RESUMEN

In vivo delivery of mRNA is promising for the study of gene expression and the treatment of diseases. Lipid nanoparticles (LNPs) enable efficient delivery of mRNA constructs, but protein expression has been assumed to be limited to the liver. With specialized LNPs, delivery to extrahepatic tissue occurs in small animal models; however, it is unclear if global delivery of mRNA to all major organs is possible in humans because delivery may be affected by differences in innate immune response and relative organ size. Furthermore, limited studies with LNPs have been performed in large animal models, such as swine, due to their sensitivity to complement activation-related pseudoallergy (CARPA). In this study, we found that exogenous protein expression occurred in all major organs when swine were injected intravenously with a relatively low dose of mRNA encapsulated in a clinically relevant LNP formulation. Exogenous protein was detected in the liver, spleen, lung, heart, uterus, colon, stomach, kidney, small intestine, and brain of the swine without inducing CARPA. Furthermore, protein expression was detected in the bone marrow, including megakaryocytes, hematopoietic stem cells, and granulocytes, and in circulating white blood cells and platelets. These results show that nearly all major organs contain exogenous protein expression and are viable targets for mRNA therapies.

RESUMEN

We would like to extend our gratitude to Dr. da Hora Passos et al. for their interest in our recently published review and meta-analysis in Critical Care. In this response, we will elaborate on the points raised by the authors. We agree with the authors that LUS, like any other diagnostic technique, is valuable and safe only when utilized by trained operators. The authors expressed uncertainty regarding the sensitivity of LUS in detecting mild ARDS or ARDS at an early stage. This variance in sensitivity is more likely due to diversity in diagnostic thresholds. We advocate for global collaboration among LUS experts to align LUS methodologies and strengthen the evidence supporting LUS in the diagnosis of ARDS and its morphological subphenotypes.

Asunto(s)

RESUMEN

The striatum, the main input nucleus of the basal ganglia, receives topographically organized input from the cortex and gives rise to the direct and indirect output pathways, which have antagonistic effects on basal ganglia output directed to the cortex. We optogenetically stimulated the direct and indirect pathways in a visual and a working memory task in mice that responded by licking. Unilateral direct pathway stimulation increased the probability of lick responses toward the contralateral, non-stimulated side and increased cortical activity globally. In contrast, indirect pathway stimulation increased the probability of responses toward the stimulated side and decreased activity in the stimulated hemisphere. Moreover, direct pathway stimulation enhanced the neural representation of a contralateral visual stimulus during the delay of the working memory task, whereas indirect pathway stimulation had the opposite effect. Our results demonstrate how these two pathways influence perceptual decisions and working memory and modify activity in the dorsal cortex.

RESUMEN

Platelets contribute to a variety of physiological processes including inflammation, sepsis and cancer. However, due to their primary role in hemostasis, platelet transfusions are largely restricted to managing thrombocytopenia and bleeding. One way to expand the utility of platelet transfusions would be to genetically engineer donor platelets with new or enhanced functions. We have previously shown that lipid nanoparticles containing mRNA (mRNA-LNP) can be used to genetically modify authentic platelets in a non-clinical crystalloid solution. Currently, platelets collected for transfusion are stored in plasma or in plasma supplemented with platelet additive solution (PAS) at supraphysiological concentrations at room temperature, or at 4 ºC if intended for use in acute hemorrhage. Here we describe a new plasma-optimized mRNA-LNP for transfecting platelets directly in plasma and plasma supplemented with PAS that is scalable to physiological and supraphysiological platelet concentrations. Transfecting platelets in clinical solutions with mRNA-LNP does not affect aspects of in vitro physiology, and transfected platelets are storable. The compatibility of this transfection system with current clinical practices could enable future mRNA-LNP based platelet products and cell therapies.

