More Like this

1. Recent progress on topological semimetal IrO 2 : electronic structures, synthesis, and transport properties

   https://doi.org/10.1088/1361-648X/ad3603  

   Zhang, T. X. ; Coughlin, A. L. ; Lu, Chi-Ken ; Heremans, J. J. ; Zhang, S. X. ( April 2024 , Journal of Physics: Condensed Matter)

   5dtransition metal oxides, such as iridates, have attracted significant interest in condensed matter physics throughout the past decade owing to their fascinating physical properties that arise from intrinsically strong spin-orbit coupling (SOC) and its interplay with other interactions of comparable energy scales. Among the rich family of iridates, iridium dioxide (IrO₂), a simple binary compound long known as a promising catalyst for water splitting, has recently been demonstrated to possess novel topological states and exotic transport properties. The strong SOC and the nonsymmorphic symmetry that IrO₂possesses introduce symmetry-protected Dirac nodal lines (DNLs) within its band structure as well as a large spin Hall effect in the transport. Here, we review recent advances pertaining to the study of this unique SOC oxide, with an emphasis on the understanding of the topological electronic structures, syntheses of high crystalline quality nanostructures, and experimental measurements of its fundamental transport properties. In particular, the theoretical origin of the presence of the fourfold degenerate DNLs in band structure and its implications in the angle-resolved photoemission spectroscopy measurement and in the spin Hall effect are discussed. We further introduce a variety of synthesis techniques to achieve IrO₂nanostructures, such as epitaxial thin films and single crystalline nanowires, with the goal of understanding the roles that each key parameter plays in the growth process. Finally, we review the electrical, spin, and thermal transport studies. The transport properties under variable temperatures and magnetic fields reveal themselves to be uniquely sensitive and modifiable by strain, dimensionality (bulk, thin film, nanowire), quantum confinement, film texture, and disorder. The sensitivity, stemming from the competing energy scales of SOC, disorder, and other interactions, enables the creation of a variety of intriguing quantum states of matter.

    

   more » « less
2. The Marine Gastropod Crepidula fornicata Remains Resilient to Ocean Acidification Across Two Life History Stages

   https://doi.org/10.3389/fphys.2021.702864  

   Reyes-Giler, Christopher L. ; Benson, Brooke E. ; Levy, Morgan ; Chen, Xuqing ; Pires, Anthony ; Pechenik, Jan A. ; Davies, Sarah W. ( August 2021 , Frontiers in Physiology)

   Rising atmospheric CO 2 reduces seawater pH causing ocean acidification (OA). Understanding how resilient marine organisms respond to OA may help predict how community dynamics will shift as CO 2 continues rising. The common slipper shell snail Crepidula fornicata is a marine gastropod native to eastern North America that has been a successful invader along the western European coastline and elsewhere. It has also been previously shown to be resilient to global change stressors. To examine the mechanisms underlying C. fornicata’s resilience to OA, we conducted two controlled laboratory experiments. First, we examined several phenotypes and genome-wide gene expression of C. fornicata in response to pH treatments (7.5, 7.6, and 8.0) throughout the larval stage and then tested how conditions experienced as larvae influenced juvenile stages (i.e., carry-over effects). Second, we examined genome-wide gene expression patterns of C. fornicata larvae in response to acute (4, 10, 24, and 48 h) pH treatment (7.5 and 8.0). Both C. fornicata larvae and juveniles exhibited resilience to OA and their gene expression responses highlight the role of transcriptome plasticity in this resilience. Larvae did not exhibit reduced growth under OA until they were at least 8 days old. These phenotypic effects were preceded by broad transcriptomic changes, which likely served as an acclimation mechanism for combating reduced pH conditions frequently experienced in littoral zones. Larvae reared in reduced pH conditions also took longer to become competent to metamorphose. In addition, while juvenile sizes at metamorphosis reflected larval rearing pH conditions, no carry-over effects on juvenile growth rates were observed. Transcriptomic analyses suggest increased metabolism under OA, which may indicate compensation in reduced pH environments. Transcriptomic analyses through time suggest that these energetic burdens experienced under OA eventually dissipate, allowing C. fornicata to reduce metabolic demands and acclimate to reduced pH. Carry-over effects from larval OA conditions were observed in juveniles; however, these effects were larger for more severe OA conditions and larvae reared in those conditions also demonstrated less transcriptome elasticity. This study highlights the importance of assessing the effects of OA across life history stages and demonstrates how transcriptomic plasticity may allow highly resilient organisms, like C. fornicata , to acclimate to reduced pH environments. 

