You are invited to the New England Society for Microscopy's Virtual Spring Symposium and Workshops! The meeting will take place on May 6th and 7th via Zoom, consisting of four technical talks, four lightning talks, and two workshops.
Please register for the Symposium (for FREE!) to receive the Zoom link. Log on at 8:45 am, Symposium will begin at 9 am.
Please register separately for each Workshop (also FREE!) to receive the links.
The Matlab workshop will use WEBEX at this link:
Matlab Workshop Registration Link: Register Here
and the Python Workshop will use Zoom. Register for this workshop from this event page.
Workshop Schedule, May 6th
10:00 AM – 4:00 PM – Workshop #1:
Image Processing, Machine Learning, and Deep Learning for Images using Matlab- with Neha Sardesai and Ram Krishnamurthy (MathWorks)
Artificial Intelligence techniques like machine and deep learning are introducing automation to the products we build and the solutions we develop. These techniques are being used to solve complex problems related to images, signals, text and controls in various domains. In this hands-on virtual lab, we will use MATLAB Online to explore, analyze, and solve exercises related to Biomedical and Microscopy Image data sets using Image Processing and AI techniques.
For more details on this workshop: Scroll to below the Abstracts section
This workshop will take place over WEBEX. To register for this Workshop, use this link:
Registration Link: Register Here
Symposium Schedule, May 7th
8:45 AM – Zoom Meeting will open for Symposium
9:00 AM –Welcome by Arthur McClelland, NESM President
9:10 AM – Tjana Ivanovic (Brandeis)
Biophysics going viral: from virus cell entry to predicting and preventing the next pandemic
9:50 AM – James LeBeau (MIT)
Unraveling the atomic scale structural and chemical origins of functional oxide behavior
10:30 AM – Coffee Break: Option of breakout rooms for chatting
10:50 AM - Stephan Kraemer (Harvard)
3D Nano-Scale Imaging via Focused Ion beam Microscopy
11:30 AM - Katherine Eremin (Harvard)
Microscopy of Paul Gauguin’s Poèmes Barbares
12:10 PM - Break for lunch
1:00 PM - Lightning talks
1:00 PM - Han Xu (Harvard)
1:10 PM- Rebecca Englke (Harvard)
1:20 PM Bibi Najma (Brandeis University)
1:30 PM Will Conway (Harvard)
1:40 PM – Coffee Break: Option of breakout rooms for chatting
Workshop Schedule, May 7th
2:00 PM - 6:00 PM - Workshop #2:
Image and Data Processing with Python- with John Russell and Ian Hunt-Isaak (Harvard)
Attendees will learn the tools to automate standard microscopy analysis tasks using Python and be well prepared for building complex analysis pipelines.
The Workshop will focus on three parts: Basics of Scientific Computing, Image Analysis with examples from Bright Field Microscopy and SEM, and a survey of more advanced topics in image analysis including big data processing and machine learning.
For more details on this workshop: Scroll to below the Abstracts section
This Workshop will take place over Zoom. Please register through this NESM Event page. The Zoom link will be sent out to Workshop registrants.
ABSTRACTS AND BIOS
Tjana Ivanovic (Brandeis)
Title:
Biophysics going viral: from virus cell entry to predicting and preventing the next pandemic
Abstract:
Viruses have superb abilities to adapt to changing environments, such as new hosts in zoonotic transmissions or the acquired host immunity. The former can lead to pandemics, and the latter to viral persistence. The known viral adaptation mechanisms include error-prone replication or the concealment of functionally constrained regions, but are not sufficient in most cases to explain the high degree of viral adaptability. In this work we aimed to identify mechanisms permitting viral functional evolution under external pressures such as the immune-system antibodies. In experiments with influenza A viruses and influenza viruses pseudotyped with Ebola glycoprotein we discovered that the resistance to neutralization is a built-in feature of the filamentous particle shape in pleomorphic viral pathogens and does not require an initiating genetic change. For this work we combined molecular virology, single-RNA fluorescence in situ hybridization, real-time single-particle imaging of membrane fusion, and stochastic simulations of membrane fusion. We showed that the shape of virus particles determines the probability of both virus attachment and membrane fusion during cell entry in the presence of neutralizing antibodies. The remaining glycoproteins not bound by the antibody can cooperate to mediate cell entry, and the larger the particle the greater the extent of inactivation it can tolerate. Fewer than 5% remaining active glycoproteins on filamentous virus particles with a length of tens of micrometers can mediate efficient cell entry. Our findings suggest a pathway for pleomorphic viruses to overcome any type of selective pressure on their cell-entry glycoproteins. Specific targeting of filamentous particles might thus help overcome antiviral drug resistance and immune evasion, or prevent pandemic adaptation.
