More Like this

1. A Deep Learning-Based Real-time Seizure Detection System

   https://doi.org/10.1109/SPMB50085.2020.9353623  

   Shawki, N. ; Elseify, T. ; Cap, T. ; Shah, V. ; Obeid, I. ; Picone, J. ( December 2020 , Proceedings of the IEEE Signal Processing in Medicine and Biology Symposium (SPMB))

   Obeid, Iyad Selesnick (Ed.)

   Electroencephalography (EEG) is a popular clinical monitoring tool used for diagnosing brain-related disorders such as epilepsy [1]. As monitoring EEGs in a critical-care setting is an expensive and tedious task, there is a great interest in developing real-time EEG monitoring tools to improve patient care quality and efficiency [2]. However, clinicians require automatic seizure detection tools that provide decisions with at least 75% sensitivity and less than 1 false alarm (FA) per 24 hours [3]. Some commercial tools recently claim to reach such performance levels, including the Olympic Brainz Monitor [4] and Persyst 14 [5]. In this abstract, we describe our efforts to transform a high-performance offline seizure detection system [3] into a low latency real-time or online seizure detection system. An overview of the system is shown in Figure 1. The main difference between an online versus offline system is that an online system should always be causal and has minimum latency which is often defined by domain experts. The offline system, shown in Figure 2, uses two phases of deep learning models with postprocessing [3]. The channel-based long short term memory (LSTM) model (Phase 1 or P1) processes linear frequency cepstral coefficients (LFCC) [6] features from each EEG channel separately. We use the hypotheses generated by the P1 model and create additional features that carry information about the detected events and their confidence. The P2 model uses these additional features and the LFCC features to learn the temporal and spatial aspects of the EEG signals using a hybrid convolutional neural network (CNN) and LSTM model. Finally, Phase 3 aggregates the results from both P1 and P2 before applying a final postprocessing step. The online system implements Phase 1 by taking advantage of the Linux piping mechanism, multithreading techniques, and multi-core processors. To convert Phase 1 into an online system, we divide the system into five major modules: signal preprocessor, feature extractor, event decoder, postprocessor, and visualizer. The system reads 0.1-second frames from each EEG channel and sends them to the feature extractor and the visualizer. The feature extractor generates LFCC features in real time from the streaming EEG signal. Next, the system computes seizure and background probabilities using a channel-based LSTM model and applies a postprocessor to aggregate the detected events across channels. The system then displays the EEG signal and the decisions simultaneously using a visualization module. The online system uses C++, Python, TensorFlow, and PyQtGraph in its implementation. The online system accepts streamed EEG data sampled at 250 Hz as input. The system begins processing the EEG signal by applying a TCP montage [8]. Depending on the type of the montage, the EEG signal can have either 22 or 20 channels. To enable the online operation, we send 0.1-second (25 samples) length frames from each channel of the streamed EEG signal to the feature extractor and the visualizer. Feature extraction is performed sequentially on each channel. The signal preprocessor writes the sample frames into two streams to facilitate these modules. In the first stream, the feature extractor receives the signals using stdin. In parallel, as a second stream, the visualizer shares a user-defined file with the signal preprocessor. This user-defined file holds raw signal information as a buffer for the visualizer. The signal preprocessor writes into the file while the visualizer reads from it. Reading and writing into the same file poses a challenge. The visualizer can start reading while the signal preprocessor is writing into it. To resolve this issue, we utilize a file locking mechanism in the signal preprocessor and visualizer. Each of the processes temporarily locks the file, performs its operation, releases the lock, and tries to obtain the lock after a waiting period. The file locking mechanism ensures that only one process can access the file by prohibiting other processes from reading or writing while one process is modifying the file [9]. The feature extractor uses circular buffers to save 0.3 seconds or 75 samples from each channel for extracting 0.2-second or 50-sample long center-aligned windows. The module generates 8 absolute LFCC features where the zeroth cepstral coefficient is replaced by a temporal domain energy term. For extracting the rest of the features, three pipelines are used. The differential energy feature is calculated in a 0.9-second absolute feature window with a frame size of 0.1 seconds. The difference between the maximum and minimum temporal energy terms is calculated in this range. Then, the first derivative or the delta features are calculated using another 0.9-second window. Finally, the second derivative or delta-delta features are calculated using a 0.3-second window [6]. The differential energy for the delta-delta features is not included. In total, we extract 26 features from the raw sample windows which add 1.1 seconds of delay to the system. We used the Temple University Hospital Seizure Database (TUSZ) v1.2.1 for developing the online system [10]. The statistics for this dataset are shown in Table 1. A channel-based LSTM model was trained using the features derived from the train set using the online feature extractor module. A window-based normalization technique was applied to those features. In the offline model, we scale features by normalizing using the maximum absolute value of a channel [11] before applying a sliding window approach. Since the online system has access to a limited amount of data, we normalize based on the observed window. The model uses the feature vectors with a frame size of 1 second and a window size of 7 seconds. We evaluated the model using the offline P1 postprocessor to determine the efficacy of the delayed features and the window-based normalization technique. As shown by the results of experiments 1 and 4 in Table 2, these changes give us a comparable performance to the offline model. The online event decoder module utilizes this trained model for computing probabilities for the seizure and background classes. These posteriors are then postprocessed to remove spurious