Continuous patient state attention model for addressing irregularity in electronic health records.

BACKGROUND: Irregular time series (ITS) are common in healthcare as patient data is recorded in an electronic health record (EHR) system as per clinical guidelines/requirements but not for research and depends on a patient's health status. Due to irregularity, it is challenging to develop machine learning techniques to uncover vast intelligence hidden in EHR big data, without losing performance on downstream patient outcome prediction tasks. METHODS: In this paper, we propose Perceiver, a cross-attention-based transformer variant that is computationally efficient and can handle long sequences of time series in healthcare. We further develop continuous patient state attention models, using Perceiver and transformer to deal with ITS in EHR. The continuous patient state models utilise neural ordinary differential equations to learn patient health dynamics, i.e., patient health trajectory from observed irregular time steps, which enables them to sample patient state at any time. RESULTS: The proposed models' performance on in-hospital mortality prediction task on PhysioNet-2012 challenge and MIMIC-III datasets is examined. Perceiver model either outperforms or performs at par with baselines, and reduces computations by about nine times when compared to the transformer model, with no significant loss of performance. Experiments to examine irregularity in healthcare reveal that continuous patient state models outperform baselines. Moreover, the predictive uncertainty of the model is used to refer extremely uncertain cases to clinicians, which enhances the model's performance. Code is publicly available and verified at https://codeocean.com/capsule/4587224 . CONCLUSIONS: Perceiver presents a computationally efficient potential alternative for processing long sequences of time series in healthcare, and the continuous patient state attention models outperform the traditional and advanced techniques to handle irregularity in the time series. Moreover, the predictive uncertainty of the model helps in the development of transparent and trustworthy systems, which can be utilised as per the availability of clinicians.

Improving Diagnostics with Deep Forest Applied to Electronic Health Records.

An electronic health record (EHR) is a vital high-dimensional part of medical concepts. Discovering implicit correlations in the information of this data set and the research and informative aspects can improve the treatment and management process. The challenge of concern is the data sources' limitations in finding a stable model to relate medical concepts and use these existing connections. This paper presents Patient Forest, a novel end-to-end approach for learning patient representations from tree-structured data for readmission and mortality prediction tasks. By leveraging statistical features, the proposed model is able to provide an accurate and reliable classifier for predicting readmission and mortality. Experiments on MIMIC-III and eICU datasets demonstrate Patient Forest outperforms existing machine learning models, especially when the training data are limited. Additionally, a qualitative evaluation of Patient Forest is conducted by visualising the learnt representations in 2D space using the t-SNE, which further confirms the effectiveness of the proposed model in learning EHR representations.

An insight in proteome profiling of Tuta absoluta larvae after entomopathogenic fungal infection.

Tuta absoluta (L.) (Lepidoptera: Gelechiidae), a major pest of solanaceous plant species, causes serious losses in the agriculture sector around the globe. For better pest management, entomopathogenic fungi such as Beauveria bassiana and Purpureocillium lilacinum, play an efficient role in suppressing the pest population. The present study was carried out to analyse the effects post fungal infections through proteome profiling using an Orbitrap Fusion Tribrid mass spectrometer. A total of 2,201 proteins were identified from the fourth instar larvae of T. absoluta, of which 442 and 423 proteins were significantly dysregulated upon infection with P. lilacinum and B. bassiana respectively. The potential proteins related to immune systems as well as detoxification processes showed significant alterations after post fungal infection. Studies on T. absoluta proteomics and genomics as well as the consequences of entomopathogenic fungal infection on the immune response of this insect could provide an initial framework for exploring more fungus-host interactions for the development of better strategies for integrated pest management.

Assessing exposure to Kilkari: a big data analysis of a large maternal mobile messaging service across 13 states in India.