   more » « less
3. Within‐ and transgenerational stress legacy effects of ocean acidification on red abalone ( Haliotis rufescens ) growth and survival

   https://doi.org/10.1111/gcb.17048  

   Neylan, Isabelle P. ; Swezey, Daniel S. ; Boles, Sara E. ; Gross, Jackson A. ; Sih, Andrew ; Stachowicz, John J. ( November 2023 , Global Change Biology)

   Understanding the mechanisms by which individual organisms respond and populations adapt to global climate change is a critical challenge. The role of plasticity and acclimation, within and across generations, may be essential given the pace of change. We investigated plasticity across generations and life stages in response to ocean acidification (OA), which poses a growing threat to both wild populations and the sustainable aquaculture of shellfish. Most studies of OA on shellfish focus on acute effects, and less is known regarding the longer term carryover effects that may manifest within or across generations. We assessed these longer term effects in red abalone (Haliotis rufescens) using a multi‐generational split‐brood experiment. We spawned adults raised in ambient conditions to create offspring that we then exposed to high pCO₂(1180 μatm; simulating OA) or low pCO₂(450 μatm; control or ambient conditions) during the first 3 months of life. We then allowed these animals to reach maturity in ambient common garden conditions for 4 years before returning the adults into high or low pCO₂treatments for 11 months and measuring growth and reproductive potential. Early‐life exposure to OA in the F1 generation decreased adult growth rate even after 5 years especially when abalone were re‐exposed to OA as adults. Adult but not early‐life exposure to OA negatively impacted fecundity. We then exposed the F2 offspring to high or low pCO₂treatments for the first 3 months of life in a fully factorial, split‐brood design. We found negative transgenerational effects of parental OA exposure on survival and growth of F2 offspring, in addition to significant direct effects of OA on F2 survival. These results show that the negative impacts of OA can last within and across generations, but that buffering against OA conditions at critical life‐history windows can mitigate these effects.

    

   more » « less
4. 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
5. Liposome‐Templated Green Synthesis of Mesoporous Metal Nanostructures with Universal Composition for Biomedical Application

   https://doi.org/10.1002/smll.202304880  

   Wang, Jinping ; Sun, Jingyu ; Khade, Rahul L. ; Chou, Tsengming ; An, Hailong ; Zhang, Yong ; Wang, Hongjun ( July 2023 , Small)

   Porous noble metal nanoparticles have received particular attention recently for their unique optical, thermal, and catalytic functions in biomedicine. However, limited progress has been made to synthesize such porous metallic nanostructures with large mesopores (≥25 nm). Here, a green yet facile synthesis strategy using biocompatible liposomes as templates to mediate the formation of mesoporous metallic nanostructures in a controllable fashion is reported. Various monodispersed nanostructures with well‐defined mesoporous shape and large mesopores (≈ 40 nm) are successfully synthesized from mono‐ (Au, Pd, and Pt), bi‐ (AuPd, AuPt, AuRh, PtRh, and PdPt), and tri‐noble metals (AuPdRh, AuPtRh, and AuPdPt). Along with a successful demonstration of its effectiveness in synthesis of various mesoporous nanostructures, the possible mechanism of liposome‐guided formation of such nanostructures via time sectioning of the synthesis process (monitoring time‐resolved growth of mesoporous structures) and computational quantum molecular modeling (analyzing chemical interaction energy between metallic cations and liposomes at the enthalpy level) is also revealed. These mesoporous metallic nanostructures exhibit a strong photothermal effect in the near‐infrared region, effective catalytic activities in hydrogen peroxide decomposition reaction, and high drug loading capacity. Thus, the liposome‐templated method provides an inspiring and robust avenue to synthesize mesoporous noble metal‐based nanostructures for versatile biomedical applications.

    

   more » « less