of-the-art system used for natural language processing tasks given its self-attention mechanism (used to process input sequences in parallel form and therefore learn context and relationships between words in the given sequences) (4). ChatGPT, an AI model created by Open AI lab in San Francisco that launched in November of 2022, is a variant of GPT and build with the GPT-3.5 large language model (5). As an LLM, ChatGPT is trained on an approximately 570 GB-sized dataset consisting of billions of words and text from multiple languages, including from CommonCrawl and WebText, internet-based book corpora, Wikipedia (8). ChatGPT also possesses a language model head along with a dialogue generation head, used for Reinforcement Learning from Human Feedback (RLHF) to understand and generate human-like text (8). Furthermore, ChatGPT performs “continual/incremental learning”, wherein the model can maintain a memory of previous input and prompts to subsequently improve the accuracy and relevance of its future responses with each iteration of text by strengthening its neural network (5). ChatGPT is also “auto-aggressive”, meaning it can predict and generate future words based on previous sequences presented to it in a uni-directional manner, whereas the comparator BERT has bi-directional architecture and therefore analyzes a word based on the word’s left and right context (4, 7-9). With all these characteristics, ChatGPT has recently emerged as a more promising candidate for original text and dialogue generation, and is a prevalent subject of ongoing discourse and research. Given its potential impact in the real-world context, GPT has been increasingly explored across many contexts and disciplines. For instance, researchers internationally are assessing the model’s performance on business, law, and medical school board exams, its ability to code in Python and JavaScript, and its capabilities of creating original literature and poems (10-13). One particular area of interest is the broader applicability of LLMs like GPT in the realm of biomedicine and healthcare. In fact, BioGPT was launched in late 2022 and was trained on millions of PubMed abstracts to be one of the first biomedical-specific versions of GPT and thereby address limitations of GPT variants in providing evidence-based, tailored responses in biomedical sciences (14). In a similar vein, GPT-4 was launched on March 15, 2023, and said to address several limitations of GPT-3 models (as evidenced by its ranking in some of the highest percentiles on standardized exams rather than in the lower percentiles, possessing greater accuracy across multiple languages, and GPT-4’s capacity to accept prompts consisting of images and diagrams, unlike GPT-3 (15)). Overall, LLMs – particularly GPT and its variants – appear to possess both promising but also concerning features if integrated into the healthcare setting. For instance, several authors have described LLMs’ potential in clinical decision support to provide further information for healthcare teams to make more informed decisions for patient care (16, 18). Others have highlighted the improved speed and efficiency of clinical practice by using programs like ChatGPT in composing patient charts or in creating efficient, standardized discharge notes (18). Furthermore, all of these could be further vetted by the healthcare team to avoid errors while saving time for healthcare providers to focus on delivering optimized patient care (19). Likewise, outside of the immediate clinical setting, individuals may use widely accessible LLMs like ChatGPT in asking their health-related questions. Thus, LLMs may not only have the potential to elevate patient care and optimize hospital administrative tasks, but also in empowering patients with healthcare knowledge that, ideally, is always verified with their healthcare teams. Despite this, concerning elements of using LLMs in healthcare may include the presence of bias and lack of transparency in the training dataset, along with ethical/legal drawbacks when LLMs provide inaccurate recommendations and potentially lead to harm to patients (10, 19, 20). Furthermore, there are also logistical concerns with integrating LLMs into the current system of EMRs or paper charts worldwide, given data privacy issues and security breaches within or between institutions (19, 20). The creation and dissemination of robust, evidence-based research, preferably conducted with a clear and standardized method, should help the scientific community further explore these concerns alongside the potential benefits of LLMs in a clinical context. There is currently a paucity of research around the specific research criteria, recommendations, and methods of evaluating the clinical utility