More Like this

1. What We Can Learn from Jury Note Taking: A Content Analysis

   Denne, E. ( March 2020 , Annual Conference of the American Psychology-Law Society.)

   Abstract: Jury notetaking can be controversial despite evidence suggesting benefits for recall and understanding. Research on note taking has historically focused on the deliberation process. Yet, little research explores the notes themselves. We developed a 10-item coding guide to explore what jurors take notes on (e.g., simple vs. complex evidence) and how they take notes (e.g., gist vs. specific representation). In general, jurors made gist representations of simple and complex information in their notes. This finding is consistent with Fuzzy Trace Theory (Reyna & Brainerd, 1995) and suggests notes may serve as a general memory aid, rather than verbatim representation. Summary: The practice of jury notetaking in the courtroom is often contested. Some states allow it (e.g., Nebraska: State v. Kipf, 1990), while others forbid it (e.g., Louisiana: La. Code of Crim. Proc., Art. 793). Some argue notes may serve as a memory aid, increase juror confidence during deliberation, and help jurors engage in the trial (Hannaford & Munsterman, 2001; Heuer & Penrod, 1988, 1994). Others argue notetaking may distract jurors from listening to evidence, that juror notes may be given undue weight, and that those who took notes may dictate the deliberation process (Dann, Hans, & Kaye, 2005). While research has evaluated the efficacy of juror notes on evidence comprehension, little work has explored the specific content of juror notes. In a similar project on which we build, Dann, Hans, and Kaye (2005) found jurors took on average 270 words of notes each with 85% including references to jury instructions in their notes. In the present study we use a content analysis approach to examine how jurors take notes about simple and complex evidence. We were particularly interested in how jurors captured gist and specific (verbatim) information in their notes as they have different implications for information recall during deliberation. According to Fuzzy Trace Theory (Reyna & Brainerd, 1995), people extract “gist” or qualitative meaning from information, and also exact, verbatim representations. Although both are important for helping people make well-informed judgments, gist-based understandings are purported to be even more important than verbatim understanding (Reyna, 2008; Reyna & Brainer, 2007). As such, it could be useful to examine how laypeople represent information in their notes during deliberation of evidence. Methods Prior to watching a 45-minute mock bank robbery trial, jurors were given a pen and notepad and instructed they were permitted to take notes. The evidence included testimony from the defendant, witnesses, and expert witnesses from prosecution and defense. Expert testimony described complex mitochondrial DNA (mtDNA) evidence. The present analysis consists of pilot data representing 2,733 lines of notes from 52 randomly-selected jurors across 41 mock juries. Our final sample for presentation at AP-LS will consist of all 391 juror notes in our dataset. Based on previous research exploring jury note taking as well as our specific interest in gist vs. specific encoding of information, we developed a coding guide to quantify juror note-taking behaviors. Four researchers independently coded a subset of notes. Coders achieved acceptable interrater reliability [(Cronbach’s Alpha = .80-.92) on all variables across 20% of cases]. Prior to AP-LS, we will link juror notes with how they discuss scientific and non-scientific evidence during jury deliberation. Coding Note length. Before coding for content, coders counted lines of text. Each notepad line with at minimum one complete word was coded as a line of text. Gist information vs. Specific information. Any line referencing evidence was coded as gist or specific. We coded gist information as information that did not contain any specific details but summarized the meaning of the evidence (e.g., “bad, not many people excluded”). Specific information was coded as such if it contained a verbatim descriptive (e.g.,“<1 of people could be excluded”). We further coded whether this information was related to non-scientific evidence or related to the scientific DNA evidence. Mentions of DNA Evidence vs. Other Evidence. We were specifically interested in whether jurors mentioned the DNA evidence and how they captured complex evidence. When DNA evidence was mention we coded the content of the DNA reference. Mentions of the characteristics of mtDNA vs nDNA, the DNA match process or who could be excluded, heteroplasmy, references to database size, and other references were coded. Reliability. When referencing DNA evidence, we were interested in whether jurors mentioned the evidence reliability. Any specific mention of reliability of DNA evidence was noted (e.g., “MT DNA is not as powerful, more prone to error”). Expert Qualification. Finally, we were interested in whether jurors noted an expert’s qualifications. All references were coded (e.g., “Forensic analyst”). Results On average, jurors took 53 lines of notes (range: 3-137 lines). Most (83%) mentioned jury instructions before moving on to case specific information. The majority of references to evidence were gist references (54%) focusing on non-scientific evidence and scientific expert testimony equally (50%). When jurors encoded information using specific references (46%), they referenced non-scientific evidence and expert testimony equally as well (50%). Thirty-three percent of lines were devoted to expert testimony with every juror including at least one line. References to the DNA evidence were usually focused on who could be excluded from the FBIs database (43%), followed by references to differences between mtDNA vs nDNA (30%), and mentions of the size of the database (11%). Less frequently, references to DNA evidence focused on heteroplasmy (5%). Of those references that did not fit into a coding category (11%), most focused on the DNA extraction process, general information about DNA, and the uniqueness of DNA. We further coded references to DNA reliability (15%) as well as references to specific statistical information (14%). Finally, 40% of jurors made reference to an expert’s qualifications. Conclusion Jury note content analysis can reveal important information about how jurors capture trial information (e.g., gist vs verbatim), what evidence they consider important, and what they consider relevant and irrelevant. In our case, it appeared jurors largely created gist representations of information that focused equally on non-scientific evidence and scientific expert testimony. This finding suggests note taking may serve not only to represent information verbatim, but also and perhaps mostly as a general memory aid summarizing the meaning of evidence. Further, jurors’ references to evidence tended to be equally focused on the non-scientific evidence and the scientifically complex DNA evidence. This observation suggests jurors may attend just as much to non-scientific evidence as they to do complex scientific evidence in cases involving complicated evidence – an observation that might inform future work on understanding how jurors interpret evidence in cases with complex information. Learning objective: Participants will be able to describe emerging evidence about how jurors take notes during trial. 

