Loss of RAS Mutations in Liquid Biopsies of Patients With Multi-Treated Metastatic Colorectal Cancer.

BACKGROUND: Liquid biopsy (LB) is a non-invasive tool to evaluate the heterogeneity of tumors. Since RAS mutations (RAS-mut) play a major role in resistance to antiepidermal growth factor receptor inhibitors (EGFR) monoclonal antibodies (Mabs), serial monitoring of RAS-mut with LB may be useful to guide treatment. The main aim of this study was to evaluate the prognostic value of the loss of RAS-mut (NeoRAS-wt) in LB, during the treatment of metastatic colorectal cancer (mCRC). METHODS: A retrospective study was conducted on patients with mCRC between January 2018 and December 2021. RAS-mut were examined in tissue biopsy, at mCRC diagnosis, and with LB, during treatment. RESULTS: Thirty-nine patients with RAS-mut mCRC were studied. LB was performed after a median of 3 lines (0-7) of systemic treatment including anti-vascular endothelial growth factor (anti-VEGF) Mabs. NeoRAS-wt was detected in 13 patients (33.3%); 9 (69.2%) of them received further treatment with anti-EGFR Mabs with a disease control rate of 44.4%. Median overall survival (OS), from the date of LB testing, was 20 months in the NeoRAS-wt group and 9 months in the persistent RAS-mut group (log-rank 2.985; P = .08), with a 12-month OS of 84.6% and 57.7%, respectively. NeoRAS-wt was identified as a predictor of survival (HR = 0.29; P = .007), with an 11-month improvement in median OS and a 71% decrease in risk of death, in heavily pretreated patients. CONCLUSIONS: In conclusion, monitoring clonal evolution in mCRC by LB may provide an additional treatment line for patients with NeoRAS-wt in advanced disease.

Effects of an online program including mindfulness, exercise therapy and patient education compared to online exercise therapy and patient education for people with Patellofemoral Pain: protocol for a randomized clinical trial.

BACKGROUND: Patellofemoral pain (PFP) is a common knee disorder that causes persistent pain, lower self-reported function and quality of life. People with PFP also present with altered psychological factors, which are associated with higher levels of pain and dysfunction. Mindfulness-based interventions (MBI) generally consist of meditative practices developed to provide a holistic approach to chronic conditions. However, the effects of MBI on clinical and psychological outcomes for people with PFP remains understudied. METHODS: This assessor-blinded, parallel, two-arm randomized clinical trial aims to investigate the effects of adding an 8-week online MBI program to exercise therapy and patient education on clinical and psychological factors for people with PFP. We also aim to investigate whether psychological factors mediate changes in pain and function. Sixty-two participants with PFP will be recruited and randomized into one of two treatment groups (Mindfulness or Control group). Both groups will receive an 8-week intervention involving exercise therapy and education delivered through an online platform. The Mindfulness group will additionally receive a MBI component including formal and informal practices. Outcomes will be assessed online at baseline, intervention endpoint (follow-up 1) and 12 months after intervention completion (follow-up 2). Comparisons between groups will be performed at all time points with linear mixed models. A mediation analysis will be performed using a 3-variable framework. DISCUSSION: Exercise therapy and patient education are considered the "best management" options for PFP. However, unsatisfactory long-term prognosis remains an issue. It is known that people with PFP present with altered psychological factors, which should be considered during the evaluation and treatment of people with PFP. Adding a MBI to the current best treatment for PFP may improve short and long-term effects by addressing the underlying psychological factors. TRIAL REGISTRATION: Registro Brasileiro de Ensaios Clínicos (ReBEC) RBR-4yhbqwk, registered in April 6, 2021.

Influence of a Biomass-Burning Event in PM2.5 Concentration and Air Quality: A Case Study in the Metropolitan Area of São Paulo.

Severe biomass burning (BB) events have become increasingly common in South America in the last few years, mainly due to the high number of wildfires observed recently. Such incidents can negatively influence the air quality index associated with PM2.5 (particulate matter, which is harmful to human health). A study performed in the Metropolitan Area of São Paulo (MASP) took place on selected days of July 2019, evaluated the influence of a BB event on air quality. Use of combined remote sensing, a surface monitoring system and data modeling and enabled detection of the BB plume arrival (light detection and ranging (lidar) ratio of (50 ± 34) sr at 532 nm, and (72 ± 45) sr at 355 nm) and how it affected the Ångström exponent (>1.3), atmospheric optical depth (>0.7), PM2.5 concentrations (>25 µg.m-3), and air quality classification. The utilization of high-order statistical moments, obtained from elastic lidar, provided a new way to observe the entrainment process, allowing understanding of how a decoupled aerosol layer influences the local urban area. This new novel approach enables a lidar system to obtain the same results as a more complex set of instruments and verify how BB events contribute from air masses aloft towards near ground ones.

