Reducing False Alarms in Wearable Seizure Detection With EEGformer: A Compact Transformer Model for MCUs.

The long-term, continuous analysis of electroencephalography (EEG) signals on wearable devices to automatically detect seizures in epileptic patients is a high-potential application field for deep neural networks, and specifically for transformers, which are highly suited for end-to-end time series processing without handcrafted feature extraction. In this work, we propose a small-scale transformer detector, the EEGformer, compatible with unobtrusive acquisition setups that use only the temporal channels. EEGformer is the result of a hardware-oriented design exploration, aiming for efficient execution on tiny low-power micro-controller units (MCUs) and low latency and false alarm rate to increase patient and caregiver acceptance.Tests conducted on the CHB-MIT dataset show a 20% reduction of the onset detection latency with respect to the state-of-the-art model for temporal acquisition, with a competitive 73% seizure detection probability and 0.15 false-positive-per-hour (FP/h). Further investigations on a novel and challenging scalp EEG dataset result in the successful detection of 88% of the annotated seizure events, with 0.45 FP/h.We evaluate the deployment of the EEGformer on three commercial low-power computing platforms: the single-core Apollo4 MCU and the GAP8 and GAP9 parallel MCUs. The most efficient implementation (on GAP9) results in as low as 13.7 ms and 0.31 mJ per inference, demonstrating the feasibility of deploying the EEGformer on wearable seizure detection systems with reduced channel count and multi-day battery duration.

A dynamic clamping approach using in silico IK1 current for discrimination of chamber-specific hiPSC-derived cardiomyocytes.

Human induced pluripotent stem cell (hiPSC)-derived cardiomyocytes (CM) constitute a mixed population of ventricular-, atrial-, nodal-like cells, limiting the reliability for studying chamber-specific disease mechanisms. Previous studies characterised CM phenotype based on action potential (AP) morphology, but the classification criteria were still undefined. Our aim was to use in silico models to develop an automated approach for discriminating the electrophysiological differences between hiPSC-CM. We propose the dynamic clamp (DC) technique with the injection of a specific IK1 current as a tool for deriving nine electrical biomarkers and blindly classifying differentiated CM. An unsupervised learning algorithm was applied to discriminate CM phenotypes and principal component analysis was used to visualise cell clustering. Pharmacological validation was performed by specific ion channel blocker and receptor agonist. The proposed approach improves the translational relevance of the hiPSC-CM model for studying mechanisms underlying inherited or acquired atrial arrhythmias in human CM, and for screening anti-arrhythmic agents.

Reducing the Energy Consumption of sEMG-Based Gesture Recognition at the Edge Using Transformers and Dynamic Inference.

Hand gesture recognition applications based on surface electromiographic (sEMG) signals can benefit from on-device execution to achieve faster and more predictable response times and higher energy efficiency. However, deploying state-of-the-art deep learning (DL) models for this task on memory-constrained and battery-operated edge devices, such as wearables, requires a careful optimization process, both at design time, with an appropriate tuning of the DL models' architectures, and at execution time, where the execution of large and computationally complex models should be avoided unless strictly needed. In this work, we pursue both optimization targets, proposing a novel gesture recognition system that improves upon the state-of-the-art models both in terms of accuracy and efficiency. At the level of DL model architecture, we apply for the first time tiny transformer models (which we call bioformers) to sEMG-based gesture recognition. Through an extensive architecture exploration, we show that our most accurate bioformer achieves a higher classification accuracy on the popular Non-Invasive Adaptive hand Prosthetics Database 6 (Ninapro DB6) dataset compared to the state-of-the-art convolutional neural network (CNN) TEMPONet (+3.1%). When deployed on the RISC-V-based low-power system-on-chip (SoC) GAP8, bioformers that outperform TEMPONet in accuracy consume 7.8×-44.5× less energy per inference. At runtime, we propose a three-level dynamic inference approach that combines a shallow classifier, i.e., a random forest (RF) implementing a simple "rest detector" with two bioformers of different accuracy and complexity, which are sequentially applied to each new input, stopping the classification early for "easy" data. With this mechanism, we obtain a flexible inference system, capable of working in many different operating points in terms of accuracy and average energy consumption. On GAP8, we obtain a further 1.03×-1.35× energy reduction compared to static bioformers at iso-accuracy.

