skip to main content

This content will become publicly available on July 7, 2023

Title: (Invited) Oxide Electronics and Recent Progress in Bipolar Applications
The discovery of oxide electronics is of increasing importance today as one of the most promising new technologies and manufacturing processes for a variety of electronic and optoelectronic applications such as next-generation displays, batteries, solar cells, memory devices, and photodetectors[1]. The high potential use seen in oxide electronics is due primarily to their high carrier mobilities and their ability to be fabricated at low temperatures[2]. However, since the majority of oxide semiconductors are n-type oxides, current applications are limited to unipolar devices, eventually developing oxide-based bipolar devices such as p-n diodes and complementary metal-oxide semiconductors. We have contributed to a wide range of oxide semiconductors and their electronics and optoelectronic device applications. Particularly, we have demonstrated n-type oxide-based thin film transistors (TFT), integrating In 2 O 3 -based n-type oxide semiconductors from binary cation materials to ternary cation species including InZnO, InGaZnO (IGZO), and InAlZnO. We have suggested channel/metallization contact strategies to achieve stable and high TFT performance[3, 4], identified vacancy-based native defect doping mechanisms[5], suggested interfacial buffer layers to promote charge injection capability[6], and established the role of third cation species on the carrier generation and carrier transport[7]. More recently, we have reported facile manufacturing of p-type SnOx through more » reactive magnetron sputtering from a Sn metal target[8]. The fabricated p-SnOx was found to be devoid of metallic phase of Sn from x-ray photoelectron spectroscopy and demonstrated stable performance in a fully oxide-based p-n heterojunction together with n-InGaZnO. The oxide-based p-n junctions exhibited a high rectification ratio greater than 10 3 at ±3 V, a low saturation current of ~2x10 -10 , and a small turn-on voltage of -0.5 V. In this presentation, we review recent achievements and still remaining issues in transition metal oxide semiconductors and their device applications, in particular, bipolar applications including p-n heterostructures and complementary metal-oxide-semiconductor devices as well as single polarity devices such as TFTs and memristors. In addition, the fundamental mechanisms of carrier transport behaviors and doping mechanisms that govern the performance of these oxide-based devices will also be discussed. ACKNOWLEDGMENT This work was supported by the U.S. National Science Foundation (NSF) Award No. ECCS-1931088. S.L. and H.W.S. acknowledge the support from the Improvement of Measurement Standards and Technology for Mechanical Metrology (Grant No. 20011028) by KRISS. K.N. was supported by Basic Science Research Program (NRF-2021R11A1A01051246) through the NRF Korea funded by the Ministry of Education. REFERENCES [1] K. Nomura et al. , Nature, vol. 432, no. 7016, pp. 488-492, Nov 25 2004. [2] D. C. Paine et al. , Thin Solid Films, vol. 516, no. 17, pp. 5894-5898, Jul 1 2008. [3] S. Lee et al. , Journal of Applied Physics, vol. 109, no. 6, p. 063702, Mar 15 2011, Art. no. 063702. [4] S. Lee et al. , Applied Physics Letters, vol. 104, no. 25, p. 252103, 2014. [5] S. Lee et al. , Applied Physics Letters, vol. 102, no. 5, p. 052101, Feb 4 2013, Art. no. 052101. [6] M. Liu et al. , ACS Applied Electronic Materials, vol. 3, no. 6, pp. 2703-2711, 2021/06/22 2021. [7] A. Reed et al. , Journal of Materials Chemistry C, 10.1039/D0TC02655G vol. 8, no. 39, pp. 13798-13810, 2020. [8] D. H. Lee et al. , ACS Applied Materials & Interfaces, vol. 13, no. 46, pp. 55676-55686, 2021/11/24 2021. « less
Authors:
; ; ; ; ; ;
Award ID(s):
1931088
Publication Date:
NSF-PAR ID:
10347718
Journal Name:
ECS Meeting Abstracts
Volume:
MA2022-01
Issue:
19
Page Range or eLocation-ID:
1071 to 1071
ISSN:
2151-2043
Sponsoring Org:
National Science Foundation
More Like this
