More Like this

1. The Temple University Digital Pathology Corpus: The Breast Tissue Subset

   https://doi.org/10.1109/SPMB52430.2021.9672275  

   Wevodau, Z. ; Doshna, B. ; Jhala, N. ; Akhtar, I. ; Obeid, I. ; Picone, J. ( December 2021 , Proceedings of the IEEE Signal Processing in Medicine and Biology Symposium (SPMB))

   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 not 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. 

   more » « less
2. Designed improvement to T-cell immunotherapy by multidimensional single cell profiling

   https://doi.org/10.1136/jitc-2020-001877  

   Bandey, Irfan N ; Adolacion, Jay R ; Romain, Gabrielle ; Paniagua, Melisa Martinez ; An, Xingyue ; Saeedi, Arash ; Liadi, Ivan ; You, Zheng ; Rajanayake, Rasindu B ; Hwu, Patrick ; et al ( March 2021 , Journal for ImmunoTherapy of Cancer)

   null (Ed.)

   Background Adoptive cell therapy based on the infusion of chimeric antigen receptor (CAR) T cells has shown remarkable efficacy for the treatment of hematologic malignancies. The primary mechanism of action of these infused T cells is the direct killing of tumor cells expressing the cognate antigen. However, understanding why only some T cells are capable of killing, and identifying mechanisms that can improve killing has remained elusive. Methods To identify molecular and cellular mechanisms that can improve T-cell killing, we utilized integrated high-throughput single-cell functional profiling by microscopy, followed by robotic retrieval and transcriptional profiling. Results With the aid of mathematical modeling we demonstrate that non-killer CAR T cells comprise a heterogeneous population that arise from failure in each of the discrete steps leading to the killing. Differential transcriptional single-cell profiling of killers and non-killers identified CD137 as an inducible costimulatory molecule upregulated on killer T cells. Our single-cell profiling results directly demonstrate that inducible CD137 is feature of killer (and serial killer) T cells and this marks a different subset compared with the CD107a pos (degranulating) subset of CAR T cells. Ligation of the induced CD137 with CD137 ligand (CD137L) leads to younger CD19 CAR T cells with sustained killing and lower exhaustion. We genetically modified CAR T cells to co-express CD137L, in trans, and this lead to a profound improvement in anti-tumor efficacy in leukemia and refractory ovarian cancer models in mice. Conclusions Broadly, our results illustrate that while non-killer T cells are reflective of population heterogeneity, integrated single-cell profiling can enable identification of mechanisms that can enhance the function/proliferation of killer T cells leading to direct anti-tumor benefit. 

   more » « less
3. Optimizing Cytotoxic T Cell Growth in a Centrifugal Bioreactor through Kinetic Growth Models

   Kitana Kaiphanliam, Brenden Fraser-Hevlin ( November 2021 , Annual meeting American Institute of Chemical Engineers)

   Cancer is the second leading cause of death globally and remains a significant issue in medicine. Immunotherapy treatments such as Chimeric Antigen Receptor T cell (CAR-T) therapies are becoming a more promising option because of their effectiveness in killing cancer cells without harming healthy tissue in the body. CAR-T therapies, however, are inaccessible to many due to the high cost—a result of inefficient cell expansion and manufacturing methods. To address this issue, we have developed the Centrifugal Fluidized Expansion (CentriFLEX) bioreactor that balances centrifugal and fluid forces, allowing the system to operate in perfusion and maintain a high cell density. Shown in past applications for similar cell types, the CentriFLEX can expand cultures up to 2.1 billion cells in an 11.4 mL chamber over the course of one week. Recently, we have used this system to expand bovine T cells as part of a collaboration with the College of Veterinary Medicine at Washington State University. Through the project, we conducted kinetic studies to model substrate consumption and metabolite production of bovine T cells and have enhanced the bioreactor design by making it more compact to fit entirely within a biosafety cabinet— mitigating contamination concerns. Current efforts have been spent determining the remaining parameters for the kinetic models and using such models to understand how the cells grow over time and in the space of a high-population density chamber. In this presentation, we will share how we use growth models that are based on a series of kinetic studies to predict substrate and metabolite levels over time in the bioreactor, allowing us to alter feed and dosing rates of medium and nutrients to maintain cell growth at the maximum specific growth rate. 