   more » « less
2. The most efficient sequence of study depends on the type of test

   https://doi.org/10.1002/acp.3740  

   Carvalho, Paulo F. ; Goldstone, Robert L. ( September 2020 , Applied Cognitive Psychology)

   Across three experiments featuring naturalistic concepts (psychology concepts) and naïve learners, we extend previous research showing an effect of the sequence of study on learning outcomes, by demonstrating that the sequence of examples during study changes the representation the learner creates of the study materials. We compared participants' performance in test tasks requiring different representations and evaluated which sequence yields better learning in which type of tests. We found that interleaved study, in which examples from different concepts are mixed, leads to the creation of relatively interrelated concepts that are represented by contrast to each other and based on discriminating properties. Conversely, blocked study, in which several examples of the same concept are presented together, leads to the creation of relatively isolated concepts that are represented in terms of their central and characteristic properties. These results argue for the integrated investigation of the benefits of different sequences of study as depending on the characteristics of the study and testing situation.

    

   more » « less
3. Issues in the Reproducibility of Deep Learning Results

   https://doi.org/10.1109/SPMB47826.2019.9037840  

   Jean-Paul, S. ; Elseify, T. ; Obeid, I. ; Picone, J. ( December 2019 , IEEE Signal Processing in Medicine and Biology Symposium (SPMB))

   Obeid, I. ; Selesnik, I. ; Picone, J. (Ed.)