  1. The discovery of oxide electronics is of increasing importance today as one of the most promising new technologies and manufacturing processes for a variety of electronic and optoelectronic applications such as next-generation displays, batteries, solar cells, and photodetectors. The high potential use seen in oxide electronics is due primarily to their high carrier mobilities and their ability to be fabricated at low temperatures. However, since the majority of oxide semiconductors are n-type oxides, current applications are limited to unipolar devices, eventually developing oxide-based bipolar devices such as p-n diodes and complementary metal-oxide semiconductors. We have contributed to wide range of oxide semiconductors and their electronics and optoelectronic device applications. Particularly, we have demonstrated n-type oxide-based thin film transistors (TFT), integrating In2O3-based n-type oxide semiconductors from binary cation materials to ternary cation species including InZnO, InGaZnO (IGZO), and InAlZnO. We have suggested channel/metallization contact strategies to achieve stable TFT performance, identified vacancy-based native defect doping mechanisms, suggested interfacial buffer layers to promote charge injection capability, and established the role of third cation species on the carrier generation and carrier transport. More recently, we have reported facile manufacturing of p-type SnOx through reactive magnetron sputtering from a Sn metal target. The fabricatedmore »p-SnOx was found to be devoid of metallic phase of Sn from x-ray photoelectron spectroscopy and demonstrated stable performance in a fully oxide based p-n heterojunction together with n-InGaZnO. The oxide-based p-n junctions exhibited a high rectification ratio greater than 103 at ±3 V, a low saturation current of ~2x10-10, and a small turn-on voltage of -0.5 V. With all the previous achievements and investigations about p-type oxide semiconductors, challenges remain for implementing p-type oxide realization. For the implementation of oxide-based p-n heterojunctions, the performance needs to be further enhanced. The current on/off ration may be limited, in our device structure, due to either high reverse saturation current (or current density) or non-ideal performance. In this study, two rational strategies are suggested to introduce an “intrinsic” layer, which is expected to reduce the reverse saturation current between p-SnOx and n-IGZO and hence increase the on/off ratio. The carrier density of n-IGZO is engineered in-situ during the sputtering process, by which compositionally homogeneous IGZO with significantly reduced carrier density is formed at the interface. Then, higher carrier density IGZO is formed continuously on the lower carrier density IGZO during the sputtering process without any exposure of the sample to the air. Alternatively, heterogeneous oxides of MgO and SiO2 are integrated into between p-SnOx and n-IGZO, by which the defects on the surface can be passivated. The interfacial properties are thoroughly investigated using transmission electron microscopy and atomic force microscopy. The I-V characteristics are compared between the set of devices integrated with two types of “intrinsic” layers. The current research results are expected to contribute to the development of p-type oxides and their industrial application manufacturing process that meets current processing requirements, such as mass production in p-type oxide semiconductors.« less
  2. In recent years, oxide electronics has emerged as one of the most promising new technologies for a variety of electrical and optoelectronic applications, including next-generation displays, solar cells, batteries, and photodetectors. Oxide electronics have a lot of potential because of their high carrier mobilities and ability to be manufactured at low temperatures. However, the preponderance of oxide semiconductors is n-type oxides, limiting present applications to unipolar devices and stifling the development of oxide-based bipolar devices like p-n diodes and complementary metal-oxide–semiconductors. We have contributed to oxide electronics, particularly on transition metal oxide semiconductors of which the cations include In, Zn, Sn and Ga. We have integrated these oxide semiconductors into thin film transistors (TFTs) as active channel layer in light of the unique combination of electronic and optical properties such as high carrier mobility (5-10 cm2/Vs), optical transparency in the visible regime (>~90%) and mild thermal budget processing (200-400°C). In this study, we achieved four different results. The first result is that unlike several previous reports on oxide p-n junctions fabricated exploiting a thin film epitaxial growth technique (known as molecular beam epitaxy, MBE) or a high-powered laser beam process (known as pulsed laser deposition, PLD) that requires ultra-high vacuummore »conditions, a large amount of power, and is limited for large-area processing, we demonstrate oxide-based heterojunction p-n diodes that consist of sputter-synthesized p-SnOx and n-IGZO of which the manufacturing routes are in-line with current manufacturing requirements. The second result is that the