ONKOTEV Score as a Predictive Tool for Thromboembolic Events in Pancreatic Cancer-A Retrospective Analysis.

BACKGROUND: Venous thromboembolism (VTE) is a frequent complication in patients with cancer and causes considerable morbidity and mortality. The risk of VTE is higher in patients with pancreatic cancer and is often associated with treatment delays or interruptions. Recently, the ONKOTEV score was proposed as a VTE risk predictor model for patients with cancer, but its validation is still ongoing. PATIENTS AND METHODS: We conducted a retrospective study to determine the incidence of VTE and to evaluate the ONKOTEV score as a VTE predictive tool in a population of patients with pancreatic cancer. RESULTS: Between February 2012 and May 2017, 165 patients were included in the study. The median age was 73 years, 45.5% of patients were female, and 55.8% had stage IV disease. Fifty-one patients had a VTE (30.9%); 23.5% had pulmonary embolism, 25.5% had deep venous thrombosis, and 51.0% had visceral VTE (VsT). At a median follow-up time of 6.3 months, cumulative incidence of VTE was less than 10% for ONKOTEV scores 0 or 1 and approximately 40% and 70% for scores 2 and ≥3, respectively. CONCLUSION: The high VTE incidence observed in this study is consistent with prior reports. Patients at high risk for VTE with no increase in hemorrhagic risk should be considered for primary thromboprophylaxis. The ONKOTEV score may stratify VTE risk in patients with pancreatic cancer, with ONKOTEV score ≥2 being associated with a higher VTE occurrence. IMPLICATIONS FOR PRACTICE: Venous thromboembolism (VTE) is a frequent complication of patients with pancreatic cancer and causes considerable morbidity, treatment delays or interruptions, and mortality. Thromboprophylaxis is not used routinely in ambulatory patients. Tools to stratify the risk of VTE are important to help select patients who may benefit from thromboprophylaxis. Recently, the ONKOTEV score was proposed as a VTE risk predictor model for patients with cancer, but its validation is still ongoing. In this patient series, ONKOTEV score ≥2 was associated with high VTE occurrence and may stratify VTE risk in patients with pancreatic cancer, suggesting that ONKOTEV can be considered to select patients with pancreatic cancer for primary thromboprophylaxis.

Comparing Different Methods for Named Entity Recognition in Portuguese Neurology Text.

Electronic Medical Records (EMRs) are written in an unstructured way, often using natural language. Information Extraction (IE) may be used for acquiring knowledge from such texts, including the automatic recognition of meaningful entities, through models for Named Entity Recognition (NER). However, while most work on the previous was made for English, this experience aimed at testing different methods in Portuguese text, more precisely, on the domain of Neurology, and take some conclusions. This paper comprised the comparison between Conditional Random Fields (CRF), bidirectional Long Short-term Memory - Conditional Random Fields (BiLSTM-CRF) and a BiLSTM-CRF with residual learning connections, using not only Portuguese texts from medical journals but also texts from the Coimbra Hospital and Universitary Centre (CHUC) Neurology Service. Furthermore, the performances of BiLSTM-CRF models using word embeddings (WEs) trained with clinical text and WEs trained with general language texts were compared. Deep learning models achieved F1-Scores of nearly 83% and 75%, respectively for relaxed and strict evaluation, on texts extracted from the medical journal. For texts collected from the Hospital, the same achieved F1-Scores of nearly 71% and 62%. This work concludes that deep learning models outperform the shallow learning models and that in-domain WEs get better results than general language WEs, even when the latter are trained with much more text than the former. Furthermore, the results show that it is possible to extract information from Hospital clinical texts with models trained with clinical cases extracted from medical journals, and thus openly available. Nevertheless, such results still require a healthcare technician to check if the information is well extracted.

Molecular ordering of PAH/MA-co-DR13 azopolymer layer-by-layer films probed by second-harmonic generation.