Identification of a serum and urine extracellular vesicle signature predicting renal outcome after kidney transplant.

BACKGROUND: A long-standing effort is dedicated towards the identification of biomarkers allowing the prediction of graft outcome after kidney transplant. Extracellular vesicles (EVs) circulating in body fluids represent an attractive candidate, as their cargo mirrors the originating cell and its pathophysiological status. The aim of the study was to investigate EV surface antigens as potential predictors of renal outcome after kidney transplant. METHODS: We characterized 37 surface antigens by flow cytometry, in serum and urine EVs from 58 patients who were evaluated before, and at 10-14 days, 3 months and 1 year after transplant, for a total of 426 analyzed samples. The outcome was defined according to estimated glomerular filtration rate (eGFR) at 1 year. RESULTS: Endothelial cells and platelets markers (CD31, CD41b, CD42a and CD62P) in serum EVs were higher at baseline in patients with persistent kidney dysfunction at 1 year, and progressively decreased after kidney transplant. Conversely, mesenchymal progenitor cell marker (CD1c, CD105, CD133, SSEEA-4) in urine EVs progressively increased after transplant in patients displaying renal recovery at follow-up. These markers correlated with eGFR, creatinine and proteinuria, associated with patient outcome at univariate analysis and were able to predict patient outcome at receiver operating characteristics curves analysis. A specific EV molecular signature obtained by supervised learning correctly classified patients according to 1-year renal outcome. CONCLUSIONS: An EV-based signature, reflecting the cardiovascular profile of the recipient, and the repairing/regenerative features of the graft, could be introduced as a non-invasive tool for a tailored management of follow-up of patients undergoing kidney transplant.

Prediction of All-Cause Mortality Following Percutaneous Coronary Intervention in Bifurcation Lesions Using Machine Learning Algorithms.

Stratifying prognosis following coronary bifurcation percutaneous coronary intervention (PCI) is an unmet clinical need that may be fulfilled through the adoption of machine learning (ML) algorithms to refine outcome predictions. We sought to develop an ML-based risk stratification model built on clinical, anatomical, and procedural features to predict all-cause mortality following contemporary bifurcation PCI. Multiple ML models to predict all-cause mortality were tested on a cohort of 2393 patients (training, n = 1795; internal validation, n = 598) undergoing bifurcation PCI with contemporary stents from the real-world RAIN registry. Twenty-five commonly available patient-/lesion-related features were selected to train ML models. The best model was validated in an external cohort of 1701 patients undergoing bifurcation PCI from the DUTCH PEERS and BIO-RESORT trial cohorts. At ROC curves, the AUC for the prediction of 2-year mortality was 0.79 (0.74-0.83) in the overall population, 0.74 (0.62-0.85) at internal validation and 0.71 (0.62-0.79) at external validation. Performance at risk ranking analysis, k-center cross-validation, and continual learning confirmed the generalizability of the models, also available as an online interface. The RAIN-ML prediction model represents the first tool combining clinical, anatomical, and procedural features to predict all-cause mortality among patients undergoing contemporary bifurcation PCI with reliable performance.

Risk stratification of patients with SARS-CoV-2 by tissue factor expression in circulating extracellular vesicles.

