News

Invited Seminar by Saga Helgadottir at the Max Planck Institute for the Science of Light, 10 May 2019

Digital video microscopy enhanced by deep learning

Saga Helgadottir
Sandoghdar Division, Max Planck Institute for the Science of Light, Erlangen, Germany
10 May 2019

Single particle tracking is essential in many branches of science and technology, from the measurement of biomolecular forces to the study of colloidal crystals. Standard methods rely on algorithmic approaches; by fine-tuning several user-defined parameters, these methods can be highly successful at tracking a well-defined kind of particle under low-noise conditions with constant and homogenous illumination. Here, we introduce an alternative data-driven approach based on a convolutional neural network, which we name DeepTrack. We show that DeepTrack outperforms algorithmic approaches, especially in the presence of noise and under poor illumination conditions. We use DeepTrack to track an optically trapped particle under very noisy and unsteady illumination conditions, where standard algorithmic approaches fail. We then demonstrate how DeepTrack can also be used to track multiple particles and non-spherical objects such as bacteria, also at very low signal-to-noise ratios. In order to make DeepTrack readily available for other users, we provide a Python software package, which can be easily personalized and optimized for specific applications.

Saga Helgadottir, Aykut Argun & Giovanni Volpe, Optica 6(4), 506—513 (2019)
doi: 10.1364/OPTICA.6.000506
arXiv: 1812.02653
GitHub: DeepTrack

Anomalous diffusion measurement with neural networks on ArXiv

Measurement of Anomalous Diffusion Using Recurrent Neural Networks

Measurement of Anomalous Diffusion Using Recurrent Neural Networks
Stefano Bo, Falko Schmidt, Ralf Eichborn & Giovanni Volpe
arXiv: 1905.02038

Anomalous diffusion occurs in many physical and biological phenomena, when the growth of the mean squared displacement (MSD) with time has an exponent different from one. We show that recurrent neural networks (RNN) can efficiently characterize anomalous diffusion by determining the exponent from a single short trajectory, outperforming the standard estimation based on the MSD when the available data points are limited, as is often the case in experiments. Furthermore, the RNN can handle more complex tasks where there are no standard approaches, such as determining the anomalous diffusion exponent from a trajectory sampled at irregular times, and estimating the switching time and anomalous diffusion exponents of an intermittent system that switches between different kinds of anomalous diffusion. We validate our method on experimental data obtained from sub-diffusive colloids trapped in speckle light fields and super-diffusive microswimmers.

Meltem Elitas from Sabanci University visits the Soft Matter Lab. Welcome!

Meltem Elitas is visiting from Sabanci University in Istanbul from 1st May until 28th June 2019.

Meltem Elitas is a faculty member at the Mechatronics Program at Sabanci University in Istanbul, Turkey. Her background is Electrical and Mechatronics Engineering; she obtained her doctorate from Bioengineerieng and Biotechnology Department at École Polytechnique Fédérale de Lausanne. She performed her postdoctoral studies at Yale University Biomedical Engineering Department. She has published more than 25 papers and conference proceedings in reputed journals. Her research interests are biomechatronics, cellular heterogeneity, cellular interactions, tumor microenvironment, live cell imaging and development of microfabricated tools for quantitative biology. She is visiting the Soft Matter Lab as part of her ongoing Marie Skłodowska-Curie project.

Invited talk by G. Volpe at DINAMO 2019, San Crístobal, Ecuador, 22-26 Apr 2019

Deep Learning Applications in Photonics and Active Matter
Giovanni Volpe
Discussions on Nano & Mesoscopic Optics (DINAMO-2019), San Crístobal, Galápagos Islands, Ecuador, 22-26 April 2019.

 

After a brief overview of artificial intelligence, machine learning and deep learning, I will present a series of recent works in which we have employed deep learning for applications in photonics and active matter. In particular, I will explain how we employed deep learning to enhance digital video microscopy [1], to estimate the properties of anomalous diffusion, and to improve the calculation of optical forces. Finally, I will provide an outlook for the application of deep learning in photonics and active matter.