Molecular orientation within azopolymer thin films is important for their nonlinear optical properties and photonic applications. We have used optical second-harmonic generation (SHG) to study the molecular orientation of Layer-by-Layer (LbL) films of a cationic polyelectrolyte (poly(allylamine hydrochloride)) and an anionic polyelectrolyte containing azochromophore side groups (MA-co-DR13) on a glass substrate. The SHG measurements indicate that there is a preferential orientation of the azochromophores in the film, leading to a significant optical nonlinearity. However, both the signal strength and its anisotropy are not homogeneous throughout the sample, indicating the presence of large orientational domains. This is corroborated with Brewster angle microscopy. The average SHG signal does not increase with film thickness, in contrast to some reports in the literature, indicating an independent orientational order for successive bilayers. Analyzing the SHG signal as a function of the input and output polarizations, a few parameters of the azochromophore orientational distribution can be deduced. Fitting the SHG signal to a simple model distribution, we have concluded that the chromophores have an angular distribution with a slight in-plane anisotropy and a mean polar angle ranging from 45° to 80° with respect to substrate normal direction, with a relatively large width of about 25°. These results show that SHG is a powerful technique for a detailed investigation of the molecular orientation in azopolymer LbL films, allowing a deeper understanding of their self-assembling mechanism and nonlinear optical properties. The inhomogeneity and anisotropy of these films may have important consequences for their applications in nonlinear optical devices.

Exposure to different glyphosate formulations on the oxidative and histological status of Rhamdia quelen.

Due to the wide use of glyphosate (GLY) in soybean cultivation, their residues in the environment may affect non-target organisms such as fish, developing toxic effects. Despite GLY being widely used in Brazil, there are few studies comparing the effects of commercial formulations in native freshwater fish species. Silver catfish (Rhamdia quelen) were exposed to three different commercial formulations of GLY 48% (Orium(®), Original(®) and Biocarb(®)) at 0.0, 2.5 and 5.0 mg/L for 96 h. The effects in thiobarbituric acid-reactive substances (TBARS), catalase (CAT), superoxide dismutase (SOD), glutathione-S-transferase (GST) and histological alterations were analysed in the liver, whereas alanine aminotransferase (ALT) and aspartate aminotransferase (AST) were studied in the plasma. In the liver, TBARS levels increased and CAT decreased in all treatments and herbicides tested in comparison with the control group. The SOD increased at 2.5 mg/L of Orium(®), Original(®) and 5.0 mg/L Orium(®) and Biocarb(®), whereas GST increased at 2.5 mg/L Orium(®) and decreased at 2.5 mg/L Biocarb(®) when compared to the control group. The main histopathological alterations in hepatic tissue were vacuolisation, leucocyte infiltration, degeneration of cytoplasm and melanomacrophage in all GLY treatments. The ALT decreased after exposure to 2.5 mg/L of Biocarb(®) and AST increased at 2.5 mg/L of Orium(®), Original(®) and 5.0 mg/L of Biocarb(®) in comparison with the control group. In summary, the oxidative damage generated by GLY may have caused the increased formation of free radicals that led to the histological alterations observed in hepatocytes.

The relationship between aerosol particles chemical composition and optical properties to identify the biomass burning contribution to fine particles concentration: a case study for São Paulo city, Brazil.

The air quality in the Metropolitan Area of São Paulo (MASP) is primarily determined by the local pollution source contribution, mainly the vehicular fleet, but there is a concern about the role of remote sources to the fine mode particles (PM2.5) concentration and composition. One of the most important remote sources of atmospheric aerosol is the biomass burning emissions from São Paulo state's inland and from the central and north portions of Brazil. This study presents a synergy of different measurements of atmospheric aerosol chemistry and optical properties in the MASP in order to show how they can be used as a tool to identify particles from local and remote sources. For the clear identification of the local and remote source contribution, aerosol properties measurements at surface level were combined with vertical profiles information. Over 15 days in the austral winter of 2012, particulate matter (PM) was collected using a cascade impactor and a Partisol sampler in São Paulo City. Mass concentrations were determined by gravimetry, black carbon concentrations by reflectance, and trace element concentrations by X-ray fluorescence. Aerosol optical properties were studied using a multifilter rotating shadowband radiometer (MFRSR), a Lidar system and satellite data. Optical properties, concentrations, size distributions, and elemental composition of atmospheric particles were strongly related and varied according to meteorological conditions. During the sampling period, PM mean mass concentrations were 17.4 ± 10.1 and 15.3 ± 6.9 µg/m3 for the fine and coarse fractions, respectively. The mean aerosol optical depths at 415 nm and Ångström exponent (AE) over the whole period were 0.29 ± 0.14 and 1.35 ± 0.11, respectively. Lidar ratios reached values of 75 sr. The analyses of the impacts of an event of biomass burning smoke transport to the São Paulo city revealed significant changing on local aerosol concentrations and optical parameters. The identification of the source contributions, local and remote, to the fine particles in MASP can be more precisely achieved when particle size composition and distribution, vertical profile of aerosols, and air mass trajectories are analyzed in combination.