RESUMEN

Schistosomiasis is an infection caused by contact with Schistosoma-contaminated water and affects more than 230 million people worldwide with varying morbidity. The roles of T helper 2 (TH2) cells and regulatory immune responses in chronic infection are well documented, but less is known about human immune responses during acute infection. Here, we comprehensively map immune responses during controlled human Schistosoma mansoni infection using male or female cercariae. Immune responses to male or female parasite single-sex infection were comparable. An early TH1-biased inflammatory response was observed at week 4 after infection, which was particularly apparent in individuals experiencing symptoms of acute schistosomiasis. By week 8 after infection, inflammatory responses were followed by an expansion of TH2 and regulatory cell subsets. This study demonstrates the shift from TH1 to both TH2 and regulatory responses, typical of chronic schistosomiasis, in the absence of egg production and provides immunological insight into the clinical manifestations of acute schistosomiasis.

Asunto(s)

RESUMEN

BACKGROUND: Travellers visiting rabies-endemic countries are at risk of rabies infection. Assessing travellers' knowledge and risk perception of rabies and risk behaviour during travel can help identify knowledge gaps and improve pre-travel risk education. METHODS: Cohort study in Dutch adult travellers, using two surveys: one before travel to assess knowledge and perception of rabies, and one after return to identify risk behaviour during travel. RESULTS: The pre-travel and post-travel survey were completed by 301 and 276 participants, respectively. 222 participants had travelled to a high-risk rabies-endemic country. 21.6 % of the participants scored their rabies knowledge as poor. Some participants were unaware cats or bats can transmit rabies (26.6 % and 13.6 %, respectively), or that post-exposure prophylaxis (PEP) is required for certain exposures such as skin abrasions without bleeding or licks on damaged skin (35.5 % and 18.9 %, respectively), while 27.9 % of participants did not know PEP needs to be administered within one day. 115 participants (51.8 %) reported any form of contact with any animal during travel. Two participants reported animal exposure, of which one took adequate PEP measures. Risk factors for animal contact abroad were regularly touching cats or dogs at home or abroad, longer travel duration, having pets during childhood and being an animal lover. CONCLUSIONS: Pre-travel rabies risk education currently does not meet travellers' needs, which is reflected in knowledge gaps and engagement in risk behaviour during travel. During pre-travel health advice, avoiding animal contact abroad should be emphasized, and additional education is required about indications for PEP.

Asunto(s)

RESUMEN

BACKGROUND: High-output intestinal fistulas and small bowel enterostomies are associated with morbidity and mortality. Current standard treatment for output reduction consists of fluid and dietary restrictions and medical therapy. There is conflicting evidence regarding the use of somatostatin analogues for output reduction. AIM: The aim of this study is to investigate whether lanreotide, added to current standard treatment, further reduces intestinal output in patients with high-output fistulas and enterostomies. METHODS: This was an open-label, multicentre, randomised controlled trial. Adult patients with a high-output intestinal fistula (>500 mL/24 h) or small bowel enterostomy (>1500 mL/24 h) more than 4 weeks post-surgery and receiving standard medical treatment (dietary- and fluid restriction, PPI, loperamide and codeine) for at least 2 weeks were eligible for inclusion. We randomised patients 1:1 between continuing standard treatment (control), and subcutaneous lanreotide 120 mg every 4 weeks with standard treatment. The primary outcome was the number of responders, with response defined as an output reduction of ≥25%, 8 weeks after randomisation. We also investigated the proportional change in output. RESULTS: We randomised 40 patients; 17 had a fistula and 23 a small bowel enterostomy. There were 9/20 responders in the intervention group and 2/20 in the control group (p = 0.013). The proportional output reduction was -26% (IQR -4 to -38) in the intervention group, compared to an increase of 4% (IQR 20 to -13) in the control group (p = 0.004). CONCLUSIONS: In patients with a high-output fistula or small bowel enterostomy, addition of lanreotide to current standard treatment can provide a clinically relevant output reduction. TRIAL REGISTRATION: EudraCT: 2013-003998-10.