synthesized p-SnOx films are devoid of metallic Sn phases (i.e., Sn0 state) with carrier density tuneability and high carrier mobility (> 2 cm2/Vs). The third result is that the charge blocking performance of the metallurgical junction is significantly enhanced by the engineering of trap/defect density of n-IGZO, which is identified using photoelectron microscopy and valence band measurements. The last result is that the resulting oxide-based p-n heterojunction exhibits a high rectification ratio greater than 103 at ±3 V (highest among the sputter-processed oxide junctions), a low saturation current of ~2×10-10 A, and a small turn-on voltage of ~0.5 V. The outcomes of the current study are expected to contribute to the development of p-type oxides and their industrial device applications such as p-n diodes of which the manufacturing routes are in-line with the current processing requirements.« less
  3. Over the two decades, amorphous oxide semiconductors (AOSs) and their thin film transistor (TFT) channel application have been intensely explored to realize high performance, transparent and flexible displays due to their high field effect mobility (μFE=5-20 cm2/Vs), visible range optical transparency, and low temperature processability (25-300 °C).[1-2] The metastable amorphous phase is to be maintained during operation by the addition of Zn and additional third cation species (e.g., Ga, Hf, or Al) as an amorphous phase stabilizer.[3-5] To limit TFT off-state currents, a thin channel layer (10-20 nm) was employed for InZnO (IZO)-based TFTs, or third cations were added to suppress carrier generations in the TFT channel. To resolve bias stress-induced instabilities in TFT performance, approaches to employ defect passivation layers or enhance channel/dielectric interfacial compatibility were demonstrated.[6-7] Metallization contact is also a dominating factor that determines the performance of TFTs. Particularly, it has been reported that high electrical contact resistance significantly sacrifices drain bias applied to the channel, which leads to undesirable power loss during TFT operation and issues for the measurement of TFT field effect mobilities. [2, 8] However, only a few reports that suggest strategies to enhance contact behaviors are available in the literature. Furthermore, the previousmore »approaches (1) require an additional fabrication complexity due to the use of additional treatments at relatively harsh conditions such as UV, plasma, or high temperatures, and (2) may lead to adverse effects on the channel material attributed to the chemical incompatibility between dissimilar materials, and exposures to harsh environments. Therefore, a simple and easy but effective buffer strategy, which does not require any additional process complexities and not sacrifice chemical compatibility, needs to be established to mitigate the contact issues and therefore achieve high performance and low power consumption AOS TFTs. The present study aims to demonstrate an approach utilizing an interfacial buffer layer, which is compositionally homogeneous to the channel to better align work functions between channel and metallization without a significant fabrication complexity and harsh treatment conditions. Photoelectron spectroscopic measurements reveal that the conducting IZO buffer, of which the work function (Φ) is 4.37 eV, relaxes a relatively large Φ difference between channel IZO (Φ=4.81 eV) and Ti (Φ=4.2-4.3 eV) metallization. The buffer is found to lower the energy barrier for charge carriers at the source to reach the effective channel region near the dielectric. In addition, the higher carrier density of the buffer and favorable chemical compatibility with the channel (compositionally the same) further contribute to a significant reduction in specific contact resistance as much as more than 2.5 orders of magnitude. The improved contact and carrier supply performance from the source to the channel lead to an enhanced field effect mobility of up to 56.49 cm2/Vs and a threshold voltage of 1.18 V, compared to 13.41 cm2/Vs and 7.44 V of IZO TFTs without a buffer. The present work is unique in that an approach to lower the potential barrier between the source and the effective channel region (located near the channel/dielectric interface, behaving similar to a buried-channel MOSFET [9]) by introducing a contact buffer layer that enhances the field effect mobility and facilitates carrier supply from the source to the effective channel region.« less
  4. 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 EEGmore »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.« less