Concept-drifts adaptation for machine learning EEG epilepsy seizure prediction.

Seizure prediction remains a challenge, with approximately 30% of patients unresponsive to conventional treatments. Addressing this issue is crucial for improving patients' quality of life, as timely intervention can mitigate the impact of seizures. In this research field, it is critical to identify the preictal interval, the transition from regular brain activity to a seizure. While previous studies have explored various Electroencephalogram (EEG) based methodologies for prediction, few have been clinically applicable. Recent studies have underlined the dynamic nature of EEG data, characterised by data changes with time, known as concept drifts, highlighting the need for automated methods to detect and adapt to these changes. In this study, we investigate the effectiveness of automatic concept drift adaptation methods in seizure prediction. Three patient-specific seizure prediction approaches with a 10-minute prediction horizon are compared: a seizure prediction algorithm incorporating a window adjustment method by optimising performance with Support Vector Machines (Backwards-Landmark Window), a seizure prediction algorithm incorporating a data-batch (seizures) selection method using a logistic regression (Seizure-batch Regression), and a seizure prediction algorithm with a dynamic integration of classifiers (Dynamic Weighted Ensemble). These methods incorporate a retraining process after each seizure and use a combination of univariate linear features and SVM classifiers. The Firing Power was used as a post-processing technique to generate alarms before seizures. These methodologies were compared with a control approach based on the typical machine learning pipeline, considering a group of 37 patients with Temporal Lobe Epilepsy from the EPILEPSIAE database. The best-performing approach (Backwards-Landmark Window) achieved results of 0.75 ± 0.33 for sensitivity and 1.03 ± 1.00 for false positive rate per hour. This new strategy performed above chance for 89% of patients with the surrogate predictor, whereas the control approach only validated 46%.

Classifier combination supported by the sleep-wake cycle improves EEG seizure prediction performance.

OBJECTIVE: Seizure prediction is a promising solution to improve the quality of life for drug-resistant patients, which concerns nearly 30% of patients with epilepsy. The present study aimed to ascertain the impact of incorporating sleep-wake information in seizure prediction. METHODS: We developed five patient-specific prediction approaches that use vigilance state information differently: i) using it as an input feature, ii) building a pool of two classifiers, each with different weights to sleep/wake training samples, iii) building a pool of two classifiers, each with only sleep/wake samples, iv) changing the alarm-threshold concerning each sleep/wake state, and v) adjusting the alarm-threshold after a sleep-wake transition. We compared these approaches with a control method that did not integrate sleep-wake information. Our models were tested with data (43 seizures and 482 hours) acquired during presurgical monitoring of 17 patients from the EPILEPSIAE database. As EPILEPSIAE does not contain vigilance state annotations, we developed a sleep-wake classifier using 33 patients diagnosed with nocturnal frontal lobe epilepsy from the CAP Sleep database. RESULTS: Although different patients may require different strategies, our best approach, the pool of weighted predictors, obtained 65% of patients performing above chance level with a surrogate analysis (against 41% in the control method). CONCLUSION: The inclusion of vigilance state information improves seizure prediction. Higher results and testing with longterm recordings from daily-life conditions are necessary to ensure clinical acceptance. SIGNIFICANCE: As automated sleep-wake detection is possible, it would be feasible to incorporate these algorithms into future devices for seizure prediction.

EEG epilepsy seizure prediction: the post-processing stage as a chronology.