The Kilkari programme is being implemented by the Government of India in 13 states. Designed by BBC Media Action and scaled in collaboration with the Ministry of Health and Family Welfare from January 2016, Kilkari had provided mobile health information to over 10 million subscribers by the time BBC Media Action transitioned the service to the government in April 2019. Despite the reach of Kilkari in terms of the absolute number of subscribers, no longitudinal analysis of subscriber exposure to health information content over time has been conducted, which may underpin effectiveness and changes in health outcomes. In this analysis, we draw from call data records to explore exposure to the Kilkari programme in India for the 2018 cohort of subscribers. We start by assessing the timing of the first successful call answered by subscribers on entry to the programme during pregnancy or postpartum, and then assess call volume, delivery, answering and listening rates over time. Findings suggest that over half of subscribers answer their first call after childbirth, with the remaining starting in the pregnancy period. The system handles upwards of 1.2 million calls per day on average. On average, 50% of calls are picked up on the first call attempt, 76% by the third and 99.5% by the ninth call attempt. Among calls picked up, over 48% were listened to for at least 50% of the total content duration and 43% were listened to for at least 75%. This is the first analysis of its kind of a maternal mobile messaging programme at scale in India. Study analyses suggest that multiple call attempts may be required to reach subscribers. However, once answered, subscribers tend to listen the majority of the call-a figure consistent across states, over time, and by health content area.

Ten lessons learnt: scaling and transitioning one of the largest mobile health communication programmes in the world to a national government.

There has been exponential growth in the numbers of 'digital development' programmes seeking to leverage technology to solve systemic challenges. However, despite promising results and a shift from pilots to scale-ups, many have failed to realise their full potential. This paper reflects on lessons learnt from scaling and transitioning one of the largest mobile health programmes in the world to the Indian government. The complementary suite of services was designed by BBC Media Action to strengthen families' reproductive, maternal, neonatal and child health behaviours. Mobile Academy was a training course to refresh frontline health workers' (FLHWs) knowledge and improve their interpersonal communication skills. Mobile Kunji was a job aid to support FLHWs' interactions with families. Kilkari delivered weekly audio information to families' phones to reinforce FLHWs' counselling. As of April 2019, when Mobile Academy and Kilkari were transitioned to the government, 206 000 FLHWs had graduated and Kilkari had reached 10 million subscribers. Lessons learnt include the following: (1) private sector business models are challenging in low-resource settings; (2) you may pilot 'apples' but scale 'oranges'; (3) trade-offs are required between ideal solution design and affordability; (4) programme components should be reassessed before scaling; (5) operational viability at scale is a prerequisite for sustainability; (6) consider the true cost of open-source software; (7) taking informed consent in low-resource settings is challenging; (8) big data offer promise, but social norms and SIM change constrain use; (9) successful government engagements require significant capacity; (10) define governance structures and roadmaps up front.

Gut-specific arylphorin mediates midgut regenerative response against Cry-induced damage in Achaea janata.

Development of insect resistance to biopesticides is a current and pertinent global issue. Earlier, it was established that lepidopteran larvae can recover from Bt intoxication via a midgut regenerative response and subsequently generate resistance. Molecular aspects of restoration of the midgut integrity following toxin exposure are emerging recently. In the present study, we bring out the pivotal role of gut arylphorin in mediating the midgut regenerative response following sublethal Bt exposure in Achaea janata. Bt-induced midgut damage was characterized by microscopic analysis using differential interference contrast (DIC) and immunofluorescence (IF). Extensive disruption of brush-border membrane, associated with the underlying cytoskeletal alterations including F-actin, α-actin and ß-tubulin was observed. Single-photon fluorescence microscopy combined with fluorescence lifetime imaging (FLIM) established the metabolic state associated with enhanced stem cell proliferation and migration from the basal side towards the luminal side following the damage. In-silico analysis revealed the phylogenetic relationship of gut arylphorin with closely related insect species and indicated the presence of two different subunits. Transient RNAi knockdown of the arylphorin resulted in diminished expression of mitotic Cyclin B mRNA levels. Human monoclonal Cyclin B antibody cross-reactivity with the Cyclin B located in the stem cells further validate the role of arylphorin as the mitogenic factor responsible for stem cell proliferation and epithelial regeneration. An in-depth understanding of resistance mechanisms will aid in the design of new strategies for the long-term usage and efficacy of Bt technology against pest control.

Midgut aminopeptidase N expression profile in castor semilooper (Achaea janata) during sublethal Cry toxin exposure.