Asunto(s)

RESUMEN

BACKGROUND: Acute respiratory distress syndrome (ARDS) is a life-threatening respiratory condition with high mortality rates, accounting for 10% of all intensive care unit admissions. Lung ultrasound (LUS) as diagnostic tool for acute respiratory failure has garnered widespread recognition and was recently incorporated into the updated definitions of ARDS. This raised the hypothesis that LUS is a reliable method for diagnosing ARDS. OBJECTIVES: We aimed to establish the accuracy of LUS for ARDS diagnosis and classification of focal versus non-focal ARDS subphenotypes. METHODS: This systematic review and meta-analysis used a systematic search strategy, which was applied to PubMed, EMBASE and cochrane databases. Studies investigating the diagnostic accuracy of LUS compared to thoracic CT or chest radiography (CXR) in ARDS diagnosis or focal versus non-focal subphenotypes in adult patients were included. Quality of studies was evaluated using the QUADAS-2 tool. Statistical analyses were performed using "Mada" in Rstudio, version 4.0.3. Sensitivity and specificity with 95% confidence interval of each separate study were summarized in a Forest plot. RESULTS: The search resulted in 2648 unique records. After selection, 11 reports were included, involving 2075 patients and 598 ARDS cases (29%). Nine studies reported on ARDS diagnosis and two reported on focal versus non-focal ARDS subphenotypes classification. Meta-analysis showed a pooled sensitivity of 0.631 (95% CI 0.450-0.782) and pooled specificity of 0.942 (95% CI 0.856-0.978) of LUS for ARDS diagnosis. In two studies, LUS could accurately differentiate between focal versus non-focal ARDS subphenotypes. Insufficient data was available to perform a meta-analysis. CONCLUSION: This review confirms the hypothesis that LUS is a reliable method for diagnosing ARDS in adult patients. For the classification of focal or non-focal subphenotypes, LUS showed promising results, but more research is needed.

Asunto(s)

RESUMEN

BACKGROUND: It is unknown whether liver surgery leads to increased RAAS activity and anti-diuretic hormone (ADH) levels and subsequent fluid accumulation. Furthermore, it is unknown whether the peri-operative fluid strategy changes this effect. METHODS: This is a pre-planned post hoc analysis of a randomised controlled trial which compared restrictive (n = 20) versus liberal fluid strategy (n = 20) in patients undergoing liver surgery. Primary outcomes for the current study were the difference in hormone levels after anaesthesia induction and after liver resection. Fluid overload was defined as a ≥10% increase in weight. RESULTS: Renin activity (6 [2.1-15.5] vs. 12 [4.6-33.5]) and ADH levels (6.0 [1.7-16.3] vs. 3.8 [1.6-14.7]) did not differ significantly before and after resection. However, aldosterone levels were significantly higher after resection (0.30 [0.17-0.49] vs. 0.69 [0.31-1.21] ). Renin activity and aldosterone levels did not differ between the groups. ADH was significantly higher in the restrictive strategy group (1.6 [1.1-2.1] vs 5.9 [3.8-16.0]). No differences in hormone levels were found in patients with and without fluid overload. DISCUSSION: Aldosterone levels increased after liver surgery but renin activity and ADH levels did not. ADH levels were higher in the restrictive group. Development of post-operative fluid overload was not associated with RAAS activity or ADH levels.