Inflammatory response following SARS-CoV-2 infection results in substantial increase of amounts of intravascular pro-coagulant extracellular vesicles (EVs) expressing tissue factor (CD142) on their surface. CD142-EV turned out to be useful as diagnostic biomarker in COVID-19 patients. Here we aimed at studying the prognostic capacity of CD142-EV in SARS-CoV-2 infection. Expression of CD142-EV was evaluated in 261 subjects admitted to hospital for pneumonia and with a positive molecular test for SARS-CoV-2. The study population consisted of a discovery cohort of selected patients (n = 60) and an independent validation cohort including unselected consecutive enrolled patients (n = 201). CD142-EV levels were correlated with post-hospitalization course of the disease and compared to the clinically available 4C Mortality Score as referral. CD142-EV showed a reliable performance to predict patient prognosis in the discovery cohort (AUC = 0.906) with an accuracy of 81.7%, that was confirmed in the validation cohort (AUC = 0.736). Kaplan-Meier curves highlighted a high discrimination power in unselected subjects with CD142-EV being able to stratify the majority of patients according to their prognosis. We obtained a comparable accuracy, being not inferior in terms of prediction of patients' prognosis and risk of mortality, with 4C Mortality Score. The expression of surface vesicular CD142 and its reliability as prognostic marker was technically validated using different immunocapture strategies and assays. The detection of CD142 on EV surface gains considerable interest as risk stratification tool to support clinical decision making in COVID-19.

Supervised and unsupervised learning to define the cardiovascular risk of patients according to an extracellular vesicle molecular signature.

Cardiovascular (CV) disease represents the most common cause of death in developed countries. Risk assessment is highly relevant to intervene at individual level and implement prevention strategies. Circulating extracellular vesicles (EVs) are involved in the development and progression of CV diseases and are considered promising biomarkers. We aimed at identifying an EV signature to improve the stratification of patients according to CV risk and likelihood to develop fatal CV events. EVs were characterized by nanoparticle tracking analysis and flow cytometry for a standardized panel of 37 surface antigens in a cross-sectional multicenter cohort (n = 486). CV profile was defined by presence of different indicators (age, sex, body mass index, hypertension, hyperlipidemia, diabetes, coronary artery disease, cardiac heart failure, chronic kidney disease, smoking habit, organ damage) and according to the 10-year risk of fatal CV events estimated using SCORE charts of European Society of Cardiology. By combining expression levels of EV antigens using unsupervised learning, patients were classified into 3 clusters: Cluster-I (n = 288), Cluster-II (n = 83), Cluster-III (n = 30). A separate analysis was conducted on patients displaying acute CV events (n = 82). Prevalence of hypertension, diabetes, chronic heart failure, and organ damage (defined as left ventricular hypertrophy and/or microalbuminuria) increased progressively from Cluster-I to Cluster-III. Several EV antigens, including markers for platelets (CD41b-CD42a-CD62P), leukocytes (CD1c-CD2-CD3-CD4-CD8-CD14-CD19-CD20-CD25-CD40-CD45-CD69-CD86), and endothelium (CD31-CD105) were independently associated with CV risk indicators and correlated to age, blood pressure, glucometabolic profile, renal function, and SCORE risk. EV profiling, obtained from minimally invasive blood sampling, allows accurate patient stratification according to CV risk profile.

Q-PPG: Energy-Efficient PPG-Based Heart Rate Monitoring on Wearable Devices.

Hearth Rate (HR) monitoring is increasingly performed in wrist-worn devices using low-cost photoplethysmography (PPG) sensors. However, Motion Artifacts (MAs) caused by movements of the subject's arm affect the performance of PPG-based HR tracking. This is typically addressed coupling the PPG signal with acceleration measurements from an inertial sensor. Unfortunately, most standard approaches of this kind rely on hand-tuned parameters, which impair their generalization capabilities and their applicability to real data in the field. In contrast, methods based on deep learning, despite their better generalization, are considered to be too complex to deploy on wearable devices. In this work, we tackle these limitations, proposing a design space exploration methodology to automatically generate a rich family of deep Temporal Convolutional Networks (TCNs) for HR monitoring, all derived from a single "seed" model. Our flow involves a cascade of two Neural Architecture Search (NAS) tools and a hardware-friendly quantizer, whose combination yields both highly accurate and extremely lightweight models. When tested on the PPG-Dalia dataset, our most accurate model sets a new state-of-the-art in Mean Absolute Error. Furthermore, we deploy our TCNs on an embedded platform featuring a STM32WB55 microcontroller, demonstrating their suitability for real-time execution. Our most accurate quantized network achieves 4.41 Beats Per Minute (BPM) of Mean Absolute Error (MAE), with an energy consumption of 47.65 mJ and a memory footprint of 412 kB. At the same time, the smallest network that obtains a MAE 8 BPM, among those generated by our flow, has a memory footprint of 1.9 kB and consumes just 1.79 mJ per inference.