Short bio:
Tijana Ivanovic is an Assistant Professor of Biochemistry at Brandeis University. She has a Bachelor of Science in Microbiology and Molecular Genetics from UCLA (1999), and a Ph.D. in Virology from Harvard University (2008). Prior to her PhD, she worked at the Aaron Diamond AIDS Research Center in New York. In postdoctoral work with Stephen Harrison (Harvard University), funded in part by a L’Oreal Fellowship for Women in Science, Ivanovic developed single-particle methods based on TIRF microscopy for the study of influenza virus cell entry, and obtained detailed molecular models of influenza virus hemagglutinin (HA) rearrangements and collaboration during membrane fusion. In 2017, she was awarded an NIH Director’s New Innovator Award (DP2) to help define the mechanisms permitting viral functional evolution under external pressure using experimental techniques that combine virology, cell biology, and biophysics. Her previous and current work has spanned HIV, reovirus, Ebola, coronavirus, and influenza. Tijana believes that deep mechanistic understanding of virus function that can come from an interdisciplinary inquiry is at the cusp of delivering breakthroughs in our understanding of how to combat viruses that are a major threat to the human population. In between single-particle experiments, she likes to host classes of kids from a nearby children’s center in her lab to talk about viruses and lasers.
James M. LeBeau, Massachusetts Institute of Technology
Title: Unraveling the atomic scale structural and chemical origins of functional oxide behavior
Abstract:
Functional oxides exhibit a range of enticing physical properties, spanning the ferroic orders. The properties can be further tuned by, or even emerge from, the introduction of structural and chemical order/disorder at the nanoscale. Relaxor ferroelectrics, for example, are a class of functional oxides characterized by diffuse dielectric and piezoelectric properties, distinguishing them from traditional ferroelectrics. To formulate models of the local polar behavior, the structural and chemical complexity in relaxor systems have largely been explored with diffuse scattering methods that provide an average of the global and local structure. In this talk, I will discuss how aberration corrected scanning transmission electron microscopy (STEM) can be used to directly separate nanoscale structural and chemical inhomogeneities in relaxor ferroelectric materials, complementing diffraction studies. These techniques are applied to the relaxor ferroelectric system — Pb(Mg1/3Nb2/3)O3-PbTiO3 (PMN-PT). Through a simultaneous acquisition images that are sensitive to chemistry (angle annular dark-field STEM) and light elements (integrated differential phase contrast STEM), we directly connect nanoscale chemical order regions, distorted oxygen octahedra, and local polarization. First, we find that contrary to the prevailing model of a binary distribution of chemically ordered regions within a disordered matrix, the degree of chemical order smoothly varies within ordered domains and approaches a minimum at anti-phase boundaries. Second, regions of correlated oxygen octahedral titling are found to be anti-correlated with regions of maximal chemical order. Comparing with the projected polarization, we observe that the regions of greatest variation in polarization correspond to the regions of maximum chemical order and maximum octahedral distortion. Based on these results, we show that both structural and chemical inhomogeneities act as a barrier for polarization rotation and thus frustrate long range polar order. In addition, I will also present STEM studies of the multiferroic material, YFeO3. I will discuss the latest advances in energy dispersive X-ray spectroscopy that enable direct detection of anti-site defects in these materials, which are shown to break inversion symmetry. The structure, chemistry, and consequences of unique antiphase boundaries will also be discussed.