The midgut of lepidopteran larvae is a multifunctional tissue that performs roles in digestion, absorption, immunity, transmission of pathogens and interaction with ingested various molecules. The proteins localized at the inner apical brush border membrane are primarily digestive proteases, but some of them, like aminopeptidase N, alkaline phosphatase, cadherins, ABC transporter C2, etc., interact with Crystal (Cry) toxins produced by Bacillus thuringiensis (Bt). In the present study, aminopeptidase N (APN) was characterized as Cry-toxin-interacting protein in the larval midgut of castor semilooper, Achaea janata. Transcriptomic and proteomic analyses revealed the presence of multiple isoforms of APNs (APN1, 2, 4, 6 and 9) which have less than 40% sequence similarity but show the presence of characteristic 'GAMENEG' and zinc-binding motifs. Feeding a sublethal dose of Cry toxin caused differential expression of various APN isoform. Further, 6thgeneration Cry-toxin-exposed larvae showed reduced expression of APN2. This report suggests that A. janata larvae exploit altered expression of APNs to overcome the deleterious effects of Cry toxicity, which might facilitate toxin tolerance in the long run.

RNA-Seq analysis and de novo transcriptome assembly of Cry toxin susceptible and tolerant Achaea janata larvae.

Larvae of most lepidopteran insect species are known to be voracious feeders and important agricultural pests throughout the world. Achaea janata larvae cause serious damage to Ricinus communis (Castor) in India resulting in significant economic losses. Microbial insecticides based on crystalline (Cry) toxins of Bacillus thuringiensis (Bt) have been effective against the pest. Excessive and indiscriminate use of Bt-based biopesticides could be counter-productive and allow susceptible larvae to eventually develop resistance. Further, lack of adequate genome and transcriptome information for the pest limit our ability to determine the molecular mechanisms of altered physiological responses in Bt-exposed susceptible and tolerant insect strains. In order to facilitate biological, biochemical and molecular research of the pest species that would enable more efficient biocontrol, we report the midgut de novo transcriptome assembly and clustering of susceptible Cry toxin-exposed and Cry toxin tolerant Achaea janata larvae with appropriate age-matched and starvation controls.

Midgut de novo transcriptome analysis and gene expression profiling of Achaea janata larvae exposed with Bacillus thuringiensis (Bt)-based biopesticide formulation.

India is the major producer and exporter of castor oil in the world. Castor semilooper, Achaea janata is one of the main castor crop pests, which causes serious economic loss of crop, hence management and control of the pest are important. Currently, Bacillus thuringiensis (Bt) based biopesticides are being used for their control. However, the insects are known to develop resistance not only against chemical pesticides but also to Bt based biopesticides. In the present study, de novo transcriptome analysis was conducted to monitor the expression pattern of larval midgut genes in Achaea janata exposed to sublethal dose of Bt formulation. A total of 34,612 and 41,109 transcripts were identified in control and toxin-exposed larval midgut samples out of which 18,836 in control and 21,046 in toxin-exposed samples are annotated. Microarray data analysis employed to monitor the gene expression upon Cry toxin exposure revealed that 375 genes were upregulated and 579 genes were downregulated during all the time points (12-60 h) of toxin exposure. The differentially expressed transcripts include i.e. Cry toxin receptors, gut proteases, arylphorin, REPATs, detoxification enzymes and aquaporins. Validation of microarray data was performed by real-time quantitative PCR using few randomly selected genes and the results obtained were in corroboration. This is the first study on transcriptome data from the castor semilooper and the results would provide valuable resources for the characterization of Bt toxin response in the pest.

Larval Mid-Gut Responses to Sub-Lethal Dose of Cry Toxin in Lepidopteran Pest Achaea janata.