Clinical Score and Machine Learning-Based Model to Predict Diagnosis of Primary Aldosteronism in Arterial Hypertension.

[Figure: see text].

Extracellular Vesicle Surface Markers as a Diagnostic Tool in Transient Ischemic Attacks.

Background and Purpose: Extracellular vesicles (EVs) are promising biomarkers for cerebral ischemic diseases, but not systematically tested in patients with transient ischemic attacks (TIAs). We aimed at (1) investigating the profile of EV-surface antigens in patients with symptoms suspicious for TIA; (2) developing and validating a predictive model for TIA diagnosis based on a specific EV-surface antigen profile. Methods: We analyzed 40 subjects with symptoms suspicious for TIA and 20 healthy controls from a training cohort. An independent cohort of 28 subjects served as external validation. Patients were stratified according to likelihood of having a real ischemic event using the Precise Diagnostic Score, defined as: unlikely (score 0­1), possible-probable (score 2­3), or very likely (score 4­8). Serum vesicles were quantified by nanoparticle tracking analysis and EV-surface antigen profile characterized by multiplex flow cytometry. Results: EV concentration increased in patients with very likely or possible-probable TIA (P<0.05) compared with controls. Nanoparticle concentration was directly correlated with the Precise Diagnostic score (R=0.712; P<0.001). After EV immuno-capturing, CD8, CD2, CD62P, melanoma-associated chondroitin sulfate proteoglycan, CD42a, CD44, CD326, CD142, CD31, and CD14 were identified as discriminants between groups. Receiver operating characteristic curve analysis confirmed a reliable diagnostic performance for each of these markers taken individually and for a compound marker derived from their linear combinations (area under the curve, 0.851). Finally, a random forest model combining the expression levels of selected markers achieved an accuracy of 96% and 78.9% for discriminating patients with a very likely TIA, in the training and external validation cohort, respectively. Conclusions: The EV-surface antigen profile appears to be different in patients with transient symptoms adjudicated to be very likely caused by brain ischemia compared with patients whose symptoms were less likely to due to brain ischemia. We propose an algorithm based on an EV-surface-antigen specific signature that might aid in the recognition of TIA.

Circulating extracellular vesicles are endowed with enhanced procoagulant activity in SARS-CoV-2 infection.

BACKGROUND: Coronavirus-2 (SARS-CoV-2) infection causes an acute respiratory syndrome accompanied by multi-organ damage that implicates a prothrombotic state leading to widespread microvascular clots. The causes of such coagulation abnormalities are unknown. The receptor tissue factor, also known as CD142, is often associated with cell-released extracellular vesicles (EV). In this study, we aimed to characterize surface antigens profile of circulating EV in COVID-19 patients and their potential implication as procoagulant agents. METHODS: We analyzed serum-derived EV from 67 participants who underwent nasopharyngeal swabs molecular test for suspected SARS-CoV-2 infection (34 positives and 33 negatives) and from 16 healthy controls (HC), as referral. A sub-analysis was performed on subjects who developed pneumonia (n = 28). Serum-derived EV were characterized for their surface antigen profile and tested for their procoagulant activity. A validation experiment was performed pre-treating EV with anti-CD142 antibody or with recombinant FVIIa. Serum TNF-α levels were measured by ELISA. FINDINGS: Profiling of EV antigens revealed a surface marker signature that defines circulating EV in COVID-19. A combination of seven surface molecules (CD49e, CD209, CD86, CD133/1, CD69, CD142, and CD20) clustered COVID (+) versus COVID (-) patients and HC. CD142 showed the highest discriminating performance at both multivariate models and ROC curve analysis. Noteworthy, we found that CD142 exposed onto surface of EV was biologically active. CD142 activity was higher in COVID (+) patients and correlated with TNF-α serum levels. INTERPRETATION: In SARS-CoV-2 infection the systemic inflammatory response results in cell-release of substantial amounts of procoagulant EV that may act as clotting initiation agents, contributing to disease severity. FUNDING: Cardiocentro Ticino Institute, Ente ospedaliero Cantonale, Lugano-Switzerland.