  5. Obeid, I. (Ed.)
    The Neural Engineering Data Consortium (NEDC) is developing the Temple University Digital Pathology Corpus (TUDP), an open source database of high-resolution images from scanned pathology samples [1], as part of its National Science Foundation-funded Major Research Instrumentation grant titled “MRI: High Performance Digital Pathology Using Big Data and Machine Learning” [2]. The long-term goal of this project is to release one million images. We have currently scanned over 100,000 images and are in the process of annotating breast tissue data for our first official corpus release, v1.0.0. This release contains 3,505 annotated images of breast tissue including 74 patients with cancerous diagnoses (out of a total of 296 patients). In this poster, we will present an analysis of this corpus and discuss the challenges we have faced in efficiently producing high quality annotations of breast tissue. It is well known that state of the art algorithms in machine learning require vast amounts of data. Fields such as speech recognition [3], image recognition [4] and text processing [5] are able to deliver impressive performance with complex deep learning models because they have developed large corpora to support training of extremely high-dimensional models (e.g., billions of parameters). Other fields that do notmore »have access to such data resources must rely on techniques in which existing models can be adapted to new datasets [6]. A preliminary version of this breast corpus release was tested in a pilot study using a baseline machine learning system, ResNet18 [7], that leverages several open-source Python tools. The pilot corpus was divided into three sets: train, development, and evaluation. Portions of these slides were manually annotated [1] using the nine labels in Table 1 [8] to identify five to ten examples of pathological features on each slide. Not every pathological feature is annotated, meaning excluded areas can include focuses particular to these labels that are not used for training. A summary of the number of patches within each label is given in Table 2. To maintain a balanced training set, 1,000 patches of each label were used to train the machine learning model. Throughout all sets, only annotated patches were involved in model development. The performance of this model in identifying all the patches in the evaluation set can be seen in the confusion matrix of classification accuracy in Table 3. The highest performing labels were background, 97% correct identification, and artifact, 76% correct identification. A correlation exists between labels with more than 6,000 development patches and accurate performance on the evaluation set. Additionally, these results indicated a need to further refine the annotation of invasive ductal carcinoma (“indc”), inflammation (“infl”), nonneoplastic features (“nneo”), normal (“norm”) and suspicious (“susp”). This pilot experiment motivated changes to the corpus that will be discussed in detail in this poster presentation. To increase the accuracy of the machine learning model, we modified how we addressed underperforming labels. One common source of error arose with how non-background labels were converted into patches. Large areas of background within other labels were isolated within a patch resulting in connective tissue misrepresenting a non-background label. In response, the annotation overlay margins were revised to exclude benign connective tissue in non-background labels. Corresponding patient reports and supporting immunohistochemical stains further guided annotation reviews. The microscopic diagnoses given by the primary pathologist in these reports detail the pathological findings within each tissue site, but not within each specific slide. The microscopic diagnoses informed revisions specifically targeting annotated regions classified as cancerous, ensuring that the labels “indc” and “dcis” were used only in situations where a micropathologist diagnosed it as such. Further differentiation of cancerous and precancerous labels, as well as the location of their focus on a slide, could be accomplished with supplemental immunohistochemically (IHC) stained slides. When distinguishing whether a focus is a nonneoplastic feature versus a cancerous growth, pathologists employ antigen targeting stains to the tissue in question to confirm the diagnosis. For example, a nonneoplastic feature of usual ductal hyperplasia will display diffuse staining for cytokeratin 5 (CK5) and no diffuse staining for estrogen receptor (ER), while a cancerous growth of ductal carcinoma in situ will have negative or focally positive staining for CK5 and diffuse staining for ER [9]. Many tissue samples contain cancerous and non-cancerous features with morphological overlaps that cause variability between annotators. The informative fields IHC slides provide could play an integral role in machine model pathology diagnostics. Following the revisions made on all the annotations, a second experiment was run using ResNet18. Compared to the pilot study, an increase of model prediction accuracy was seen for the labels indc, infl, nneo, norm, and null. This increase is correlated with an increase in annotated area and annotation accuracy. Model performance in identifying the suspicious label decreased by 25% due to the decrease of 57% in the total annotated area described by this label. A summary of the model performance is given in Table 4, which shows the new prediction accuracy and the absolute change in error rate compared to Table 3. The breast tissue subset we are developing includes 3,505 annotated breast pathology slides from 296 patients. The average size of a scanned SVS