Almost one-third of epileptic patients fail to achieve seizure control through anti-epileptic drug administration. In the scarcity of completely controlling a patient's epilepsy, seizure prediction plays a significant role in clinical management and providing new therapeutic options such as warning or intervention devices. Seizure prediction algorithms aim to identify the preictal period that Electroencephalogram (EEG) signals can capture. However, this period is associated with substantial heterogeneity, varying among patients or even between seizures from the same patient. The present work proposes a patient-specific seizure prediction algorithm using post-processing techniques to explore the existence of a set of chronological events of brain activity that precedes epileptic seizures. The study was conducted with 37 patients with Temporal Lobe Epilepsy (TLE) from the EPILEPSIAE database. The designed methodology combines univariate linear features with a classifier based on Support Vector Machines (SVM) and two post-processing techniques to handle pre-seizure temporality in an easily explainable way, employing knowledge from network theory. In the Chronological Firing Power approach, we considered the preictal as a sequence of three brain activity events separated in time. In the Cumulative Firing Power approach, we assumed the preictal period as a sequence of three overlapping events. These methodologies were compared with a control approach based on the typical machine learning pipeline. We considered a Seizure Prediction horizon (SPH) of 5 mins and analyzed several values for the Seizure Occurrence Period (SOP) duration, between 10 and 55 mins. Our results showed that the Cumulative Firing Power approach may improve the seizure prediction performance. This new strategy performed above chance for 62% of patients, whereas the control approach only validated 49% of its models.

Addressing data limitations in seizure prediction through transfer learning.

According to the literature, seizure prediction models should be developed following a patient-specific approach. However, seizures are usually very rare events, meaning the number of events that may be used to optimise seizure prediction approaches is limited. To overcome such constraint, we analysed the possibility of using data from patients from an external database to improve patient-specific seizure prediction models. We present seizure prediction models trained using a transfer learning procedure. We trained a deep convolutional autoencoder using electroencephalogram data from 41 patients collected from the EPILEPSIAE database. Then, a bidirectional long short-term memory and a classifier layers were added on the top of the encoder part and were optimised for 24 patients from the Universitätsklinikum Freiburg individually. The encoder was used as a feature extraction module. Therefore, its weights were not changed during the patient-specific training. Experimental results showed that seizure prediction models optimised using pretrained weights present about four times fewer false alarms while maintaining the same ability to predict seizures and achieved more 13% validated patients. Therefore, results evidenced that the optimisation using transfer learning was more stable and faster, saving computational resources. In summary, adopting transfer learning for seizure prediction models represents a significant advancement. It addresses the data limitation seen in the seizure prediction field and offers more efficient and stable training, conserving computational resources. Additionally, despite the compact size, transfer learning allows to easily share data knowledge due to fewer ethical restrictions and lower storage requirements. The convolutional autoencoder developed in this study will be shared with the scientific community, promoting further research.

Removing artefacts and periodically retraining improve performance of neural network-based seizure prediction models.

The development of seizure prediction models is often based on long-term scalp electroencephalograms (EEGs) since they capture brain electrical activity, are non-invasive, and come at a relatively low-cost. However, they suffer from major shortcomings. First, long-term EEG is usually highly contaminated with artefacts. Second, changes in the EEG signal over long intervals, known as concept drift, are often neglected. We evaluate the influence of these problems on deep neural networks using EEG time series and on shallow neural networks using widely-used EEG features. Our patient-specific prediction models were tested in 1577 hours of continuous EEG, containing 91 seizures from 41 patients with temporal lobe epilepsy who were undergoing pre-surgical monitoring. Our results showed that cleaning EEG data, using a previously developed artefact removal method based on deep convolutional neural networks, improved prediction performance. We also found that retraining the models over time reduced false predictions. Furthermore, the results show that although deep neural networks processing EEG time series are less susceptible to false alarms, they may need more data to surpass feature-based methods. These findings highlight the importance of robust data denoising and periodic adaptation of seizure prediction models.

The goal of explaining black boxes in EEG seizure prediction is not to explain models' decisions.