Profiling Inflammatory Extracellular Vesicles in Plasma and Cerebrospinal Fluid: An Optimized Diagnostic Model for Parkinson's Disease.

Extracellular vesicles (EVs) play a central role in intercellular communication, which is relevant for inflammatory and immune processes implicated in neurodegenerative disorders, such as Parkinson's Disease (PD). We characterized and compared distinctive cerebrospinal fluid (CSF)-derived EVs in PD and atypical parkinsonisms (AP), aiming to integrate a diagnostic model based on immune profiling of plasma-derived EVs via artificial intelligence. Plasma- and CSF-derived EVs were isolated from patients with PD, multiple system atrophy (MSA), AP with tauopathies (AP-Tau), and healthy controls. Expression levels of 37 EV surface markers were measured by a flow cytometric bead-based platform and a diagnostic model based on expression of EV surface markers was built by supervised learning algorithms. The PD group showed higher amount of CSF-derived EVs than other groups. Among the 17 EV surface markers differentially expressed in plasma, eight were expressed also in CSF of a subgroup of PD, 10 in MSA, and 6 in AP-Tau. A two-level random forest model was built using EV markers co-expressed in plasma and CSF. The model discriminated PD from non-PD patients with high sensitivity (96.6%) and accuracy (92.6%). EV surface marker characterization bolsters the relevance of inflammation in PD and it underscores the role of EVs as pathways/biomarkers for protein aggregation-related neurodegenerative diseases.

Development of a Prediction Score to Avoid Confirmatory Testing in Patients With Suspected Primary Aldosteronism.

CONTEXT: The diagnostic work-up of primary aldosteronism (PA) includes screening and confirmation steps. Case confirmation is time-consuming, expensive, and there is no consensus on tests and thresholds to be used. Diagnostic algorithms to avoid confirmatory testing may be useful for the management of patients with PA. OBJECTIVE: Development and validation of diagnostic models to confirm or exclude PA diagnosis in patients with a positive screening test. DESIGN, PATIENTS, AND SETTING: We evaluated 1024 patients who underwent confirmatory testing for PA. The diagnostic models were developed in a training cohort (n = 522), and then tested on an internal validation cohort (n = 174) and on an independent external prospective cohort (n = 328). MAIN OUTCOME MEASURE: Different diagnostic models and a 16-point score were developed by machine learning and regression analysis to discriminate patients with a confirmed diagnosis of PA. RESULTS: Male sex, antihypertensive medication, plasma renin activity, aldosterone, potassium levels, and the presence of organ damage were associated with a confirmed diagnosis of PA. Machine learning-based models displayed an accuracy of 72.9%-83.9%. The Primary Aldosteronism Confirmatory Testing (PACT) score correctly classified 84.1% at training and 83.9% or 81.1% at internal and external validation, respectively. A flow chart employing the PACT score to select patients for confirmatory testing correctly managed all patients and resulted in a 22.8% reduction in the number of confirmatory tests. CONCLUSIONS: The integration of diagnostic modeling algorithms in clinical practice may improve the management of patients with PA by circumventing unnecessary confirmatory testing.