   The Neuronix high-performance computing cluster allows us to conduct extensive machine learning experiments on big data [1]. This heterogeneous cluster uses innovative scheduling technology, Slurm [2], that manages a network of CPUs and graphics processing units (GPUs). The GPU farm consists of a variety of processors ranging from low-end consumer grade devices such as the Nvidia GTX 970 to higher-end devices such as the GeForce RTX 2080. These GPUs are essential to our research since they allow extremely compute-intensive deep learning tasks to be executed on massive data resources such as the TUH EEG Corpus [2]. We use TensorFlow [3] as the core machine learning library for our deep learning systems, and routinely employ multiple GPUs to accelerate the training process. Reproducible results are essential to machine learning research. Reproducibility in this context means the ability to replicate an existing experiment – performance metrics such as error rates should be identical and floating-point calculations should match closely. Three examples of ways we typically expect an experiment to be replicable are: (1) The same job run on the same processor should produce the same results each time it is run. (2) A job run on a CPU and GPU should produce identical results. (3) A job should produce comparable results if the data is presented in a different order. System optimization requires an ability to directly compare error rates for algorithms evaluated under comparable operating conditions. However, it is a difficult task to exactly reproduce the results for large, complex deep learning systems that often require more than a trillion calculations per experiment [5]. This is a fairly well-known issue and one we will explore in this poster. Researchers must be able to replicate results on a specific data set to establish the integrity of an implementation. They can then use that implementation as a baseline for comparison purposes. A lack of reproducibility makes it very difficult to debug algorithms and validate changes to the system. Equally important, since many results in deep learning research are dependent on the order in which the system is exposed to the data, the specific processors used, and even the order in which those processors are accessed, it becomes a challenging problem to compare two algorithms since each system must be individually optimized for a specific data set or processor. This is extremely time-consuming for algorithm research in which a single run often taxes a computing environment to its limits. Well-known techniques such as cross-validation [5,6] can be used to mitigate these effects, but this is also computationally expensive. These issues are further compounded by the fact that most deep learning algorithms are susceptible to the way computational noise propagates through the system. GPUs are particularly notorious for this because, in a clustered environment, it becomes more difficult to control which processors are used at various points in time. Another equally frustrating issue is that upgrades to the deep learning package, such as the transition from TensorFlow v1.9 to v1.13, can also result in large fluctuations in error rates when re-running the same experiment. Since TensorFlow is constantly updating functions to support GPU use, maintaining an historical archive of experimental results that can be used to calibrate algorithm research is quite a challenge. This makes it very difficult to optimize the system or select the best configurations. The overall impact of all of these issues described above is significant as error rates can fluctuate by as much as 25% due to these types of computational issues. Cross-validation is one technique used to mitigate this, but that is expensive since you need to do multiple runs over the data, which further taxes a computing infrastructure already running at max capacity. GPUs are preferred when training a large network since these systems train at least two orders of magnitude faster than CPUs [7]. Large-scale experiments are simply not feasible without using GPUs. However, there is a tradeoff to gain this performance. Since all our GPUs use the NVIDIA CUDA® Deep Neural Network library (cuDNN) [8], a GPU-accelerated library of primitives for deep neural networks, it adds an element of randomness into the experiment. When a GPU is used to train a network in TensorFlow, it automatically searches for a cuDNN implementation. NVIDIA’s cuDNN implementation provides algorithms that increase the performance and help the model train quicker, but they are non-deterministic algorithms [9,10]. Since our networks have many complex layers, there is no easy way to avoid this randomness. Instead of comparing each epoch, we compare the average performance of the experiment because it gives us a hint of how our model is performing per experiment, and if the changes we make are efficient. In this poster, we will discuss a variety of issues related to reproducibility and introduce ways we mitigate these effects. For example, TensorFlow uses a random number generator (RNG) which is not seeded by default. TensorFlow determines the initialization point and how certain functions execute using the RNG. The solution for this is seeding all the necessary components before training the model. This forces TensorFlow to use the same initialization point and sets how certain layers work (e.g., dropout layers). However, seeding all the RNGs will not guarantee a controlled experiment. Other variables can affect the outcome of the experiment such as training using GPUs, allowing multi-threading on CPUs, using certain layers, etc. To mitigate our problems with reproducibility, we first make sure that the data is processed in the same order during training. Therefore, we save the data from the last experiment and to make sure the newer experiment follows the same order. If we allow the data to be shuffled, it can affect the performance due to how the model was exposed to the data. We also specify the float data type to be 32-bit since Python defaults to 64-bit. We try to avoid using 64-bit precision because the numbers produced by a GPU can vary significantly depending on the GPU architecture [11-13]. Controlling precision somewhat reduces differences due to computational noise even though technically it increases the amount of computational noise. We are currently developing more advanced techniques for preserving the efficiency of our training process while also maintaining the ability to reproduce models. In our poster presentation we will demonstrate these issues using some novel visualization tools, present several examples of the extent to which these issues influence research results on electroencephalography (EEG) and digital pathology experiments and introduce new ways to manage such computational issues. 

   more » « less
4. Heterformer: Transformer-based Deep Node Representation Learning on Heterogeneous Text-Rich Networks

   https://doi.org/10.1145/3580305.3599376  

   Jin, Bowen ; Zhang, Yu ; Zhu, Qi ; Han, Jiawei ( August 2023 , ACM)

   Proc. 2023 ACM SIGKDD Int. Conf. on Knowledge Discovery and Data Mining (Ed.)

   Representation learning on networks aims to derive a meaningful vector representation for each node, thereby facilitating downstream tasks such as link prediction, node classification, and node clustering. In heterogeneous text-rich networks, this task is more challenging due to (1) presence or absence of text: Some nodes are associated with rich textual information, while others are not; (2) diversity of types: Nodes and edges of multiple types form a heterogeneous network structure. As pretrained language models (PLMs) have demonstrated their effectiveness in obtaining widely generalizable text representations, a substantial amount of effort has been made to incorporate PLMs into representation learning on text-rich networks. However, few of them can jointly consider heterogeneous structure (network) information as well as rich textual semantic information of each node effectively. In this paper, we propose Heterformer, a Heterogeneous Network-Empowered Transformer that performs contextualized text encoding and heterogeneous structure encoding in a unified model. Specifically, we inject heterogeneous structure information into each Transformer layer when encoding node texts. Meanwhile, Heterformer is capable of characterizing node/edge type heterogeneity and encoding nodes with or without texts. We conduct comprehensive experiments on three tasks (i.e., link prediction, node classification, and node clustering) on three large-scale datasets from different domains, where Heterformer outperforms competitive baselines significantly and consistently. 

   more » « less
5. 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