Many state-of-the-art methods for seizure prediction, using the electroencephalogram, are based on machine learning models that are black boxes, weakening the trust of clinicians in them for high-risk decisions. Seizure prediction concerns a multidimensional time-series problem that performs continuous sliding window analysis and classification. In this work, we make a critical review of which explanations increase trust in models' decisions for predicting seizures. We developed three machine learning methodologies to explore their explainability potential. These contain different levels of model transparency: a logistic regression, an ensemble of 15 support vector machines, and an ensemble of three convolutional neural networks. For each methodology, we evaluated quasi-prospectively the performance in 40 patients (testing data comprised 2055 hours and 104 seizures). We selected patients with good and poor performance to explain the models' decisions. Then, with grounded theory, we evaluated how these explanations helped specialists (data scientists and clinicians working in epilepsy) to understand the obtained model dynamics. We obtained four lessons for better communication between data scientists and clinicians. We found that the goal of explainability is not to explain the system's decisions but to improve the system itself. Model transparency is not the most significant factor in explaining a model decision for seizure prediction. Even when using intuitive and state-of-the-art features, it is hard to understand brain dynamics and their relationship with the developed models. We achieve an increase in understanding by developing, in parallel, several systems that explicitly deal with signal dynamics changes that help develop a complete problem formulation.

Unsupervised EEG preictal interval identification in patients with drug-resistant epilepsy.

Typical seizure prediction models aim at discriminating interictal brain activity from pre-seizure electrographic patterns. Given the lack of a preictal clinical definition, a fixed interval is widely used to develop these models. Recent studies reporting preictal interval selection among a range of fixed intervals show inter- and intra-patient preictal interval variability, possibly reflecting the heterogeneity of the seizure generation process. Obtaining accurate labels of the preictal interval can be used to train supervised prediction models and, hence, avoid setting a fixed preictal interval for all seizures within the same patient. Unsupervised learning methods hold great promise for exploring preictal alterations on a seizure-specific scale. Multivariate and univariate linear and nonlinear features were extracted from scalp electroencephalography (EEG) signals collected from 41 patients with drug-resistant epilepsy undergoing presurgical monitoring. Nonlinear dimensionality reduction was performed for each group of features and each of the 226 seizures. We applied different clustering methods in searching for preictal clusters located until 2 h before the seizure onset. We identified preictal patterns in 90% of patients and 51% of the visually inspected seizures. The preictal clusters manifested a seizure-specific profile with varying duration (22.9 ± 21.0 min) and starting time before seizure onset (47.6 ± 27.3 min). Searching for preictal patterns on the EEG trace using unsupervised methods showed that it is possible to identify seizure-specific preictal signatures for some patients and some seizures within the same patient.

Impact of extreme wildfires from the Brazilian Forests and sugarcane burning on the air quality of the biggest megacity on South America.

Recently, extreme wildfires have damaged important ecosystems worldwide and have affected urban areas miles away due to long-range transport of smoke plumes. We performed a comprehensive analysis to clarify how smoke plumes from Pantanal and Amazon forests wildfires and sugarcane harvest burning also from interior of the state of São Paulo (ISSP) were transported and injected into the atmosphere of the Metropolitan Area of São Paulo (MASP), where they worsened air quality and increased greenhouse gas (GHG) levels. To classify event days, multiple biomass burning fingerprints as carbon isotopes, Lidar ratio and specific compounds ratios were combined with back trajectories modeling. During smoke plume event days in the MASP fine particulate matter concentrations exceeded the WHO standard (>25 µg m-3), at 99 % of the air quality monitoring stations, and peak CO2 excess were 100 % to 1178 % higher than non-event days. We demonstrated how external pollution events such as wildfires pose an additional challenge for cities, regarding public health threats associated to air quality, and reinforces the importance of GHG monitoring networks to track local and remote GHG emissions and sources in urban areas.

Ensemble Deep Neural Network for Automatic Classification of EEG Independent Components.

OBJECTIVE: Independent component analysis (ICA) is commonly used to remove noisy artifacts from multi-channel scalp electroencephalogram (EEG) signals. ICA decomposes EEG into different independent components (ICs) and then, experts remove the noisy ones. This process is highly time-consuming and experts are not always available. To surpass this drawback, research is going on to develop models to automatically conduct IC classification. Current state-of-the-art models use power spectrum densities (PSDs) and topoplots to classify ICs. The performance of these methods may be limited by disregarding the IC time-series that would contribute to fully simulate the visual inspection performed by experts. METHODS: We present a novel ensemble deep neural network that combines time-series, PSDs, and topoplots to classify ICs. Moreover, we study the ability to use our model in transfer learning approaches. RESULTS: Experimental results showed that using time-series improves IC classification. Results also indicated that transfer learning obtained higher performance than simply training a new model from scratch. CONCLUSION: Researchers should develop IC classifiers using the three sources of information. Moreover, transfer learning approaches should be considered when producing new deep learning models. SIGNIFICANCE: This work improves IC classification, enhancing the automatic removal of EEG artifacts. Additionally, since labelled ICs are generally not publicly available, the possibility of using our model in transfer learning studies may motivate other researchers to develop their own classifiers.

