Conference Agenda

TOM4 S01: Bio-Medical Optics
Tuesday, 13/Sept/2022:
4:30pm - 6:00pm

Location: B325

3rd floor, 32 seats

4:30pm - 4:45pm
ID: 142 / TOM4 S01: 1
TOM 4 Bio-Medical Optics

Single-shot 3D endoscopic imaging exploiting a diffuser and neural networks

Julian Lich1, Tom Glosemeyer1, Jürgen Czarske1,2, Robert Kuschmierz1,2

1Laboratory of Measurement and Sensor Systems, TU Dresden, Germany; 2Competence Center for Biomedical Computational Laser Systems (BIOLAS), TU Dresden, Germany

Lens-based endoscopes offer high lateral resolution, but suffer from rigid imaging properties, such as a fixed focal plane. We present a miniaturized 0.5 mm diameter endoscope in which the objective lens is replaced by an optical diffuser. The intensity information of the object space is scattered and passed to a camera via a coherent fibre bundle. The image is reconstructed by a neural network. The field of view and resolution depend on the object distance. 3D-single-shot imaging up to video rate can be enabled. The approach shows great potential for applications like robust 3D fluorescence imaging.

4:45pm - 5:00pm
ID: 216 / TOM4 S01: 2
TOM 4 Bio-Medical Optics

Linearly modulated multi-focal diffractive lens for multi-sheet excitation of flow-driven samples in a light-sheet fluorescence microscope

Meike Hofmann1, Shima Gharbi Ghebjagh1, Chao Fan1, Yuchao Feng1, Karen Lemke2, Stefan Sinzinger1

1Technische Universität Ilmenau, Germany; 2Institut für Bioprozess- und Analysenmesstechnik Heiligenstadt, Germany

Light-sheet fluorescence microscopy (LSFM) with single light-sheet illumination enables rapid 3D-imaging of living cells. In this paper we show the design, fabrication and characterization of a diffractive optical element producing several light sheets along an inclined tube for applications in flow-driven imaging. The element, which is based on a multi-focal Fresnel zone plate and a linear grating, generates in combination with a refractive cylindrical lens five thin light sheets of equal intensity.

5:00pm - 5:15pm
ID: 271 / TOM4 S01: 3
TOM 4 Bio-Medical Optics

C-reactive protein detection using a lensless fibre optic fluorescence sensor

Victoria Esteso1,2, Pietro Lombardi1,2, Francesco Chiavaioli3, Maja Colautti1,2, Steffen Howitz4, Paolo Cecchi5, Mario Agio6, Ambra Giannetti3, Costanza Toninelli1,2

1CNR-INO, Italy; 2LENS, Italy; 3Istituto di Fisica Applicata "Nello Carrara", Italy; CNR-IFAC (Italy); 4GeSiM Gesellschaft fuer Silizium-Mikrosysteme mbH (Germany); 5Cecchi s.r.l. (Italy); 6Univ. Siegen (Germany)

Herein we show a prototype based on a lens-less fibre optic for fluorescence detection of labelled bio-assay which in this work it has been tested for early diagnosis of sepsis through detection of C-reactive protein (CRP) biomarker. In particular the rationalized optical design of the assay substrate allows to improve the coupling among the emitter and the optical fibre while enhancing the collection efficiency of the fluorescence signal for small numerical apertures. The prototype has been tested in a well-stablish and reproducible standard as it is the immunoglobuline IgG/Anti-IgG assay, reporting an enhancement above one order of magnitude compared to a commercial equipment. The limit-of-detection (LOD) achieved with this prototype for the CRP biomarker in matrices mimicking a real sample is in the clinically relevant range for early diagnosis of sepsis. These results demonstrate the validity of this prototype as an affordable, easy-to-use, compatible with micro-well arrays device for sepsis diagnosis ideal for hospital benches. Moreover, it can be extended to other biomarkers and fluid samples.

5:15pm - 5:30pm
ID: 400 / TOM4 S01: 4
TOM 4 Bio-Medical Optics

Physics-informed machine learning for microscopy

Emmanouil Xypakis1,2, Valeria deTuris1, Fabrizio Gala3, Giancarlo Ruocco1, Marco Leonetti1,2,4

1Center for Life Nano- and Neuro-Science, Istituto Italiano di Tecnologia, Viale Regina Elena 291, 00161, Rome, Italy; 2D-TAILS srl, 00161, Rome, Italy; 3Crestoptics S.p.A. (Italy); 4Soft and Living Matter Laboratory, Institute of Nanotechnology, Consiglio Nazionale delle Ricerche, 00185, Rome, Italy

We developed a physics-informed deep neural network architecture able to achieve signal to noise ratio improvement starting from low exposure noisy data. Our model is based on the nature of the photon detection process characterized by a Poisson probability distribution which we included in the training loss function. Our approach surpasses previous algorithms performance for microscopy data, moreover, the generality of the physical concepts employed here, makes it readily exportable to any imaging context.