An Ensemble of Hyperdimensional Classifiers: Hardware-Friendly Short-Latency Seizure Detection With Automatic iEEG Electrode Selection.

We propose a new algorithm for detecting epileptic seizures. Our algorithm first extracts three features, namely mean amplitude, line length, and local binary patterns that are fed to an ensemble of classifiers using hyperdimensional (HD) computing. These features are embedded into prototype vectors representing ictal (during seizures) and interictal (between seizures) brain states are constructed. These vectors can be computed at different spatial scales ranging from a single electrode up to many electrodes. This flexibility allows our algorithm to identify the electrodes that discriminate best between ictal and interictal brain states. We assess our algorithm on the SWEC-ETHZ iEEG dataset that includes 99 short-time iEEG seizures recorded with 36 to 100 electrodes from 16 drug-resistant epilepsy patients. Using k-fold cross-validation and all electrodes, our algorithm surpasses state-of-the-art algorithms yielding significantly shorter latency (8.81 s vs. 11.57 s) in seizure onset detection, and higher specificity (97.31% vs. 94.84%) and accuracy (96.85% vs. 95.42%). We can further reduce the latency of our algorithm to 3.74 s by allowing a slightly higher percentage of false alarms (2% specificity loss). Using only the top 10% of the electrodes ranked by our algorithm, we still maintain superior latency, sensitivity, and specificity compared to the other algorithms with all the electrodes. We finally demonstrate the suitability of our algorithm to deployment on low-cost embedded hardware platforms, thanks to its robustness to noise/artifacts affecting the signal, its low computational complexity, and the small memory-footprint on a RISC-V microcontroller.

Prediction of hyperaldosteronism subtypes when adrenal vein sampling is unilaterally successful.

OBJECTIVE: Adrenal venous sampling (AVS) is the gold standard to discriminate patients with unilateral primary aldosteronism (UPA) from bilateral disease (BPA). AVS is technically demanding and in cases of unsuccessful cannulation of adrenal veins, the results may not always be interpreted. The aim of our study was to develop diagnostic models to distinguish UPA from BPA, in cases of unilateral successful AVS and the presence of contralateral suppression of aldosterone secretion. DESIGN: Retrospective evaluation of 158 patients referred to a tertiary hypertension unit who underwent AVS. We randomly assigned 110 patients to a training cohort and 48 patients to a validation cohort to develop and test the diagnostic models. METHODS: Supervised machine learning algorithms and regression models were used to develop and validate two prediction models and a simple 19-point score system to stratify patients according to their subtype diagnosis. RESULTS: Aldosterone levels at screening and after confirmatory testing, lowest potassium, ipsilateral and contralateral imaging findings at CT scanning, and contralateral ratio at AVS, were associated with a diagnosis of UPA and were included in the diagnostic models. Machine learning algorithms correctly classified the majority of patients both at training and validation (accuracy: 82.9-95.7%). The score system displayed a sensitivity/specificity of 95.2/96.9%, with an AUC of 0.971. A flow-chart integrating our score correctly managed all patients except 3 (98.1% accuracy), avoiding the potential repetition of 77.2% of AVS procedures. CONCLUSIONS: Our score could be integrated in clinical practice and guide surgical decision-making in patients with unilateral successful AVS and contralateral suppression.

Immune profiling of plasma-derived extracellular vesicles identifies Parkinson disease.