Interpretable EEG seizure prediction using a multiobjective evolutionary algorithm.

Seizure prediction might be the solution to tackle the apparent unpredictability of seizures in patients with drug-resistant epilepsy, which comprise about a third of all patients with epilepsy. Designing seizure prediction models involves defining the pre-ictal period, a transition stage between inter-ictal brain activity and the seizure discharge. This period is typically a fixed interval, with some recent studies reporting the evaluation of different patient-specific pre-ictal intervals. Recently, researchers have aimed to determine the pre-ictal period, a transition stage between regular brain activity and a seizure. Authors have been using deep learning models given the ability of such models to automatically perform pre-processing, feature extraction, classification, and handling temporal and spatial dependencies. As these approaches create black-box models, clinicians may not have sufficient trust to use them in high-stake decisions. By considering these problems, we developed an evolutionary seizure prediction model that identifies the best set of features while automatically searching for the pre-ictal period and accounting for patient comfort. This methodology provides patient-specific interpretable insights, which might contribute to a better understanding of seizure generation processes and explain the algorithm's decisions. We tested our methodology on 238 seizures and 3687 h of continuous data, recorded on scalp recordings from 93 patients with several types of focal and generalised epilepsies. We compared the results with a seizure surrogate predictor and obtained a performance above chance for 32% patients. We also compared our results with a control method based on the standard machine learning pipeline (pre-processing, feature extraction, classifier training, and post-processing), where the control marginally outperformed our approach by validating 35% of the patients. In total, 54 patients performed above chance for at least one method: our methodology or the control one. Of these 54 patients, 21 ([Formula: see text]38%) were solely validated by our methodology, while 24 ([Formula: see text]44%) were only validated by the control method. These findings may evidence the need for different methodologies concerning different patients.

EPIC: Annotated epileptic EEG independent components for artifact reduction.

Scalp electroencephalogram is a non-invasive multi-channel biosignal that records the brain's electrical activity. It is highly susceptible to noise that might overshadow important data. Independent component analysis is one of the most used artifact removal methods. Independent component analysis separates data into different components, although it can not automatically reject the noisy ones. Therefore, experts are needed to decide which components must be removed before reconstructing the data. To automate this method, researchers have developed classifiers to identify noisy components. However, to build these classifiers, they need annotated data. Manually classifying independent components is a time-consuming task. Furthermore, few labelled data are publicly available. This paper presents a source of annotated electroencephalogram independent components acquired from patients with epilepsy (EPIC Dataset). This dataset contains 77,426 independent components obtained from approximately 613 hours of electroencephalogram, visually inspected by two experts, which was already successfully utilised to develop independent component classifiers.

On the clinical acceptance of black-box systems for EEG seizure prediction.

Seizure prediction may be the solution for epileptic patients whose drugs and surgery do not control seizures. Despite 46 years of research, few devices/systems underwent clinical trials and/or are commercialized, where the most recent state-of-the-art approaches, as neural networks models, are not used to their full potential. The latter demonstrates the existence of social barriers to new methodologies due to data bias, patient safety, and legislation compliance. In the form of literature review, we performed a qualitative study to analyze the seizure prediction ecosystem to find these social barriers. With the Grounded Theory, we draw hypotheses from data, while with the Actor-Network Theory we considered that technology shapes social configurations and interests, being fundamental in healthcare. We obtained a social network that describes the ecosystem and propose research guidelines aiming at clinical acceptance. Our most relevant conclusion is the need for model explainability, but not necessarily intrinsically interpretable models, for the case of seizure prediction. Accordingly, we argue that it is possible to develop robust prediction models, including black-box systems to some extent, while avoiding data bias, ensuring patient safety, and still complying with legislation, if they can deliver human- comprehensible explanations. Due to skepticism and patient safety reasons, many authors advocate the use of transparent models which may limit their performance and potential. Our study highlights a possible path, by using model explainability, on how to overcome these barriers while allowing the use of more computationally robust models.