and limitations of GPT models. Thus, the objectives of our scoping review are: Objectives: To elucidate the research landscape regarding the clinical utility of LLMs – with a particular focus on ChatGPT/GPT-3 and GPT-4 and provide a summary of its effectiveness, accuracy, and efficacy (i.e., in the realms of clinical decision-making support, operational efficiency, and patient communication) To explore the ethical, legal, socioeconomic, and other implications of integrating LLMs in the clinical setting To scope practical barriers to the implementation of LLMs in the clinical context (i.e. costs, EMRs integration, data sharing across institutions) and key factors of successful implementation of LLMs in healthcare settings To propose a standardized, more rigorous methodology of investigating and evaluating LLMs’ clinical utility for future research studies To recommend specific clinical contexts that LLMs should be further investigated and may have the highest potential for impacting patient care. References: Yang X, Chen A, PourNejatian N, Shin HC, Smith KE, Parisien C, Compas C, Martin C, Costa AB, Flores MG, Zhang Y, Magoc T, Harle CA, Lipori G, Mitchell DA, Hogan WR, Shenkman EA, Bian J, Wu Y. A large language model for electronic health records. NPJ Digit Med. 2022 Dec 26;5(1):194. doi: 10.1038/s41746-022-00742-2. Weiss, K., Khoshgoftaar, T.M. & Wang, D. A survey of transfer learning. J Big Data 3, 9 (2016). https://doi.org/10.1186/s40537-016-0043-6 Kojima T, Shane SG, Reid M, Matsuo Y, Iwasaw Y. Large Language Models are Zero-Shot Reasoners. arXiv. 2022 May 24: v4. https://doi.org/10.48550/arXiv.2205.11916 Vaswani A, Shazeer N, Parmar N, Uszkoreit J, Jones L, Gomez A, Kaiser L, Polosukhin I. Attention Is All You Need. arXiv. 2017 June 12;v1. https://doi.org/10.48550/arXiv.1706.03762 Open AI. https://openai.com/blog/chatgpt. Accessed March 17, 2023. Liu S, Wright AP, Patterson BL, Wanderer JP, Turer RW, Nelson SD, McCoy AB, Sittig DF, Wright A. Assessing the Value of ChatGPT for Clinical Decision Support Optimization. medRxiv [Preprint]. 2023 Feb 23:2023.02.21.23286254. doi: 10.1101/2023.02.21.23286254. Invgate. https://blog.invgate.com/gpt-3-vs-bert. Accessed March 17, 2023. Devlin J, Chang MW, Lee K, Toutanova K. BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. arXiv [Preprint]. 2018 Oct; v1. https://doi.org/10.48550/arXiv.1810.04805 Ji Z, Wei Q, Xu H. BERT-based Ranking for Biomedical Entity Normalization. AMIA Jt Summits Transl Sci Proc. 2020 May 30;2020:269-277. Sallam M. The Utility of ChatGPT as an Example of Large Language Models in Healthcare Education, Research and Practice: Systematic Review on the Future Perspectives and Potential Limitations. medRxiv [Pre-print]. 2023 Feb 21: doi: https://doi.org/10.1101/2023.02.19.23286155 Gilson A, Safranek CW, Huang T, Socrates V, Chi L, Taylor RA, Chartash D. How Does ChatGPT Perform on the United States Medical Licensing Examination? The Implications of Large Language Models for Medical Education and Knowledge Assessment. JMIR Med Educ. 2023 Feb 8;9:e45312. doi: 10.2196/45312. PMID: 36753318 Kelly, S. ChatGPT passes exams from law and business schools. https://www.cnn.com/2023/01/26/tech/chatgpt-passes-exams/index.html. Accessed March 17, 2023. Chaarani, J. ChatGPT wrote a poem about winter, but is it truly art? This Waterloo AI ethicist weighs in. https://www.cbc.ca/news/canada/kitchener-waterloo/chatgpt-ai-text-university-waterloo-maura-grossman-1.6703819#:~:text=ChatGPT%20is%20artificial%20intelligence%20chatbot,still%20spotlighting%20concerns%20around%20accuracy.. Accessed March 17, 2023. Luo R, Sun L, Xia Y, Qin T, Zhang S, Poon H, Liu T. BioGPT: Generative Pre-trained Transformer for Biomedical text Generation and Mining. arXiv [Pre-print]. 2023 Jan;v2. https://doi.org/10.48550/arXiv.2210.10341 Open AI. https://openai.com/research/gpt-4. Accessed March 17, 2023. Levine DM, Tuwani R, Kompa B, Varma A, Finlayson SG, Mehrotra A, Beam A. The Diagnostic and Triage Accuracy of the GPT-3 Artificial Intelligence Model. medRxiv [Preprint]. 2023 Feb 1:2023.01.30.23285067. doi: 10.1101/2023.01.30.23285067. Patel SB, Lam K. ChatGPT: the future of discharge summaries? Lancet Digit Health. 2023 Mar;5(3):e107-e108. doi: 10.1016/S2589-7500(23)00021-3. Stewart, Jonathon et al "Applications of Natural Language Processing at Emergency Department Triage: A Systematic Review." medRxiv (2022): 2022.12.20.22283735. Web. 21 Feb. 2023. Stokel-Walker C, Van Noorden R. What ChatGPT and generative AI mean for science. Nature. 2023 Feb;614(7947):214-216. doi: 10.1038/d41586-023-00340-6. Nov, Oded, Nina Singh, and Devin M. Mann. "Putting ChatGPT’s Medical Advice to the (Turing) Test." medRxiv (2023): 2023.01.23.23284735. Web. 22 Feb. 2023.