Bio:
James earned his B.S. in Materials Science & Engineering from Rensselaer Polytechnic Institute in 2006 and his Ph.D. from the University of California Santa Barbara in 2010. After his graduate work, he joined the Department of Materials Science and Engineering at North Carolina State University as a faculty member in January 2011. In 2019 he joined the faculty in the Department of Materials Science & Engineering at the Massachusetts Institute of Technology. His research focuses on applying and developing (scanning) transmission electron microscopy techniques to determine the atomic structure and chemistry of materials to inform our understanding of functional oxides, mechanical, optical, and quantum properties. He has been honored with awards that include the Presidential Early Career Award for Scientists and Engineers (PECASE), NSF CAREER award, an AFOSR Young Investigator grant, the Microscopy Society of America Burton Medal, and the Microanalysis Society K.F.J Heinrich award.
Katherine Eremin (Harvard)
Title:
Microscopy of Paul Gauguin’s Poèmes Barbares
K. Eremin, G. Rayner, A. McClelland, T. Cavanaugh, K. Smith, M. Walton, M. Vermeulen
Abstract:
Paul Gauguin painted the enigmatic composition, Poèmes Barbares, during his second trip to French Polynesia in 1896. This painting, a fusion of Tahitian, Christian and Buddhist mythologies, entered the Harvard Art Museums in 1951 as part of a bequest of French impressionist and post-impressionist works by Maurice Wertheim, a 1906 Harvard graduate, investment banker, and philanthropist. Close technical study of the paintings in the Wertheim collection was prohibited by the terms of the bequest which require that these works are always on view together. Closure of the Harvard Art Museums for a major building renovation and expansion between 2008 and 2014, provided a unique opportunity to closely examine paintings in the Wertheim collection. Radiography of Poèmes Barbares revealed an entirely different composition below the surface - a Tahitian landscape with two horses and riders rotated ninety degrees counterclockwise to the surface figures. To more fully understand the underlying image detailed analysis was undertaken. The initial phase involved non-invasive X-ray fluorescence and visible reflectance spectroscopies. This in-situ mapping was complemented by analysis of samples from the edges of the painting using optical microscopy, scanning electron microscopy with energy dispersive microanalysis, Fourier transform infrared spectroscopy, Raman spectroscopy and hyperspectral microscopy. Cross-sections revealed multiple layers with complex pigment mixtures within both paintings and identified some key differences between the palettes.
Bio:
Katherine Eremin received her PhD in 1994 from the University of Cambridge. From there she went to the National Museums of Scotland as a research scientist. She joined the Straus Center for Conservation and Technical Studies at the Harvard Art Museums in 2004 and is currently the Patricia Cornwell Senior Conservation Scientist. She specializes in the study of inorganic artefacts within the museums, in particular pigments, glass and metals from the near and far east. Recent publications include the materials and techniques used in Islamic manuscripts, ancient Chinese Jades and historic enamels and stained glass, drawing on the wide ranging collections of the Harvard Art Museums.
Lightning Talk Abstracts:
Han Xu (Harvard), Alexander Rzhevskii, Gili Naveh
Title:Detecting Riboflavin Crosslinking Induced Structural Change in Collagen with Raman Spectroscopy
Abstract:Type I collagen is a fibrillar protein that serves as a backbone for tissues that function under high loads such as bone, dentin, tendons and ligaments. Though single collagen fibril can sustain tension loads, it is the 3D organization of the fibrils into fibers and into higher hierarchical levels, that significantly enhances its mechanical strength. This 3D organization is dependent on highly-regulated intermolecular covalent crosslinks. Pathological reduction of collagen crosslinks leads to deformation of the affected tissue, resulting in disorderly functions. This phenomenon has been linked to keratoconus, where progressive thinning, scarring and protrusion of the cornea eventually causes severe visual impairment. Crosslinking of corneal collagen with riboflavin and ultraviolet A light is the standard of care for keratoconus. Riboflavin induced crosslinks increase the stiffness of the cornea by up to a factor of 1.8. However, the structural effects of crosslinking on the collagen was not thoroughly investigated. Past studies commonly use auto-fluorescence of the collagenous tissue to identify crosslink formation, which is detecting the structural changes of the collagen molecules. Raman spectroscopy is a vibrational spectroscopy that provides information of chemical bonds and is sensitive to structural detail of molecules and orientation. Our study is aimed at elucidating the effects of riboflavin crosslinking on type I collagen structure using Raman spectroscopy.