RESUMEN

BACKGROUND: Incidence of central venous catheter (CVC)-related thrombosis in critically ill patients remains ambiguous and its association with potential hazardous sequelae unknown. The primary aim of the study was to evaluate the epidemiology of CVC-related thrombosis; secondary aims were to assess the association of catheter-related thrombosis with catheter-related infection, pulmonary embolism and mortality. METHODS: This was a single-center, prospective observational study conducted at a tertiary intensive care unit (ICU) in the Netherlands. The study population consisted of CVC placements in adult ICU patients with a minimal indwelling time of 48 h. CVC-related thrombosis was diagnosed with ultrasonography. Primary outcomes were prevalence and incidence, incidence was reported as the number of cases per 1000 indwelling days. RESULTS: 173 CVCs in 147 patients were included. Median age of patients was 64.0 [IQR: 52.0, 72.0] and 71.1 % were male. Prevalence of thrombosis was 0.56 (95 % CI: 0.49, 0.63) and incidence per 1000 indwelling days was 65.7 (95 % CI: 59.0, 72.3). No association with catheter-related infection was found (p = 0.566). There was a significant association with pulmonary embolism (p = 0.022). All 173 CVCs were included in the survival analysis. Catheter-related thrombosis was associated with a lower 28-day mortality risk (hazard ratio: 0.39, 95 % CI: 0.17, 0.87). CONCLUSION: In critically ill patients, prevalence and incidence of catheter-related thrombosis were high. Catheter-related thrombosis was not associated with catheter-related infections, but was associated with pulmonary embolism and a decreased mortality risk.

Asunto(s)

RESUMEN

BACKGROUND: Acute respiratory distress syndrome (ARDS) is a frequent cause of hypoxemic respiratory failure with a mortality rate of approximately 30%. Identifying ARDS subphenotypes based on "focal" or "non-focal" lung morphology has the potential to better target mechanical ventilation strategies of individual patients. However, classifying morphology through chest radiography or computed tomography is either inaccurate or impractical. Lung ultrasound (LUS) is a non-invasive bedside tool that can accurately distinguish "focal" from "non-focal" lung morphology. We hypothesize that LUS-guided personalized mechanical ventilation in ARDS patients leads to a reduction in 90-day mortality compared to conventional mechanical ventilation. METHODS: The Personalized Mechanical Ventilation Guided by UltraSound in Patients with Acute Respiratory Distress Syndrome (PEGASUS) study is an investigator-initiated, international, randomized clinical trial (RCT) that plans to enroll 538 invasively ventilated adult intensive care unit (ICU) patients with moderate to severe ARDS. Eligible patients will receive a LUS exam to classify lung morphology as "focal" or "non-focal". Thereafter, patients will be randomized within 12 h after ARDS diagnosis to receive standard care or personalized ventilation where the ventilation strategy is adjusted to the morphology subphenotype, i.e., higher positive end-expiratory pressure (PEEP) and recruitment maneuvers for "non-focal" ARDS and lower PEEP and prone positioning for "focal" ARDS. The primary endpoint is all-cause mortality at day 90. Secondary outcomes are mortality at day 28, ventilator-free days at day 28, ICU length of stay, ICU mortality, hospital length of stay, hospital mortality, and number of complications (ventilator-associated pneumonia, pneumothorax, and need for rescue therapy). After a pilot phase of 80 patients, the correct interpretation of LUS images and correct application of the intervention within the safe limits of mechanical ventilation will be evaluated. DISCUSSION: PEGASUS is the first RCT that compares LUS-guided personalized mechanical ventilation with conventional ventilation in invasively ventilated patients with moderate and severe ARDS. If this study demonstrates that personalized ventilation guided by LUS can improve the outcomes of ARDS patients, it has the potential to shift the existing one-size-fits-all ventilation strategy towards a more individualized approach. TRIAL REGISTRATION: The PEGASUS trial was registered before the inclusion of the first patient, https://clinicaltrials.gov/ (ID: NCT05492344).