The lack of homogeneity in field application of Bacillus thuringiensis formulation often results in ingestion of sub-lethal doses of the biopesticide by a fraction of pest population and there by promotes the toxin tolerance and resistance in long term. Gut regeneration seems to be one of the possible mechanism by which this is accomplished. However, the existing information is primarily derived from in vitro studies using mid-gut cell cultures. Present study illustrates cellular and molecular changes in mid-gut epithelium of a Bt-susceptible polyphagous insect pest castor semilooper, Achaea janata in response to a Cry toxin formulation. The present report showed that prolonged exposure to sub-lethal doses of Cry toxin formulation has deleterious effect on larval growth and development. Histological analysis of mid-gut tissue exhibits epithelial cell degeneration, which is due to necrotic form of cell death followed by regeneration through enhanced proliferation of mid-gut stem cells. Cell death is demonstrated by confocal microscopy, flow-cytometry, and DNA fragmentation analysis. Cell proliferation in control vs. toxin-exposed larvae is evaluated by bromodeoxyuridine (BrdU) labeling and toluidine blue staining. Intriguingly, in situ mRNA analysis detected the presence of arylphorin transcripts in larval mid-gut epithelial cells. Quantitative PCR analysis further demonstrates altered expression of arylphorin gene in toxin-exposed larvae when compared with the control. The coincidence of enhanced mid-gut cell proliferation coincides with the elevated arylphorin expression upon Cry intoxication suggests that it might play a role in the regeneration of mid-gut epithelial cells.

Insecticidal Effects of Hemocoelic Delivery of Bacillus thuringiensis Cry Toxins in Achaea janata Larvae.

Insecticidal effects of Bacillus thuringiensis Cry toxins in hemocoel of larvae have not been properly evaluated. In the present study, hemocoelic injection of four representative Cry toxins i.e., Cry1Aa, Cry1Ab, Cry1Ac, and DOR5 to an economically important lepidopteran insect pest Achaea janata, induced larval mortality, reduced larval growth rate and gave rise to smaller pupae, all in a dose-dependent manner. We observed extensive degeneration as well as the disintegration of larval tissues, most notably, fat body, and the possible involvement of lysosomal enzymes in tissue histolysis. The resultant "hypoproteinemia" and most relevantly, the drastic reduction of 80-85 kDa hexamerin proteins levels of hemolymph could be attributed to the pathological state of the fat body induced by Cry toxin injection. Formation of non-viable larval-pupal intermediates and emergence of defective adults also indicate toxicity effects of Cry toxins during metamorphosis. Thus, findings from our study suggest Cry toxins in larval hemocoel are also toxic to A. janata larval survival and subsequent development.

Differential oxidative stress responses in castor semilooper, Achaea janata.

Balance between reactive oxygen species (ROS) and the antioxidant (AO) defense mechanisms is vital for organism survival. Insects serve as an ideal model to elucidate oxidative stress responses as they are prone to different kinds of stress during their life cycle. The present study demonstrates the modulation of AO enzyme gene expression in the insect pest, Achaea janata (castor semilooper), when subjected to different oxidative stress stimuli. Antioxidant enzymes' (catalase (Cat), superoxide dismutase (Sod), glutathione-S-transferase (GST) and glutathione peroxidase (Gpx)) partial coding sequences were cloned and characterized from larval whole body. Tissue expression studies reveal a unique pattern of AO genes in the larval tissues with maximum expression in the gut and fat body. Ontogeny profile depicts differential expression pattern through the larval developmental stages for each AO gene studied. Using quantitative RT-PCR, the expression pattern of these genes was monitored during sugar-induced (d-galactose feeding), infection-induced (Gram positive, Gram negative and non-pathogenic bacteria) and pesticide-induced oxidative stress (Bt Cry toxin). d-Galactose feeding differentially modulates the expression of AO genes in the larval gut and fat body. Immune challenge with Escherichia coli induces robust upregulation of AO genes when compared to Bacillus coagulans and Bacillus cereus in the larval fat body and gut. Cry toxin feeding predominantly induced GST upregulation in the gut. The current study suggests that though there are multiple ways of generation of oxidative stress in the insect, the organism tailors its response by insult- and tissue-specific recruitment of the antioxidant players and their differential regulation for each inducer.

The use of therapeutic hypothermia after cardiac arrest in a pregnant patient.

Therapeutic hypothermia is an effective intervention for the postresuscitative care of patients who have sustained a cardiac arrest. There has been only 1 documented case of successful resuscitation of a pregnant patient and fetus with therapeutic hypothermia, with an abbreviated developmental follow-up of the child. A 33-year-old woman in her 20th week of pregnancy presented to our emergency department after experiencing a cardiac arrest. After successful resuscitation and a discussion with a multidisciplinary team about expected outcomes, the mother and fetus were successfully treated with therapeutic hypothermia, and a healthy baby was delivered 19 weeks later. The mother's cardiac and neurologic function was normal 36 months after the arrest, and the child has reached all growth and neurodevelopmental milestones. We present a case demonstrating excellent immediate and long-term maternal-fetal neurologic, cardiac, and developmental outcomes after the use of therapeutic hypothermia after cardiac arrest in a pregnant patient.