OBJECTIVE: To develop a diagnostic model based on plasma-derived extracellular vesicle (EV) subpopulations in Parkinson disease (PD) and atypical parkinsonism (AP), we applied an innovative flow cytometric multiplex bead-based platform. METHODS: Plasma-derived EVs were isolated from PD, matched healthy controls, multiple system atrophy (MSA), and AP with tauopathies (AP-Tau). The expression levels of 37 EV surface markers were measured by flow cytometry and correlated with clinical scales. A diagnostic model based on EV surface markers expression was built via supervised machine learning algorithms and validated in an external cohort. RESULTS: Distinctive pools of EV surface markers related to inflammatory and immune cells stratified patients according to the clinical diagnosis. PD and MSA displayed a greater pool of overexpressed immune markers, suggesting a different immune dysregulation in PD and MSA vs AP-Tau. The receiver operating characteristic curve analysis of a compound EV marker showed optimal diagnostic performance for PD (area under the curve [AUC] 0.908; sensitivity 96.3%, specificity 78.9%) and MSA (AUC 0.974; sensitivity 100%, specificity 94.7%) and good accuracy for AP-Tau (AUC 0.718; sensitivity 77.8%, specificity 89.5%). A diagnostic model based on EV marker expression correctly classified 88.9% of patients with reliable diagnostic performance after internal and external validations. CONCLUSIONS: Immune profiling of plasmatic EVs represents a crucial step toward the identification of biomarkers of disease for PD and AP.

Circulating extracellular vesicles as non-invasive biomarker of rejection in heart transplant.

BACKGROUND: Circulating extracellular vesicles (EVs) are raising considerable interest as a non-invasive diagnostic tool, as they are easily detectable in biologic fluids and contain a specific set of nucleic acids, proteins, and lipids reflecting pathophysiologic conditions. We aimed to investigate differences in plasma-derived EV surface protein profiles as a biomarker to be used in combination with endomyocardial biopsies (EMBs) for the diagnosis of allograft rejection. METHODS: Plasma was collected from 90 patients (53 training cohort, 37 validation cohort) before EMB. EV concentration was assessed by nanoparticle tracking analysis. EV surface antigens were measured using a multiplex flow cytometry assay composed of 37 fluorescently labeled capture bead populations coated with specific antibodies directed against respective EV surface epitopes. RESULTS: The concentration of EVs was significantly increased and their diameter decreased in patients undergoing rejection as compared with negative ones. The trend was highly significant for both antibody-mediated rejection and acute cellular rejection (p < 0.001). Among EV surface markers, CD3, CD2, ROR1, SSEA-4, human leukocyte antigen (HLA)-I, and CD41b were identified as discriminants between controls and acute cellular rejection, whereas HLA-II, CD326, CD19, CD25, CD20, ROR1, SSEA-4, HLA-I, and CD41b discriminated controls from patients with antibody-mediated rejection. Receiver operating characteristics curves confirmed a reliable diagnostic performance for each single marker (area under the curve range, 0.727-0.939). According to differential EV-marker expression, a diagnostic model was built and validated in an external cohort of patients. Our model was able to distinguish patients undergoing rejection from those without rejection. The accuracy at validation in an independent external cohort reached 86.5%. Its application for patient management has the potential to reduce the number of EMBs. Further studies in a higher number of patients are required to validate this approach for clinical purposes. CONCLUSIONS: Circulating EVs are highly promising as a new tool to characterize cardiac allograft rejection and to be complementary to EMB monitoring.

An extracellular vesicle epitope profile is associated with acute myocardial infarction.

The current standard biomarker for myocardial infarction (MI) is high-sensitive troponin. Although powerful in clinical setting, search for new markers is warranted as early diagnosis of MI is associated with improved outcomes. Extracellular vesicles (EVs) attracted considerable interest as new blood biomarkers. A training cohort used for diagnostic modelling included 30 patients with STEMI, 38 with stable angina (SA) and 30 matched-controls. Extracellular vesicle concentration was assessed by nanoparticle tracking analysis. Extracellular vesicle surface-epitopes were measured by flow cytometry. Diagnostic models were developed using machine learning algorithms and validated on an independent cohort of 80 patients. Serum EV concentration from STEMI patients was increased as compared to controls and SA. EV levels of CD62P, CD42a, CD41b, CD31 and CD40 increased in STEMI, and to a lesser extent in SA patients. An aggregate marker including EV concentration and CD62P/CD42a levels achieved non-inferiority to troponin, discriminating STEMI from controls (AUC = 0.969). A random forest model based on EV biomarkers discriminated the two groups with 100% accuracy. EV markers and RF model confirmed high diagnostic performance at validation. In conclusion, patients with acute MI or SA exhibit characteristic EV biomarker profiles. EV biomarkers hold great potential as early markers for the management of patients with MI.