Mouse tail type I collagen fibers were freshly extracted and crosslinked with 0.5% riboflavin solution under 450nm light. Raman spectra of crosslinked and control (not crosslinked) fibers were retrieved in the range of 750 to 1800 cm-1, where most of collagen’s signature peaks can be found. Our results showed amplitude changes after crosslinking at 940 cm-1 as well as 1660cm-1 which indicate an increase of β-sheet secondary structure by 8.6% and a decrease in the α-helix content by 11.6%. Additionally, changes were seen in areas attributed to CN stretch/NH deformation (1247-1275 cm-1, AmideIII) and CH3 and CH2 deformation (1309-1400 cm-1) in collagen. This detailed structural information can help monitor the effectiveness of the crosslinking treatment and aid in optimizing dosage more precisely. Moreover, the detection of crosslinks in biological samples with Raman spectroscopy opens the possibility to many other in-depth structural studies of natural and engineered tissues.
Bio: Han Xu is a postdoctoral research fellow in Dr. Gili Naveh’s lab at Harvard School of Dental Medicine, Department of Oral Medicine, Infection and Immunity. With a background in polymeric biomaterial and biomechanics, Han is working on engineering the periodontal ligament to prevent or minimize orthodontic tooth movement relapse. Han’s project combines soft tissue engineering with optical clearing of mineralized tissue and powerful imaging methods such as high resolution micro-CT and Raman spectroscopy/microscopy, in order to study the periodontal ligament in its native, three-dimensional context.
Rebecca Englke (Harvard)
Title: Imaging dislocations in twisted bilayer 2D materials: a Moire measuring stick
Abstract: Control of the twist degree of freedom between 2D layers can tune the band structure and otherwise modify material properties, through the formation of a Moire pattern and--at small enough angles--relaxation into domains separated by dislocations. Dark field imaging in the TEM allows us to visualize the domains and dislocations in such materials. We generalize the concept of the Moire pattern beyond twist to include the layers' relative strain, and use dark field images to compute the strain tensor in the material. In this way we obtain information about the microscopic structure without requiring high resolution imaging, and open the door to studying new types of Moire patterns resulting from anisotropic distortions.
Bibi Najma (Brandeis University)
Title: Liquid to Gel Transitions in 3D Active Networks
Abstract: We are investigating the origin of self-amplifying
deformations in 3D active networks composed of
cytoskeletal filaments (microtubules), crosslinkers, and
molecular motors. The motors hydrolyze ATP and convert
chemical energy into mechanical work as they slide
adjacent microtubules. This input of energy drives the
cytoskeletal network away from thermodynamic
equilibrium. We shear-aligned this suspension in thin
microfluidic channels and investigate how the kinematics
of the instability can reveal the underlying rheology of
the network. In particular, we investigate the role of
microtubule length on the rheology of the network and
show that increasing the length of the polymer leads to a
transition from an active liquid to an active solid.
Will Conway (Harvard)
Title: Towards Measuring NDC80c Kinetics and Cooperativity from FLIM-FRET Fluctuations
Abstract: The coupling of the kinetochore to microtubules drives many aspects of chromosome motion during cell division. Despite recent advances in understanding the composition and assembly of the kinetochore, the mechanism of force generation at the kinetochore remains poorly understood. The NDC80complex (NDC80c) is believed to be the primary coupler between microtubules and the kinetochore. The two most widely discussed models of forces production are the biased diffusion model, in which the N80c is proposed to transiently and dynamically associate with microtubules, and a conformational wave model, in which the NDC80c is proposed to engage in stabile, cooperative interactions with microtubules. Our group has previously developed a FLIM-FRET based method to precisely measure the fraction of NDC80c at individual kinetochores in U2OS cells bound to microtubules. We propose an extension of this methodology to measure the dynamics and cooperativity of NDC80c binding and unbinding to microtubules at kinetochores via analysis of fluctuations in the FLIM-FRET signal. We will present theory, simulations and preliminary pilot experiments using diffusing dyes and DNA hairpins arguing that the proposed approach is promising. The same methodology could be applied to study the kinetics of other biochemical reactions in-vivo, on timescales from seconds to microsecond.