detections. The online postprocessor receives and saves 8 seconds of class posteriors in a buffer for further processing. It applies multiple heuristic filters (e.g., probability threshold) to make an overall decision by combining events across the channels. These filters evaluate the average confidence, the duration of a seizure, and the channels where the seizures were observed. The postprocessor delivers the label and confidence to the visualizer. The visualizer starts to display the signal as soon as it gets access to the signal file, as shown in Figure 1 using the “Signal File” and “Visualizer” blocks. Once the visualizer receives the label and confidence for the latest epoch from the postprocessor, it overlays the decision and color codes that epoch. The visualizer uses red for seizure with the label SEIZ and green for the background class with the label BCKG. Once the streaming finishes, the system saves three files: a signal file in which the sample frames are saved in the order they were streamed, a time segmented event (TSE) file with the overall decisions and confidences, and a hypotheses (HYP) file that saves the label and confidence for each epoch. The user can plot the signal and decisions using the signal and HYP files with only the visualizer by enabling appropriate options. For comparing the performance of different stages of development, we used the test set of TUSZ v1.2.1 database. It contains 1015 EEG records of varying duration. The any-overlap performance [12] of the overall system shown in Figure 2 is 40.29% sensitivity with 5.77 FAs per 24 hours. For comparison, the previous state-of-the-art model developed on this database performed at 30.71% sensitivity with 6.77 FAs per 24 hours [3]. The individual performances of the deep learning phases are as follows: Phase 1’s (P1) performance is 39.46% sensitivity and 11.62 FAs per 24 hours, and Phase 2 detects seizures with 41.16% sensitivity and 11.69 FAs per 24 hours. We trained an LSTM model with the delayed features and the window-based normalization technique for developing the online system. Using the offline decoder and postprocessor, the model performed at 36.23% sensitivity with 9.52 FAs per 24 hours. The trained model was then evaluated with the online modules. The current performance of the overall online system is 45.80% sensitivity with 28.14 FAs per 24 hours. Table 2 summarizes the performances of these systems. The performance of the online system deviates from the offline P1 model because the online postprocessor fails to combine the events as the seizure probability fluctuates during an event. The modules in the online system add a total of 11.1 seconds of delay for processing each second of the data, as shown in Figure 3. In practice, we also count the time for loading the model and starting the visualizer block. When we consider these facts, the system consumes 15 seconds to display the first hypothesis. The system detects seizure onsets with an average latency of 15 seconds. Implementing an automatic seizure detection model in real time is not trivial. We used a variety of techniques such as the file locking mechanism, multithreading, circular buffers, real-time event decoding, and signal-decision plotting to realize the system. A video demonstrating the system is available at: https://www.isip.piconepress.com/projects/nsf_pfi_tt/resources/videos/realtime_eeg_analysis/v2.5.1/video_2.5.1.mp4. The final conference submission will include a more detailed analysis of the online performance of each module. ACKNOWLEDGMENTS Research reported in this publication was most recently supported by the National Science Foundation Partnership for Innovation award number IIP-1827565 and the Pennsylvania Commonwealth Universal Research Enhancement Program (PA CURE). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the official views of any of these organizations. REFERENCES [1] A. Craik, Y. He, and J. L. Contreras-Vidal, “Deep learning for electroencephalogram (EEG) classification tasks: a review,” J. Neural Eng., vol. 16, no. 3, p. 031001, 2019. https://doi.org/10.1088/1741-2552/ab0ab5. [2] A. C. Bridi, T. Q. Louro, and R. C. L. Da Silva, “Clinical Alarms in intensive care: implications of alarm fatigue for the safety of patients,” Rev. Lat. Am. Enfermagem, vol. 22, no. 6, p. 1034, 2014. https://doi.org/10.1590/0104-1169.3488.2513. [3] M. Golmohammadi, V. Shah, I. Obeid, and J. Picone, “Deep Learning Approaches for Automatic Seizure Detection from Scalp Electroencephalograms,” in Signal Processing in Medicine and Biology: Emerging Trends in Research and Applications, 1st ed., I. Obeid, I. Selesnick, and J. Picone, Eds. New York, New York, USA: Springer, 2020, pp. 233–274. https://doi.org/10.1007/978-3-030-36844-9_8. [4] “CFM Olympic Brainz Monitor.” [Online]. Available: https://newborncare.natus.com/products-services/newborn-care-products/newborn-brain-injury/cfm-olympic-brainz-monitor. [Accessed: 17-Jul-2020]. [5] M. L. Scheuer, S. B. Wilson, A. Antony, G. Ghearing, A. Urban, and A. I. Bagic, “Seizure Detection: Interreader Agreement and Detection Algorithm Assessments Using a Large Dataset,” J. Clin. Neurophysiol., 2020. https://doi.org/10.1097/WNP.0000000000000709. [6] A. Harati, M. Golmohammadi, S. Lopez, I. Obeid, and J. Picone, “Improved EEG Event Classification Using Differential Energy,” in Proceedings of the IEEE Signal Processing in Medicine and Biology Symposium, 2015, pp. 1–4. https://doi.org/10.1109/SPMB.2015.7405421. [7] V. Shah, C. Campbell, I. Obeid, and J. Picone, “Improved Spatio-Temporal Modeling in Automated Seizure Detection using Channel-Dependent Posteriors,” Neurocomputing, 2021. [8] W. Tatum, A. Husain, S. Benbadis, and P. Kaplan, Handbook of EEG Interpretation. New York City, New York, USA: Demos Medical Publishing, 2007. [9] D. P. Bovet and C. Marco, Understanding the Linux Kernel, 3rd ed. O’Reilly Media, Inc., 2005. https://www.oreilly.com/library/view/understanding-the-linux/0596005652/. [10] V. Shah et al., “The Temple University Hospital Seizure Detection Corpus,” Front. Neuroinform., vol. 12, pp. 1–6, 2018. https://doi.org/10.3389/fninf.2018.00083. [11] F. Pedregosa et al., “Scikit-learn: Machine Learning in Python,” J. Mach. Learn. Res., vol. 12, pp. 2825–2830, 2011. https://dl.acm.org/doi/10.5555/1953048.2078195. [12] J. Gotman, D. Flanagan, J. Zhang, and B. Rosenblatt, “Automatic seizure detection in the newborn: Methods and initial evaluation,” Electroencephalogr. Clin. Neurophysiol., vol. 103, no. 3, pp. 356–362, 1997. https://doi.org/10.1016/S0013-4694(97)00003-9. 