file is 363 MB. The annotations are stored in an XML format. A CSV version of the annotation file is also available which provides a flat, or simple, annotation that is easy for machine learning researchers to access and interface to their systems. Each patient is identified by an anonymized medical reference number. Within each patient’s directory, one or more sessions are identified, also anonymized to the first of the month in which the sample was taken. These sessions are broken into groupings of tissue taken on that date (in this case, breast tissue). A deidentified patient report stored as a flat text file is also available. Within these slides there are a total of 16,971 total annotated regions with an average of 4.84 annotations per slide. Among those annotations, 8,035 are non-cancerous (normal, background, null, and artifact,) 6,222 are carcinogenic signs (inflammation, nonneoplastic and suspicious,) and 2,714 are cancerous labels (ductal carcinoma in situ and invasive ductal carcinoma in situ.) The individual patients are split up into three sets: train, development, and evaluation. Of the 74 cancerous patients, 20 were allotted for both the development and evaluation sets, while the remain 34 were allotted for train. The remaining 222 patients were split up to preserve the overall distribution of labels within the corpus. This was done in hope of creating control sets for comparable studies. Overall, the development and evaluation sets each have 80 patients, while the training set has 136 patients. In a related component of this project, slides from the Fox Chase Cancer Center (FCCC) Biosample Repository (https://www.foxchase.org/research/facilities/genetic-research-facilities/biosample-repository -facility) are being digitized in addition to slides provided by Temple University Hospital. This data includes 18 different types of tissue including approximately 38.5% urinary tissue and 16.5% gynecological tissue. These slides and the metadata provided with them are already anonymized and include diagnoses in a spreadsheet with sample and patient ID. We plan to release over 13,000 unannotated slides from the FCCC Corpus simultaneously with v1.0.0 of TUDP. Details of this release will also be discussed in this poster. Few digitally annotated databases of pathology samples like TUDP exist due to the extensive data collection and processing required. The breast corpus subset should be released by November 2021. By December 2021 we should also release the unannotated FCCC data. We are currently annotating urinary tract data as well. We expect to release about 5,600 processed TUH slides in this subset. We have an additional 53,000 unprocessed TUH slides digitized. Corpora of this size will stimulate the development of a new generation of deep learning technology. In clinical settings where resources are limited, an assistive diagnoses model could support pathologists’ workload and even help prioritize suspected cancerous cases. ACKNOWLEDGMENTS This material is supported by the National Science Foundation under grants nos. CNS-1726188 and 1925494. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation. REFERENCES [1] N. Shawki et al., “The Temple University Digital Pathology Corpus,” 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 City, New York, USA: Springer, 2020, pp. 67 104. https://www.springer.com/gp/book/9783030368432. [2] J. Picone, T. Farkas, I. Obeid, and Y. Persidsky, “MRI: High Performance Digital Pathology Using Big Data and Machine Learning.” Major Research Instrumentation (MRI), Division of Computer and Network Systems, Award No. 1726188, January 1, 2018 – December 31, 2021. https://www. isip.piconepress.com/projects/nsf_dpath/. [3] A. Gulati et al., “Conformer: Convolution-augmented Transformer for Speech Recognition,” in Proceedings of the Annual Conference of the International Speech Communication Association (INTERSPEECH), 2020, pp. 5036-5040. https://doi.org/10.21437/interspeech.2020-3015. [4] C.-J. Wu et al., “Machine Learning at Facebook: Understanding Inference at the Edge,” in Proceedings of the IEEE International Symposium on High Performance Computer Architecture (HPCA), 2019, pp. 331–344. https://ieeexplore.ieee.org/document/8675201. [5] I. Caswell and B. Liang, “Recent Advances in Google Translate,” Google AI Blog: The latest from Google Research, 2020. [Online]. Available: https://ai.googleblog.com/2020/06/recent-advances-in-google-translate.html. [Accessed: 01-Aug-2021]. [6] V. Khalkhali, N. Shawki, V. Shah, M. Golmohammadi, I. Obeid, and J. Picone, “Low Latency Real-Time Seizure Detection Using Transfer Deep Learning,” in Proceedings of the IEEE Signal Processing in Medicine and Biology Symposium (SPMB), 2021, pp. 1 7. https://www.isip. piconepress.com/publications/conference_proceedings/2021/ieee_spmb/eeg_transfer_learning/. [7] J. Picone, T. Farkas, I. Obeid, and Y. Persidsky, “MRI: High Performance Digital Pathology Using Big Data and Machine Learning,” Philadelphia, Pennsylvania, USA, 2020. https://www.isip.piconepress.com/publications/reports/2020/nsf/mri_dpath/. [8] I. Hunt, S. Husain, J. Simons, I. Obeid, and J. Picone, “Recent Advances in the Temple University Digital Pathology Corpus,” in Proceedings of the IEEE Signal Processing in Medicine and Biology Symposium (SPMB), 2019, pp. 1–4. https://ieeexplore.ieee.org/document/9037859. [9] A. P. Martinez, C. Cohen, K. Z. Hanley, and X. (Bill) Li, “Estrogen Receptor and Cytokeratin 5 Are Reliable Markers to Separate Usual Ductal Hyperplasia From Atypical Ductal Hyperplasia and Low-Grade Ductal Carcinoma In Situ,” Arch. Pathol. Lab. Med., vol. 140, no. 7, pp. 686–689, Apr. 2016. https://doi.org/10.5858/arpa.2015-0238-OA.« less