SARS-CoV-2 is a unique event, having emerged suddenly as a highly infectious viral pathogen for human populations. Previous phylogenetic analyses show its closest known evolutionary relative to be a virus detected in bats (RaTG13), with a common assumption that SARS-CoV-2 evolved from a zoonotic ancestor via recent genetic changes (likely in the Spike protein receptor-binding domain or RBD) that enabled it to infect humans. We used detailed phylogenetic analysis, ancestral sequence reconstruction, and in situ molecular dynamics simulations to examine the Spike-RBD’s functional evolution, finding that the common ancestral virus with RaTG13, dating to no later than 2013, possessed high binding affinity to the human ACE2 receptor. This suggests that SARS-CoV-2 likely possessed a latent capacity to bind to human cellular targets (though this may not have been sufficient for successful infection) and emphasizes the importance of expanding efforts to catalog and monitor viruses circulating in both human and non-human populations. Supplementary Information The online version contains supplementary material available at 10.1007/s00239-021-10034-0.

People in the 21st century are being raised in an Internet enabled world where social media has become an integral part of people’s daily routine, with communication just a click away. As per the latest survey, the number of individuals using social media is over 39.6 million worldwide, with the average user having 8.7 accounts on various networking sites. Social media provides an opportunity to connect with people and share data in the form of posts, text etc, with this package of pros and yet various individuals are trying to misuse it by spreading hatred towards a group, individuals, a topic or an activity. Due to which cyberbullying has come into play, affecting the psychological state of the person. Where prevention is much needed, for which many researchers have come together and established many such technologies and programs for automatically detecting the events of cyber bullying on social media and preventing them by analysing the pattern of the posted comments or images. Thus the purpose of this research is to track and monitor the threats using supervised machine learning and mining.

Prescription data are invaluable for healthcare research and intelligence, yet, extraction of these data is challenging as this information is intertwined in the unstructured and non-grammatical text in prescription images. Moreover, text extraction from images in itself is hard, particularly for handwritten text. While piecemeal solutions exist, they are either limited to a small set of entities of interest or have very low accuracy and are not scalable. In this paper, we present two algorithms: the C-Cube algorithm for digitization of computer-printed prescriptions and the 3-Step Filtering algorithm for handwritten prescriptions. While a brute-force approach would match every word that is received from an optical character reader (OCR) with all possible entries in the database, this approach is inefficient and imprecise. The premise of our algorithms is an application of pattern intelligence to select a much smaller set of words (from the words returned by the OCR) as potential entities of interest. We rigorously tested the two algorithms on a corpus of more than 10,000 prescriptions’ images, taking the brute-force technique as the baseline methodology. Regarding latencies, we found that the C-Cube and the 3-Step Filtering algorithms were 588 and 231 times faster than the brute-force approach. In terms of accuracies, we found that the F-score of the C-cube algorithm was 90% higher than the F-score of the brute-force approach whereas the F-score for the 3-Step filtering algorithm was found to be 8,600% higher. The algorithms are decidedly faster and more accurate than the brute-force approach. These attributes make them suitable for implementation in real-time environments as well as for use in batch-mode for various applications. We expect the algorithms to play a significant role in the digitalization of healthcare information and briefly discuss a few applications.