   more » « less
2. Drying, more than warming, alters ecosystem functioning in streams with different energy pathways

   https://doi.org/10.1111/1365-2435.14351  

   Nelson, Daniel ; Busch, Michelle H. ; Kopp, Darin A. ; Allen, Daniel C. ( May 2023 , Functional Ecology)

   Empirical evidence and theory suggest that climate warming and an increase in the frequency and duration of drying events will alter the metabolic balance of freshwater ecosystems. However, the impacts of climate change on ecosystem metabolism may depend on whether energy inputs are of autochthonous or allochthonous origin. To date, few studies have examined how warming and drying may interact to alter stream metabolism, much less how their impacts may depend on the energy‐base of the food web.

   To address this research gap, we conducted a multi‐factorial experiment using outdoor mesocosms to investigate the individual and synergistic effects of warming and drought on metabolic processes in stream mesocosms with green (algal‐based) vs. mixed (algal‐ and detritus‐based) vs. brown (detritus‐based) energy pathways.

   We set up 48 mesocosms with one of three different levels of shade and leaf litter input combinations to create mesocosms with different primary energy channels. In addition, we warmed half of the mesocosms by ~2–3°C. We assessed changes in ecosystem respiration (ER), gross primary production (GPP), net ecosystem production (NEP) and organic matter biomass in warmed and ambient temperature mesocosms before a 24 day drying event and after rewetting.

   Surprisingly, experimental warming had little effect on metabolic processes. Drying, however, led to decreased rates of ER and GPP and led to an overall reduction in NEP. Although the effects of drying were similar across energy channel treatments, reductions in ER and GPP were primarily driven by decreases in biomass of benthic and filamentous algae.

   Overall, we demonstrate that drying led to lower rates of NEP in mesocosms regardless of energy inputs. While warming showed little effect in our study, our results suggest that an increase in the frequency of stream drying events could greatly alter the metabolic balance of many aquatic ecosystems.

   Read the freePlain Language Summaryfor this article on the Journal blog.