References

[1] S. Helgadottir, A. Argun and G. Volpe, Digital video microscopy enhanced by deep learning. arXiv 1812.02653 (2018).

 

Presentation by Laura Pérez at the OSA Biophotonics Congress, Tucson, 16 Apr 2019

FORMA: Force Reconstruction via Maximum-likelihood-estimator Analysis

Laura Pérez García, Jaime Donlucas Pérez, Giorgio Volpe, Alejandro V. Areola & Giovanni Volpe
OSA Biophotonics Congress, Tucson (AZ), USA
16 April 2019

Microscopic force characterization is often done by using a microscopic colloidal particle which probes local forces. These particles are often held by a harmonic trapping potential with stiffness k so that a homogeneous force acting on the particle results in a displacement Δx from the equilibrium position and the force can, therefore, be measured as k Δx . To perform such measurement, it is necessary to determine the value of k , which is often done by measuring the Brownian fluctuations of the particle around its stable equilibrium position. This is achieved by measuring the particle position as a function of time, x (t) , and then using some calibration algorithms; the most commonly employed techniques are the potential analysis that relies on the fact that the force is derived from a potential; and the power spectral density (PSD) and the auto-correlation function (ACF) methods that require a regular sampling in time. Besides the previous requirements, all methods depend on the choice of some analysis parameters. This has inhibited the applicability of force measurement methods to characterize force fields with non-conservative components or where the particle freely explores an extended potential landscape. We propose a method for Force Reconstruction via Maximum-likelihood-estimator Analysis (FORMA) that exploits the fact that in the proximity of an equilibrium position the force field can be approximated by a linear form and, therefore, optimally estimated using a linear Maximum-likelihood-estimator (MLE).

Session: Biological Applications
10:30 AM–12:00 AM, Tuesday, April 16, 2019

More information can be found on the link: https://www.osapublishing.org/abstract.cfm?uri=OMA-2019-AT2E.2

 

Presentation by Saga Helgadottir at the OSA Biophotonics Congress, Tucson, 16 Apr 2019

Digital video microscopy enhanced by deep learning

Saga Helgadottir, Aykut Argun & Giovanni Volpe
OSA Biophotonics Congress, Tucson (AZ), USA
16 April 2019

Single particle tracking is essential in many branches of science and technology, from the measurement of biomolecular forces to the study of colloidal crystals. Standard methods rely on algorithmic approaches; by fine-tuning several user-defined parameters, these methods can be highly successful at tracking a well-defined kind of particle under low-noise conditions with constant and homogenous illumination. Here, we introduce an alternative data-driven approach based on a convolutional neural network, which we name DeepTrack. We show that DeepTrack outperforms algorithmic approaches, especially in the presence of noise and under poor illumination conditions. We use DeepTrack to track an optically trapped particle under very noisy and unsteady illumination conditions, where standard algorithmic approaches fail. We then demonstrate how DeepTrack can also be used to track multiple particles and non-spherical objects such as bacteria, also at very low signal-to-noise ratios. In order to make DeepTrack readily available for other users, we provide a Python software package, which can be easily personalized and optimized for specific applications.

Session: Biological Applications
10:30 AM–12:00 AM, Tuesday, April 16, 2019

More information can be found on the link: https://www.osapublishing.org/abstract.cfm?uri=OMA-2019-AT2E.5

 

Presentation by F. Schmidt at the OSA Biophotonics Congress, Tucson, 16 Apr 2019

Light-driven Assembly and Optical Manipulation of Active Colloidal Molecules

Falko Schmidt, Benno Liebchen, Hartmut Loewen & Giovanni Volpe
OSA Biophotonics Congress, Tucson (AZ), USA
16 April 2019

Active matter, consisting of self-propelled units locally injecting energy into the system, opens new horizons for the creation of functional soft materials with designable properties. Experiencing a constant energy input, allows active matter to self-assemble into phases with a complex architecture and functionality such as living clusters which dynamically form, reshape and break-up but would be forbidden in equilibrium material by the entropy maximization (or free energy minimization) principle. The challenge to control this active self-assembly has evoked widespread efforts typically hinging on an engineering of the properties of individual motile constituents. Here, we provide a different route, where activity occurs as an emergent phenomenon only when individual building blocks bind together, in a way which we control by laser light. Using experiments and simulations of two species of immotile microspheres, we exemplify this route by creating active molecules featuring a complex array of behaviors, becoming migrators, spinners and rotators. The possibility to control the dynamics of active self-assembly via light-controllable nonreciprocal interactions will inspire new approaches to understand living matter and to design active materials.