Background Bariatric surgery, besides causing significant weight reduction, leads to improvement in type 2 diabetes mellitus (T2DM). However, there is a scarcity of data on the prediction of diabetes resolution in non-Western population. Objective To evaluate the impact of bariatric surgery on T2DM and to assess the accuracy of pre-operative scoring systems in predicting remission. Study setting A tertiary care academic centre, India. Methodology We used a retrospective cohort of all diabetic patients (n = 244) who underwent bariatric surgery at our centre in the past 10 years. The cohort was followed up for diabetes remission, and pre-operative scoring systems were analysed against the observed results. Results Of 244 patients, we were able to contact 156 patients. The median period of follow-up was 38 months. The mean body mass index (BMI) of the study group decreased from 45.4 to 33.4 kg/m2 (%excess BMI loss = 61.2%). The number of patients dependent on oral anti-diabetic pharmacotherapy and on insulin decreased from 133 (85.3%) to 40 (25.6%) and from 31 (19.9%) to 7 (4.5%), respectively. Remission was analysed for 96 patients, who submitted complete biochemical investigations. The median follow-up period for this sub-cohort was 36 months. 38 (39.6%) patients were in complete remission, 15 (15.6%) patients in partial remission and 34 (38.5%) patients showed an improved glycaemic control. The three pre-operative scores, Advanced-DiaRem, DiaRem and ABCD, showed predictive accuracies of 81.1%, 75.6% and 77.8%, respectively. Conclusions Besides leading to excess BMI loss of 61.2%, bariatric surgery also resulted in diabetes remission in 55.2% of the patients. Amongst various pre-operative scores, Advanced-DiaRem has the highest predictive accuracy for T2DM remission.

Banded sleeve gastrectomy (BSG) was developed to restrict progressive dilation of the gastric sleeve, which remains a commonly implicated reason for weight regain following SG. The present study attempted to perform a systematic review and meta-analysis comparing the two procedures. Literature search was performed across PubMed and Google Scholar, using the keywords "Banded Sleeve Gastrectomy", "Sleeve gastrectomy", "Banded", "BSG" and "LSG". It yielded 4267 articles, six of which have been included in this review. Better weight loss outcomes at 3 and 5 years are noted following BSG, with a margin of 6.39% and 9.97% in %TWL at respective time points. No difference in impact on co-morbidities was noted. A revision rate of 7.1% was seen after BSG, with increased regurgitation as the most common indication.

In recent years, there has been a resurgence in methods that use distributed (neural) representations to represent and reason about semantic knowledge for robotics applications. However, while robots often observe previously unknown concepts, these representations typically assume that all concepts are known a priori, and incorporating new information requires all concepts to be learned afresh. Our work relaxes this limiting assumption of existing representations and tackles the incremental knowledge graph embedding problem by leveraging the principles of a range of continual learning methods. Through an experimental evaluation with several knowledge graphs and embedding representations, we provide insights about trade-offs for practitioners to match a semantics-driven robotics applications to a suitable continual knowledge graph embedding method., 8 pages, 4 figures. Accepted for publication in IEEE Robotics and Automation Letters (RA-L)

SARS-CoV-2 is a unique event, having emerged suddenly as a highly infectious viral pathogen for human populations. Previous phylogenetic analyses show its closest known evolutionary relative to be a virus detected in bats (RaTG13), with a common assumption that SARS-CoV-2 evolved from a zoonotic ancestor via recent genetic changes (likely in the Spike protein receptor binding domain – or RBD) that enabled it to infect humans. We used detailed phylogenetic analysis, ancestral sequence reconstruction, and in situ molecular dynamics simulations to examine the Spike-RBD’s functional evolution, finding that the common ancestral virus with RaTG13, dating to at least 2013, possessed high binding affinity to the human ACE2 receptor. This suggests that SARS-CoV-2 likely possessed a latent capacity to bind to human cellular targets (though this may not have been sufficient for successful infection) and emphasizes the importance to expand the cataloging and monitoring of viruses circulating in both human and non-human populations.