Development and Validation of Prediction Models for Subtype Diagnosis of Patients With Primary Aldosteronism.

CONTEXT: Primary aldosteronism (PA) comprises unilateral (lateralized [LPA]) and bilateral disease (BPA). The identification of LPA is important to recommend potentially curative adrenalectomy. Adrenal venous sampling (AVS) is considered the gold standard for PA subtyping, but the procedure is available in few referral centers. OBJECTIVE: To develop prediction models for subtype diagnosis of PA using patient clinical and biochemical characteristics. DESIGN, PATIENTS AND SETTING: Patients referred to a tertiary hypertension unit. Diagnostic algorithms were built and tested in a training (N = 150) and in an internal validation cohort (N = 65), respectively. The models were validated in an external independent cohort (N = 118). MAIN OUTCOME MEASURE: Regression analyses and supervised machine learning algorithms were used to develop and validate 2 diagnostic models and a 20-point score to classify patients with PA according to subtype diagnosis. RESULTS: Six parameters were associated with a diagnosis of LPA (aldosterone at screening and after confirmatory testing, lowest potassium value, presence/absence of nodules, nodule diameter, and computed tomography results) and were included in the diagnostic models. Machine learning algorithms displayed high accuracy at training and internal validation (79.1%-93%), whereas a 20-point score reached an area under the curve of 0.896, and a sensitivity/specificity of 91.7/79.3%. An integrated flowchart correctly addressed 96.3% of patients to surgery and would have avoided AVS in 43.7% of patients. The external validation on an independent cohort confirmed a similar diagnostic performance. CONCLUSIONS: Diagnostic modelling techniques can be used for subtype diagnosis and guide surgical decision in patients with PA in centers where AVS is unavailable.

Hyperdimensional Computing With Local Binary Patterns: One-Shot Learning of Seizure Onset and Identification of Ictogenic Brain Regions Using Short-Time iEEG Recordings.

OBJECTIVE: We develop a fast learning algorithm combining symbolic dynamics and brain-inspired hyperdimensional computing for both seizure onset detection and identification of ictogenic (seizure generating) brain regions from intracranial electroencephalography (iEEG). METHODS: Our algorithm first transforms iEEG time series from each electrode into symbolic local binary pattern codes, from which a holographic distributed representation of the brain state of interest is constructed across all the electrodes and over time in a hyperdimensional space. The representation is used to quickly learn from few seizures, detect their onset, and identify the spatial brain regions that generated them. RESULTS: We assess our algorithm on our dataset that contains 99 short-time iEEG recordings from 16 drug-resistant epilepsy patients being implanted with 36-100 electrodes. For the majority of the patients (ten out of 16), our algorithm quickly learns from one or two seizures and perfectly (100%) generalizes on novel seizures using k-fold cross-validation. For the remaining six patients, the algorithm requires three to six seizures for learning. Our algorithm surpasses the state-of-the-art including deep learning algorithms by achieving higher specificity (94.84% versus 94.77%) and macroaveraging accuracy (95.42% versus 94.96%), and 74× lower memory footprint, but slightly higher average latency in detection (15.9 s versus 14.7 s). Moreover, the algorithm can reliably identify (with a p-value ) the relevant electrodes covering an ictogenic brain region at two levels of granularity: cerebral hemispheres and lobes. CONCLUSION AND SIGNIFICANCE: Our algorithm provides: 1) a unified method for both learning and classification tasks with end-to-end binary operations; 2) one-shot learning from seizure examples; 3) linear computational scalability for increasing number of electrodes; and 4) generation of transparent codes that enables post-translational support for clinical decision making. Our source code and anonymized iEEG dataset are freely available at http://ieeg-swez.ethz.ch.