Session: Nanothermodynamics
8:00 AM–10:00 AM, Tuesday, April 16, 2019
Chair: Agnese Callegari; Bilkent University, Turkey

Presentation by Alessandro Magazzù at the OSA Biophotonics Congress, Tucson, 16 Apr 2019

Dynamics of optically trapped particles tuned by critical Casimir forces and torques

Alessandro Magazzù, Agnese Callegari, Juan Pablo Staforelli, Andrea Gambassi, Siegfried Dietrich & Giovanni Volpe.
OSA Biophotonics Congress, Tucson (AZ), USA 16 April 2019

Fluctuations have always played a crucial role in physics, especially when spatially confined by objects. Density fluctuations of the composition of a binary critical mixture emerge when its temperature is in proximity of the critical point. If these fluctuations are confined between two objects (e.g., two colloids, or a colloid and a planar surface), they give rise to Critical Casimir forces (CCFs). Although, these forces were predicted theoretically in 1978 in analogy to quantum-electrodynamical (QED) Casimir Forces they have never aroused a lot of attentions. They have always been considered mostly like a curiosity, until recently. Thanks to the development of nano-technology, CCFs seem to have establish their role in nano-science. They have been measured only recently, proving their relevance at nanoscale.

Session: Enhancing Techniques
14:00 –16:00, Tuesday, April 16, 2019
Chair: Frank Cichos; University Leipzig, Germany

 

Presentation by A. Argun at OSA Life Sciences Conference, Tucson, 14-17 April 2019

Statistics of Brownian particles held in non-harmonic potentials in an active bath

Aykut Argun and Giovanni Volpe
OSA Life Sciences Conference,
Tucson, 14-17 April 2019

Abstract: 

Active systems are subject to persistent noise that arise from biological media or artificial activity like self-propelled particles. Therefore, these systems are  intrinsically out of equilibrium and can only be studied within the framework of non-equilibrium physics. So far, steady-state behavior and dynamical fluctuations of Brownian particles in active baths have been investigated both theoretically and experimentally. While some of the equilibrium properties can be retained by using an effective temperature, for most systems this generalization is not possible. Here, we extend the existing studies to non-harmonic potential cases, where other qualitative distinctions of the active matter emerge.

Digital Video Microscopy Enhanced by Deep Learning published in Optica

Digital video microscopy enhanced by deep learning

Digital video microscopy enhanced by deep learning
(Cover article)
Saga Helgadottir, Aykut Argun & Giovanni Volpe
Optica 6(4), 506—513 (2019)
doi: 10.1364/OPTICA.6.000506
arXiv: 1812.02653
GitHub: DeepTrack

Single particle tracking is essential in many branches of science and technology, from the measurement of biomolecular forces to the study of colloidal crystals. Standard methods rely on algorithmic approaches; by fine-tuning several user-defined parameters, these methods can be highly successful at tracking a well-defined kind of particle under low-noise conditions with constant and homogenous illumination. Here, we introduce an alternative data-driven approach based on a convolutional neural network, which we name DeepTrack. We show that DeepTrack outperforms algorithmic approaches, especially in the presence of noise and under poor illumination conditions. We use DeepTrack to track an optically trapped particle under very noisy and unsteady illumination conditions, where standard algorithmic approaches fail. We then demonstrate how DeepTrack can also be used to track multiple particles and non-spherical objects such as bacteria, also at very low signal-to-noise ratios. In order to make DeepTrack readily available for other users, we provide a Python software package, which can be easily personalized and optimized for specific applications.