Laparoscopic Sleeve gastrectomy (LSG) is the most commonly performed bariatric surgical procedure worldwide. There is wide variation however in post-operative weight loss on long term follow-up, and residual gastric volume (RGV) is believed to be an important variable. Multiple studies have correlated RGV as assessed by Computerized Tomography volumetry with excess weight loss (EWL%) following LSG, but definite consensus is lacking. This article systematically reviews the published studies in English literature to ascertain whether any correlation exists between the RGV and EWL% following LSG. Ten studies were included in this review, and significant differences were noted in the technique of RGV assessment, and timing of RGV and EWL% assessment. Five studies found a statistically significant correlation between the RGV and EWL%. One study found a correlation which did not reach statistical significance. Two additional studies reported that the resected volume rather than RGV correlated with the EWL%. Meta-analysis of studies reporting correlation between RGV and EWL% showed that up to 26.3% (95% CI: 5.1%-56.1%) of variability in EWL% can be explained by variations in RGV. A lower RGV is likely to result in a better post-operative weight loss following LSG. There is need for standardization of technique and timing of RGV assessment.

Background India, a low-middle income and a developing country is combating with a triple burden of malnutrition with a very cost-effective measure, infant and young child feeding (IYCF) practices. But there are a lot of challenges in its implementation which need to be catered. The objective of the present qualitative study was to assess IYCF practices among mothers of children aged 6 months to 2 years in a rural area of Haryana. Method Qualitative study was carried out among mothers of children 6 months-2 years in villages of Ballabgarh block of Haryana using focussed group discussion (FGD) and in-depth interview methods. All recordings of FGDs and IDIs were transcribed into verbatim and codes were generated. Thematic analysis of the transcript of in-depth interview and FGD was performed with the help of Doc Tools in MS Word 2016. Results The mothers had good knowledge about breastfeeding, importance of colostrum, and weaning practices of infants and children of less than 2 years. Though there is evidence of some cultural misbeliefs, most of the taboos are obsolete now. There was a knowledge gap regarding initiation and composition of complementary feeding practices. The awareness about food diversity, effects of junk food, and recommended complementary feeding practices was less. Conclusion There is need of creating awareness among mothers regarding importance of IYCF practices to reduce infant and under 5 mortality in rural area.

Introduction Multiple sclerosis (MS) is a disease that causes progressive neurological disability. Treatments are available that are protective against MS relapses and it is thought that reduction of early neuroinflammation may improve long term prognosis. At present there is no biomarker that can predict which patients may have a more severe disease course, and potentially benefit from more aggressive therapy. Long noncoding RNAs (lncRNAs) are emerging as potential disease biomarkers that could be of interest in prognostication of MS. Methods We identified a discovery cohort of 20 patients, ten of which had a mild MS phenotype and ten with severe MS phenotype according to the Age-Related MS Severity Scale (ARMSS). RNAseq was performed on RNA extracted from whole blood and bioinformatic analysis restricted to lncRNAs. Our goal was to select the most significant lncRNAs and quantify these using custom digital droplet RT-qPCR assays in a validation cohort of 44 participants (with mild or severe MS). Results Eight lncRNA candidates were identified from the discovery cohort. Of these, four lncRNAs remained significantly differentially expressed in the validation cohort (ENSG00000260302, ENSG00000270972, ENSG00000272512 and ENSG00000223387). Little is known about the precise roles of these lncRNAs but based on expression data they appear to be important to immune function and are of potential biological significance to MS pathogenesis. Conclusions This study is the first to investigate possible lncRNA biomarkers to differentiate phenotypic severity in MS. Although the findings are preliminary based on our small sample size, they are sufficient to identify hypotheses for future investigation, and give guidance regarding the design of future studies.