Asunto(s)

RESUMEN

Angiogenesis, the growth of new blood vessels from pre-existing vasculature, is essential for the development of new organ systems, but transcriptional control of angiogenesis remains incompletely understood. Here we show that FOXC1 is essential for retinal angiogenesis. Endothelial cell (EC)-specific loss of Foxc1 impairs retinal vascular growth and expression of Slc3a2 and Slc7a5, which encode the heterodimeric CD98 (LAT1/4F2hc) amino acid transporter and regulate the intracellular transport of essential amino acids and activation of the mammalian target of rapamycin (mTOR). EC-Foxc1 deficiency diminishes mTOR activity, while administration of the mTOR agonist MHY-1485 rescues perturbed retinal angiogenesis. EC-Foxc1 expression is required for retinal revascularization and resolution of neovascular tufts in a model of oxygen-induced retinopathy. Foxc1 is also indispensable for pericytes, a critical component of the blood-retina barrier during retinal angiogenesis. Our findings establish FOXC1 as a crucial regulator of retinal vessels and identify therapeutic targets for treating retinal vascular disease.

Asunto(s)

RESUMEN

Sub-cellular compartmentalization of metabolism has important implications for the local production of metabolites and redox co-factors, as well as pathway regulation. 4'-phosphopantetheinyl (4'PP) groups are essential co-factors derived from coenzyme A and added to target proteins in both the cytoplasm and mitochondria by p hospho p antetheinyl transferase (PPTase) enzymes. Mammals express only one PPTase, thought to localize to the cytoplasm: aminoadipate semialdehyde dehydrogenase phosphopantetheinyl transferase (AASDHPPT); raising the question of how mitochondrial proteins are 4'PP-modified. We found that AASDHPPT is required for mitochondrial respiration and oxidative metabolism via the mitochondrial fatty acid synthesis (mtFAS) pathway. Moreover, we discovered that a pool of AASDHPPT localizes to the mitochondrial matrix via an N-terminal mitochondrial targeting sequence contained within the first 13 amino acids of the protein. Our data show that mitochondrial localization of AASDHPPT is required to support mtFAS function, and further identify two variants in Aasdhppt that are likely pathogenic in humans.

RESUMEN

Genetic manipulation of animal models is a fundamental research tool in biology and medicine but is challenging in large animals. In rodents, models can be readily developed by knocking out genes in embryonic stem cells or by knocking down genes through in vivo delivery of nucleic acids. Swine are a preferred animal model for studying the cardiovascular and immune systems, but there are limited strategies for genetic manipulation. Lipid nanoparticles (LNPs) efficiently deliver small interfering RNA (siRNA) to knock down circulating proteins, but swine are sensitive to LNP-induced complement activation-related pseudoallergy (CARPA). We hypothesized that appropriately administering optimized siRNA-LNPs could knock down circulating levels of plasminogen, a blood protein synthesized in the liver. siRNA-LNPs against plasminogen (siPLG) reduced plasma plasminogen protein and hepatic plasminogen mRNA levels to below 5% of baseline values. Functional assays showed that reducing plasminogen levels modulated systemic blood coagulation. Clinical signs of CARPA were not observed, and occasional mild and transient hepatotoxicity was present in siPLG-treated animals at 5 h post-infusion, which returned to baseline by 7 days. These findings advance siRNA-LNPs in swine models, enabling genetic engineering of blood and hepatic proteins, which can likely expand to proteins in other tissues in the future.

RESUMEN

The OligoMetastatic Esophagogastric Cancer (OMEC) project aims to provide clinical practice guidelines for the definition, diagnosis, and treatment of esophagogastric oligometastatic disease (OMD). Guidelines were developed according to AGREE II and GRADE principles. Guidelines were based on a systematic review (OMEC-1), clinical case discussions (OMEC-2), and a Delphi consensus study (OMEC-3) by 49 European expert centers for esophagogastric cancer. OMEC identified patients for whom the term OMD is considered or could be considered. Disease-free interval (DFI) was defined as the time between primary tumor treatment and detection of OMD. Moderate to high quality of evidence was found (i.e. 1 randomized and 4 non-randomized phase II trials) resulting in moderate recommendations. OMD is considered in esophagogastric cancer patients with 1 organ with ≤ 3 metastases or 1 involved extra-regional lymph node station. In addition, OMD continues to be considered in patients with OMD without progression in number of metastases after systemic therapy. 18F-FDG PET/CT imaging is recommended for baseline staging and for restaging after systemic therapy when local treatment is considered. For patients with synchronous OMD or metachronous OMD and a DFI ≤ 2 years, recommended treatment consists of systemic therapy followed by restaging to assess suitability for local treatment. For patients with metachronous OMD and DFI > 2 years, upfront local treatment is additionally recommended. These multidisciplinary European clinical practice guidelines for the uniform definition, diagnosis and treatment of esophagogastric OMD can be used to standardize inclusion criteria in future clinical trials and to reduce variation in treatment.