An unusual cause of dysphagia: pericardial effusion after implantable cardioverter-defibrillator placement.

BACKGROUND: Dysphagia is a known complication of pericardial effusions. Most cases of pericardial effusions are idiopathic, infectious, and neoplastic, but can also occur after cardiac procedures. OBJECTIVE: To report the case of a patient who developed dysphagia from a sub-acute pericardial effusion caused by the placement of an implantable cardioverter-defibrillator (ICD). CASE REPORT: A 62-year-old woman presented to the Emergency Department (ED) with a 2-day history of dysphagia. Imaging revealed a large pericardial effusion compressing the esophagus from the mid-thoracic level to the gastroesophageal junction. Ten days prior, a dual-chamber ICD with small-diameter active fixation leads was placed in the patient. There had been no apparent complications from the procedure, however, over this 10-day period she developed a sub-acute pericardial effusion from an incidental perforation during ICD lead placement that led to the extrinsic compression of the esophagus and her presenting symptom of dysphagia. The patient underwent pericardiocentesis for the pericardial effusion and she was discharged in stable condition. CONCLUSION: This case report highlights the importance of recognizing a non-cardiac complaint such as dysphagia as the primary symptom of a critical cardiac condition. With an increase in cardiac procedures anticipated, clinicians should consider the possibility of a pericardial effusion as a cause of dysphagia, especially for those patients with recent cardiac procedures.

Shock on T versus direct current voltage for induction of ventricular fibrillation: a randomized prospective comparison.

VF is induced during ICD implantation to determine efficacy of therapy. Establishing the best clinical method of induction of VF would potentially be beneficial in reducing the number of induction attempts and reducing the frequency of inadvertent induction of VT. Commonly used methods to induce VF include shock in the T wave vulnerable period (T shock) and high frequency stimulation. This study compared the efficacy of T shock with a new induction method using a 9-V DC pulse. The study was a randomized, prospective, case crossover trial in patients receiving ICDs. VF was induced by T shock and DC in a randomized sequence during an ICD implant. VF was induced at least four times in each patient (two T shocks and two DC inductions) and with each induction; attempts were continued with modifications until successful. A paired evaluation between the T shock/DC induction was performed in 37 patients (28 men, age 64 +/- 12 years) with a left ventricular ejection fraction of 0.40 +/- 0.20. Arrhythmia indications were VT (n = 23), VF (n = 10), and VT/VF (n = 4). Drug therapy included amiodarone (n = 10), metoprolol (n = 6), digoxin (n = 1), and lidocaine (n = 1). The average T shock voltage was 207.0 +/- 16.1 V. The S1 cycle drive length was consistently 400 ms, and the mean S2 coupling interval was 317.8 +/- 19.6 ms. The length of time DC applied averaged 3.8 +/- 1.4 seconds. A total of 148 episodes of VF were included in the analysis. T shock induced VF with a cycle length of 213.5 +/- 35.1 ms, and DC induced VF with a cycle length of 214.6 +/- 34.5 ms (P = 0.86). Although VF was eventually induced for each randomization, the number of attempts required were dependent on the method of induction. The successful DC first attempt VF induction rate was 96%, with three patients requiring two attempts during one of the DC inductions. T shock had a 68% first attempt success rate with 21 patients requiring multiple T shocks to induce VF. All nine female patients had at least one unsuccessful first attempt T shock, which contributed to an overall unsuccessful first attempt induction rate significantly higher in women then men (36.1% vs 12.5%, P = 0.001). A constant DC voltage induction of VF may be more effective than T shock for induction of VF in a clinical setting because it reduces the number of attempts required to induce VF. By either method, VF appears to be more difficult to induce in women. DC induction has the advantage of simple programming of only duration of stimulation. These findings have implications particularly for ICD implantation with conscious sedation.