Background: Atypical teratoid rhabdoid tumor of the central nervous system (CNS ATRT) is a malignancy that commonly affects young children. The biological mechanisms contributing to tumor aggressiveness and resistance to conventional therapies in ATRT are unknown. Previous studies have shown the activity of insulin like growth factor-I receptor (IGF-1R) in ATRT tumor specimens and cell lines. IGF-1R has been shown to cross-talk with other receptor tyrosine kinases (RTKs) in a number of cancer types, leading to enhanced cell proliferation. Objective: This study aims to evaluate the role of IGF-1 receptor cross-talk in ATRT biology and the potential for therapeutic targeting. Methods: Cell lines derived from CNS ATRT specimens were analyzed for IGF-1 mediated cell proliferation. A comprehensive receptor tyrosine kinase (RTK) screen was conducted following IGF-1 stimulation. Bioinformatic analysis of publicly available cancer growth inhibition data to identify correlation between IC50 of a VEGFR inhibitor and IGF-1R expression. Results: Comprehensive RTK screen identified VEGFR-2 cross-activation following IGF-1 stimulation. Bioinformatics analysis demonstrated a positive correlation between IC50 values of VEGFR inhibitor Axitinib and IGF-1R expression, supporting the critical influence of IGF-1R in modulating response to anti-angiogenic therapies. Conclusion: Overall, our data present a novel experimental framework to evaluate and utilize receptor cross-talk mechanisms to select effective drugs and combinations for future therapeutic trials in ATRT.

Background Association between low counts of herpesvirus-specific T cells and subsequent relapse of hematologic malignancy has been shown in two retrospective studies. Methods Here we present results of a prospective validation study. Multiple subsets of Epstein-Barr virus (EBV)–specific T cells were measured in 69 patients on day 56 and 84, using intracellular flow cytometry after incubation of blood mononuclear cells (MNCs) with EBV peptides or lysate. Results All EBV T-cell subsets measured, both on day 56 and 84, were lower in patients who did versus did not subsequently relapse. This was most significant for day 56 EBV lysate-stimulated CD8 T cells producing interferon-gamma. Patients with day 56 counts of this subset >5/µL had a significantly lower likelihood of relapse compared with those with ≤5/µL (subhazard ratio, 5.7; P = 0.007). Similar significant associations were shown for a total of seven EBV T-cell subsets on day 56 and nine subsets on day 84. However, sensitivity and specificity of relapse prediction using the count of any subset was low (area under the curve of receiver-operator characteristic curve was Discussion In conclusion, the association between EBV T-cell counts and subsequent relapse is valid. However, its clinical utility appears to be limited.

10035 Background: Neuroblastoma (NB) is the most common extracranial solid tumor in children. Despite the development of risk stratification tools to improve prognostication, prediction of patient survival outcomes in NB remains poor. In this study we used an unbiased machine-learning algorithm to develop and validate a transcriptomic signature capable of predicting 5-year overall (OS) and event-free survival (EFS) in these patients. Methods: The TARGET-Neuroblastoma dataset (n = 243) was used as the training set. Two independent NB cohorts, E-MTAB-179 (n = 478) and GSE85047 (n = 266) were used as validation sets. Elastic net regression was employed to identify transcripts associated with EFS. Association of the developed signature with EFS and OS was evaluated using univariate Cox proportional hazards (CoxPH), Kaplan-Meier, and 5-year receiver-operator characteristic curves in validation cohorts. Further, the independent prognostic value of the signature was assessed using multivariate CoxPH models with relevant clinicopathologic variables including age, INSS stage, and N-Myc amplification status in both validation sets. Finally, a nomogram was developed to integrate the signature with prognostic clinicopathologic variables to evaluate their combined efficacy for prediction of 5-year EFS and OS. Results: We identified a 21-gene signature that demonstrates significant association with EFS and OS in both E-MTAB-178 and GSE49710 validation cohorts. Moreover, the signature is independent of clinicopathological variables and can be effectively incorporated into a risk model, improving the prognostic performance. Several genes within the signature have been previously implicated in NB, including ECEL1, HOXC9 and ARAF1. Conclusions: To the best of our knowledge, we are the first to use an unbiased machine learning approach to generate a transcriptomic gene signature for neuroblastoma prognosis externally validated in multiple cohorts across platforms. This 21-gene transcriptomic signature significantly associated with EFS and OS in this disease. Combining this signature with current prognostic clinicopathologic variables will improve risk stratification of affected patients and may inform effective clinical decision-making.[Table: see text]