Asunto(s)

RESUMEN

BACKGROUND: Neuroprostheses are used to electrically stimulate the brain, modulate neural activity and restore sensory and motor function following injury or disease, such as blindness, paralysis, and other movement and psychiatric disorders. Recordings are often made simultaneously with stimulation, allowing the monitoring of neural signals and closed-loop control of devices. However, stimulation-evoked artifacts may obscure neural activity, particularly when stimulation and recording sites are nearby. Several methods have been developed to remove stimulation artifacts, but it remains challenging to validate and compare these methods because the 'ground-truth' of the neuronal signals may be contaminated by artifacts. NEW METHOD: Here, we delivered stimulation to the visual cortex via a high-channel-count prosthesis while recording neuronal activity and stimulation artifacts. We quantified the waveforms and temporal properties of stimulation artifacts from the cortical visual prosthesis (CVP) and used them to build a dataset, in which we simulated the neuronal activity and the stimulation artifacts. We illustrate how to use the simulated data to evaluate the performance of six software-based artifact removal methods (Template subtraction, Linear interpolation, Polynomial fitting, Exponential fitting, SALPA and ERAASR) in a CVP application scenario. RESULTS: We here focused on stimulation artifacts caused by electrical stimulation through a high-channel-count cortical prosthesis device. We find that the Polynomial fitting and Exponential fitting methods outperform the other methods in recovering spikes and multi-unit activity. Linear interpolation and Template subtraction recovered the local-field potentials. CONCLUSION: Polynomial fitting and Exponential fitting provided a good trade-off between the quality of the recovery of spikes and multi-unit activity (MUA) and the computational complexity for a cortical prosthesis.

Asunto(s)

RESUMEN

T cells modified to express intelligently designed chimeric antigen receptors (CARs) are exceptionally powerful therapeutic agents for relapsed and refractory blood cancers and have the potential to revolutionize therapy for many other diseases. To circumvent the complexity and cost associated with broad-scale implementation of ex vivo manufactured adoptive cell therapy products, alternative strategies to generate CAR T cells in vivo by direct infusion of nanoparticle-formulated nucleic acids or engineered viral vectors under development have received a great deal of attention in the past few years. Here, we outline the ex vivo manufacturing process as a motivating framework for direct in vivo strategies and discuss emerging data from preclinical models to highlight the potency of the in vivo approach, the applicability for new disease indications, and the remaining challenges associated with clinical readiness, including delivery specificity, long term efficacy, and safety.

Asunto(s)