Bio: I am a fifth-year physics graduate student in the Needleman lab. I am broadly interested in the physical basis of self-assembly in living cells. Most of my PhD has focused on characterizing the biophysics of the kinetochore-microtubule interaction using a combination of fluorescence microscopy, cellular tomography, predictive modeling and novel method development aimed at measuring binding dynamics in live cells.
Matlab Workshop details:
Title: Image Processing, Machine Learning, and Deep Learning for Images using MATLAB
Artificial Intelligence techniques like machine and deep learning are introducing automation to the products we build and the solutions we develop. These techniques are being used to solve complex problems related to images, signals, text and controls in various domains.
In this hands-on virtual lab, we will use MATLAB Online to explore, analyze, and solve exercises related to Biomedical and Microscopy Image data sets using Image Processing and AI techniques. The exercises will include:
• Full workflows with image processing apps
• Image segmentation, analysis, and morphology
• Explore concepts of supervised learning and feature extraction in Machine Learning
• Build and evaluate machine learning models for classification and regression
• Train deep neural networks on GPUs in the cloud.
• Create deep learning models from scratch for image and signal data.
• Explore pretrained models and use transfer learning.
Bio: Neha Sardesai is an Education Application Engineer, and is the dedicated technical resource for MIT, Northeastern University, and Broad Institute. She partners with university customers to understand their technical and business challenges, and identifies how MathWorks products can help address these challenges in education and research. She demonstrates the value of MATLAB and Simulink to grow their adoption in curriculum, research, and commercial projects. She received her Ph.D. in Electrical Engineering with a focus on Biomedical Instrumentation from the University of Maryland, Baltimore County in 2016. She has been working at MathWorks for 3 years.
Ram Krishnamurthy is an Education Application Engineer with MathWorks. He helps faculty, researchers, and students do their best work by collaborating with them on curriculum and research matters. He received his Master's Degree in Computer Engineering from the University of Colorado, Boulder in 2016. His specialization is in Image Processing, Computer Vision and AI.
Registration Link: Register Here
Python Workshop details:
As a part of NESM’s spring 2021 meeting, we will be offering a one day workshop on python workshop. After this workshop attendees will have the tools to automate standard microscopy analysis tasks using Python and be well prepared for building complex analysis pipelines.
The workshop will be broken up into three parts each comprising demonstrations of different tools by the instructors as well as hands on exercises completed by the participants working in small groups.
The first section will be on the Basics of Scientific Computing. In this section the focus will be on using numpy for high performance array operations and using matplotlib for plotting and visualization.
The second section will focus on Image Analysis with examples from Bright Field Microscopy and SEM. We will cover the basics of Image IO, Visualization, Preprocessing and Quantification. This will be accomplished by tools from the robust ecosystem of open source libraries such as scipy,ndimage and scikit-image.
Finally we will provide a survey of more advanced topics in image analysis including big data processing and machine learning.
Prerequisites:
This Bootcamp will require some basic Python knowledge (variables, if statements, for loops). These can be learned via any of the following resources:
-A Whirlwind Tour of Python (Sections 1-9)
-Learn Python (Don't go farther than the Functions section)
-Learn to Code In Python by repl.it
If you have any questions or concerns please feel free to email John and Ian at:
johnrussell@g.harvard.edu, ianhuntisaak@g.harvard.edu
John Russell
John is a Ph.D. candidate in applied physics at Harvard University, working in the Hekstra Lab. He uses nonlinear optical microscopy to study microbial metabolism and physiology at the single cell level. His work also involves developing high performance computational tools for quantitative image analysis. These techniques are being used to study the metabolic consequences of aneuploidy, the state of having an imbalanced number of chromosomes, in budding yeast. Outside of lab he can often be found running along the Charles river as well as biking or skiing depending on the season.
Ian Hunt-Isaak
Ian is a Ph.D. candidate in Applied Physics at Harvard University in the Hekstra lab. He studies how heterogeneity develops in populations of yeast when exposed to environmental stressors. As part of that work he has developed Python code to control a custom microscope and laser integrate and automate the collection of single cell raman spectra with traditional microscopy. Working with John, they have also built a state of the art data analysis pipeline for segmenting, and tracking cells over a time series. In his spare time he enjoys gardening, fixing bikes and running. He also contributes to several open source Python projects such as Matplotlib, ipywidgets, and Jupyterlab.