Document digitization has important industrial applications and automation of this process can unlock significant business value, especially when documents are in the form of image files that are not directly machine readable. A medical prescription contains information about a patient's health status and is an important tool for prevention of drug abuse and patient safety. In our organization, in the interests of patient safety and to minimize dispensation errors, the customer is required to upload a picture of the prescription with her order if it contains any prescription drug. Once this prescription is digitized, the system decides the eligibility for dispensation of the ordered medicines based on the information extracted from the prescription. Since we process thousands of medicine orders everyday for which a large number of man-hours are required for processing the uploaded prescriptions, a key opportunity was identified in this process to make it more efficient and faster. In this paper, we describe how artificial intelligence and ingenious solutions have enabled us to automate different elements of the screening process and present metrics for evaluating the performance of the automated process. Our paper demonstrates that attended robotic process automation reduces the human effort to verify a prescription and facilitates a superior employee experience. We also underpin the importance of prescription data to the healthcare community at large and advocate that such automations further the cause of healthcare research.

With technological advancements and increment in content based advertisement, the use of Short Message Service (SMS) on phones has increased to such a significant level that devices are sometimes flooded with a number of spam SMS. These spam messages can lead to loss of private data as well. There are many content-based machine learning techniques which have proven to be effective in filtering spam emails. Modern day researchers have used some stylistic features of text messages to classify them to be ham or spam. SMS spam detection can be greatly influenced by the presence of known words, phrases, abbreviations and idioms. This paper aims to compare different classifying techniques on different datasets collected from previous research works, and evaluate them on the basis of their accuracies, precision, recall and CAP Curve. The comparison has been performed between traditional machine learning techniques and deep learning methods.

INTRODUCTION: Elevated levels of LIN28A expression in AT/RT samples provide the rationale to search for agents that may interfere with cellular pathways regulated by this molecule. Ornithine decarboxylase (ODC) modulates eIF-5A through polyamines production and has been shown to directly influence the LIN28/Let-7 axis. The anti-protozoan drug α-difluoromethylornithine (DFMO) irreversibly inhibits ODC and its activity is potentiated by GC7 and Aurothioglucose (AuTG). METHODS: Bioinformatics: ATRT Pediatric Patient Dataset from Children’s Brain Tumour Tissue Consortium was evaluated (n=32). In-vitro studies: Expression levels of LIN28 and ODC in three CNS ATRT cell lines, BT12, BT16 and KCCF1, were examined by western blotting. Cells were treated with increasing concentrations of DFMO, GC7 and AuTG and viability was measured by alamar blue assay. RESULTS: mRNA expression profile of patients suggest strong evidence of co-expression between three proteins: LIN28A – ODC1: 0.37 Pearson Coefficient (p = 0.037), ODC1 – EIF5A: 0.35 Pearson Coefficient (p = 0.049) and LIN28A – EIF5A: 0.54 Pearson Coefficient (p = 0.001). Drug sensitivity studies indicated DFMO, GC7 and AuTG were cytotoxic in-vitro. IC50 values of DFMO for BT12, BT16 and KCCF1 were 80uM, 50uM and 80uM respectively. GC-7 was cytotoxic at IC50 of 115uM, 25uM and 25uM and AuTG displayed IC50 of 15uM, 42 uM and 50uM. Markers of apoptosis, including PARP cleavage, were present at 48 hours although kinetics varied between cell lines. Drug combination studies showed significant synergy between DMFO and GC7. DISCUSSION: Data from this study identified alterations in expression and activity of components in LIN28 mediated pathways in ATRT. Interference in the LIN28 pathway by the tested drugs leads to apoptosis at physiologically attainable concentrations. Therapeutic efficacy can be further enhanced by mechanistically validated drug combinations. We discuss in detail the implications of these findings with respect to the biology and novel therapeutics development for ATRT.