RESUMEN

INTRODUCTION: The OligoMetastatic Esophagogastric Cancer (OMEC) project aims to provide clinical practice guidelines for the definition, diagnosis, and treatment of esophagogastric oligometastatic disease (OMD). METHODS: Guidelines were developed according to AGREE II and GRADE principles. Guidelines were based on a systematic review (OMEC-1), clinical case discussions (OMEC-2), and a Delphi consensus study (OMEC-3) by 49 European expert centers for esophagogastric cancer. OMEC identified patients for whom the term OMD is considered or could be considered. Disease-free interval (DFI) was defined as the time between primary tumor treatment and detection of OMD. RESULTS: Moderate to high quality of evidence was found (i.e. 1 randomized and 4 non-randomized phase II trials) resulting in moderate recommendations. OMD is considered in esophagogastric cancer patients with 1 organ with ≤ 3 metastases or 1 involved extra-regional lymph node station. In addition, OMD continues to be considered in patients with OMD without progression in number of metastases after systemic therapy. 18F-FDG PET/CT imaging is recommended for baseline staging and for restaging after systemic therapy when local treatment is considered. For patients with synchronous OMD or metachronous OMD and a DFI ≤ 2 years, recommended treatment consists of systemic therapy followed by restaging to assess suitability for local treatment. For patients with metachronous OMD and DFI > 2 years, upfront local treatment is additionally recommended. DISCUSSION: These multidisciplinary European clinical practice guidelines for the uniform definition, diagnosis and treatment of esophagogastric OMD can be used to standardize inclusion criteria in future clinical trials and to reduce variation in treatment.

Asunto(s)

RESUMEN

Over a decade ago, longitudinal multiomics analysis was pioneered for early disease detection and individually tailored precision health interventions. However, high sample processing costs, expansive multiomics measurements along with complex data analysis have made this approach to precision/personalized medicine impractical. Here we describe in a case report, a more practical approach that uses fewer measurements, annual sampling, and faster decision making. We also show how this approach offers promise to detect an exceedingly rare and potentially fatal condition before it fully manifests. Specifically, we describe in the present case report how longitudinal multiomics monitoring (LMOM) helped detect a precancerous pancreatic tumor and led to a successful surgical intervention. The patient, enrolled in an annual blood-based LMOM since 2018, had dramatic changes in the June 2021 and 2022 annual metabolomics and proteomics results that prompted further clinical diagnostic testing for pancreatic cancer. Using abdominal magnetic resonance imaging, a 2.6 cm lesion in the tail of the patient's pancreas was detected. The tumor fluid from an aspiration biopsy had 10,000 times that of normal carcinoembryonic antigen levels. After the tumor was surgically resected, histopathological findings confirmed it was a precancerous pancreatic tumor. Postoperative omics testing indicated that most metabolite and protein levels returned to patient's 2018 levels. This case report illustrates the potentials of blood LMOM for precision/personalized medicine, and new ways of thinking medical innovation for a potentially life-saving early diagnosis of pancreatic cancer. Blood LMOM warrants future programmatic translational research with the goals of precision medicine, and individually tailored cancer diagnoses and treatments.

Asunto(s)

RESUMEN

Many cognitive problems can be decomposed into series of subproblems that are solved sequentially by the brain. When subproblems are solved, relevant intermediate results need to be stored by neurons and propagated to the next subproblem, until the overarching goal has been completed. We will here consider visual tasks, which can be decomposed into sequences of elemental visual operations. Experimental evidence suggests that intermediate results of the elemental operations are stored in working memory as an enhancement of neural activity in the visual cortex. The focus of enhanced activity is then available for subsequent operations to act upon. The main question at stake is how the elemental operations and their sequencing can emerge in neural networks that are trained with only rewards, in a reinforcement learning setting. We here propose a new recurrent neural network architecture that can learn composite visual tasks that require the application of successive elemental operations. Specifically, we selected three tasks for which electrophysiological recordings of monkeys' visual cortex are available. To train the networks, we used RELEARNN, a biologically plausible four-factor Hebbian learning rule, which is local both in time and space. We report that networks learn elemental operations, such as contour grouping and visual search, and execute sequences of operations, solely based on the characteristics of the visual stimuli and the reward structure of a task. After training was completed, the activity of the units of the neural network elicited by behaviorally relevant image items was stronger than that elicited by irrelevant ones, just as has been observed in the visual cortex of monkeys solving the same tasks. Relevant information that needed to be exchanged between subroutines was maintained as a focus of enhanced activity and passed on to the subsequent subroutines. Our results demonstrate how a biologically plausible learning rule can train a recurrent neural network on multistep visual tasks.

Asunto(s)