   more » « less
4. Quantifying the efficacy of checkpoint inhibitors on CD8+ cytotoxic T cells for immunotherapeutic applications via single-cell interaction

   https://doi.org/10.1038/s41419-020-03173-7  

   Sullivan, Matthew Ryan ; Ugolini, Giovanni Stefano ; Sarkar, Saheli ; Kang, Wenjing ; Smith, Evan Carlton ; Mckenney, Seamus ; Konry, Tania ( November 2020 , Cell Death & Disease)

   The inhibition of the PD1/PDL1 pathway has led to remarkable clinical success for cancer treatment in some patients. Many, however, exhibit little to no response to this treatment. To increase the efficacy of PD1 inhibition, additional checkpoint inhibitors are being explored as combination therapy options. TSR-042 and TSR-033 are novel antibodies for the inhibition of the PD1 and LAG3 pathways, respectively, and are intended for combination therapy. Here, we explore the effect on cellular interactions of TSR-042 and TSR-033 alone and in combination at the single-cell level. Utilizing our droplet microfluidic platform, we use time-lapse microscopy to observe the effects of these antibodies on calcium flux in CD8⁺T cells upon antigen presentation, as well as their effect on the cytotoxic potential of CD8⁺T cells on human breast cancer cells. This platform allowed us to investigate the interactions between these treatments and their impacts on T-cell activity in greater detail than previously applied in vitro tests. The novel parameters we were able to observe included effects on the exact time to target cell killing, contact times, and potential for serial-killing by CD8⁺T cells. We found that inhibition of LAG3 with TSR-033 resulted in a significant increase in calcium fluctuations of CD8⁺T cells in contact with dendritic cells. We also found that the combination of TSR-042 and TSR-033 appears to synergistically increase tumor cell killing and the single-cell level. This study provides a novel single-cell-based assessment of the impact these checkpoint inhibitors have on cellular interactions with CD8⁺T cells.

    

   more » « less
5. Determining Kinetic Parameters for a Mathematical Model to Optimize Growth of Cancer-Fighting T Cells in a Novel Bioreactor

   Moore, Z ( November 2021 , Annual Biomedical Research Conference for Minority Students (ABRCMS) Virtual)

   T cell transfer immunotherapy is a highly effective cancer treatment in which the immune system’s inherent ability to fight cancer is amplified by increasing the amount of T cells that are deemed most active within a patient. T cells are a lymphocyte produced as an immune response to cancerous cells. Despite this advanced form of biological therapy, current T cell expansion methods are inefficient, resulting in high manufacturing costs, which brings question to the efficacy of T cell therapies. To address this issue, the recent development of a centrifugal bioreactor aims to rapidly expand T cells for cancer immunotherapy treatments at higher cell densities and in a shorter amount of time compared to current systems on the market. We hypothesize that by producing a mathematical model of a proof-of-concept T cell line to determine substrate consumption and metabolite production over time, we will be able to optimize growth of the cell line in the bioreactor. A series of three studies were performed to produce the growth model: (1) measuring yield coefficients of lactate, ammonium ion, and glucose, (2) determining the Monod constant and maximum specific growth rate, and (3) finding critical metabolite concentrations. To measure yield coefficients, T cells were grown in a 6-well plate at 1 x 105 cells/mL in 4 mL of medium with 100 uL samples taken and frozen each day over a 5-day period. At the end of the study, samples are thawed and used with lactate and ammonium assay kits for microplate reading to determine metabolite levels over time. To determine the Monod constant and maximum specific growth rate, T cells were grown in 12-well plates at pre-calculated varying glucose concentrations in 4 mL of medium in triplicates. Cells were counted for a minimum of six days to determine expansion over time to develop a linearized growth plot. To find critical metabolite concentrations, ammonium and lactate were added to glucose-free T cell medium at four different concentrations in triplicates utilizing a 12-well plate with a seeding density of 1 x 105 cells/mL in 4 mL of medium. The T cells then remained undisturbed in culture and were counted on day three. Once all parameters are determined, we can apply them to the growth model to determine levels of glucose, lactate, and ammonium as the T cells grow to high densities in the bioreactor and, as a result, optimize the manufacturing process for cancer immunotherapy treatments. 

   more » « less