    

   more » « less
3. Improving Predictions of Stream CO 2 Concentrations and Fluxes Using a Stream Network Model: A Case Study in the East River Watershed, CO, USA

   https://doi.org/10.1029/2021GB006972  

   Saccardi, Brian ; Winnick, Matthew ( December 2021 , Global Biogeochemical Cycles)

   Inland waters are an important component of the global carbon budget. However, our ability to predict carbon fluxes from stream systems remains uncertain, aspCO₂varies within streams at scales of 1–100 m. This makes direct monitoring of large‐scale CO₂fluxes impractical. We incorporate CO₂input and output fluxes into a stream network advection‐reaction model, representing the first process‐based representation of stream CO₂dynamics at watershed scales. This model includes groundwater (GW) CO₂inputs, water column (WC), benthic hyporheic zone (BHZ) respiration, downstream advection, and atmospheric exchange. We evaluate this model against existing statistical methods including upscaling and multiple linear regressions through comparisons to high‐resolution streampCO₂data collected across the East River Watershed in the Colorado Rocky Mountains (USA). The stream network model accurately captures GW, evasion, and respiration‐drivenpCO₂variability and significantly outperforms multiple linear regressions for predictingpCO₂. Further, the model provides estimates of CO₂contributions from internal versus external sources suggesting that streams transition from GW‐ to BHZ‐dominated sources between 3rd and 4th Strahler orders, with GW, BHZ, and WC accounting for 49.3%, 50.6%, and 0.1% of CO₂fluxes from the watershed, respectively. Lastly, stream network model atmospheric CO₂fluxes are 4‐12x times smaller than upscaling technique predictions, largely due to relationships between streampCO₂and gas exchange velocities. Taken together, this stream network model improves our ability to predict stream CO₂dynamics and efflux. Furthermore, future applications to regional and global scales may result in a significant downward revision of global flux estimates.

    

   more » « less
4. Monitoring covariance in multivariate time series: Comparing machine learning and statistical approaches

   https://doi.org/10.1002/qre.3551  

   Weix, Derek ; Cath, Tzahi Y. ; Hering, Amanda S. ( April 2024 , Quality and Reliability Engineering International)

   In complex systems with multiple variables monitored at high‐frequency, variables are not only temporally autocorrelated, but they may also be nonlinearly related or exhibit nonstationarity as the inputs or operation changes. One approach to handling such variables is to detrend them prior to monitoring and then apply control charts that assume independence and stationarity to the residuals. Monitoring controlled systems is even more challenging because the control strategy seeks to maintain variables at prespecified mean levels, and to compensate, correlations among variables may change, making monitoring the covariance essential. In this paper, a vector autoregressive model (VAR) is compared with a multivariate random forest (MRF) and a neural network (NN) for detrending multivariate time series prior to monitoring the covariance of the residuals using a multivariate exponentially weighted moving average (MEWMA) control chart. Machine learning models have an advantage when the data's structure is unknown or may change. We design a novel simulation study with nonlinear, nonstationary, and autocorrelated data to compare the different detrending models and subsequent covariance monitoring. The machine learning models have superior performance for nonlinear and strongly autocorrelated data and similar performance for linear data. An illustration with data from a reverse osmosis process is given.

    

   more » « less
5. Dimension-reduction simplifies the analysis of signal crosstalk in a bacterial quorum sensing pathway

   https://doi.org/10.1038/s41598-021-99169-0  

   Miller, Taylor ; Patel, Keval ; Rodriguez, Coralis ; Stabb, Eric V. ; Hagen, Stephen J. ( October 2021 , Scientific Reports)

   Many pheromone sensing bacteria produce and detect more than one chemically distinct signal, or autoinducer. The pathways that detect these signals are typically noisy and interlocked through crosstalk and feedback. As a result, the sensing response of individual cells is described by statistical distributions that change under different combinations of signal inputs. Here we examine how signal crosstalk reshapes this response. We measure how combinations of two homoserine lactone (HSL) input signals alter the statistical distributions of individual cell responses in the AinS/R- and LuxI/R-controlled branches of theVibrio fischeribioluminescence pathway. We find that, while the distributions of pathway activation in individual cells vary in complex fashion with environmental conditions, these changes have a low-dimensional representation. For both the AinS/R and LuxI/R branches, the distribution of individual cell responses to mixtures of the two HSLs is effectively one-dimensional, so that a single tuning parameter can capture the full range of variability in the distributions. Combinations of crosstalking HSL signals extend the range of responses for each branch of the circuit, so that signals in combination allow population-wide distributions that are not available under a single HSL input. Dimension reduction also simplifies the problem of identifying the HSL conditions to which the pathways and their outputs are most sensitive. A comparison of the maximum sensitivity HSL conditions to actual HSL levels measured during culture growth indicates that the AinS/R and LuxI/R branches lack sensitivity to population density except during the very earliest and latest stages of growth respectively.

    

   more » « less