Philip Kollmannsberger

This paper develops an automated deep learning pipeline for Diptera wing landmark and semilandmark annotation, achieving high accuracy comparable to human observers and outperforming existing methods.

This paper quantifies microtissue growth dynamics using optical flow and mitosis detection from label-free brightfield imaging, revealing distinct spatial patterns of cell proliferation and tissue expansion.

The paper introduces Mega-expansion microscopy, a fluorescence imaging technique that enables high-resolution structural visualization of whole cells using conventional microscopes.

This paper introduces programmable protein standards for benchmarking sub-10-nm fluorescence microscopy, demonstrating their utility in assessing fluorophore interactions and imaging performance.

The paper combines expansion microscopy with two-dye-imager DNA-PAINT to achieve nanometer-resolution 3D imaging of expanded cells.

This paper introduces a single-step expansion microscopy method enabling SMLM with spontaneously blinking dyes in expanded samples, achieving nanometer-resolution imaging of cellular structures.

The paper uses the Cellular Potts Model to show that tension-dependent local proliferation drives curvature-driven tissue growth via surface tension, without soluble growth factors.

This paper develops a geometric deep learning framework for improved particle averaging in single-molecule localization microscopy, achieving comparable results to state-of-the-art methods without explicit 3D alignment.

This study develops a CNN-based system for mosquito species identification using wing images, achieving high accuracy and demonstrating practical usability in real-world conditions.

The paper develops an optimized Ex-dSTORM method to resolve endogenous protein assemblies with nanometer precision in cells.

The paper improves single-molecule localization microscopy by extending temporal context using a hybrid U-Net and Transformer model, achieving better resolution than existing methods.

The paper uses endogenous tagging to study Unc-13 nanocluster reorganization during presynaptic homeostatic potentiation in Drosophila.

This study examines the ultrastructural changes in active zones of RIM1α knock-out mice compared to wild-type, revealing altered synaptic vesicle organization and architecture.

The paper introduces ReCSAI, a compressed sensing neural network for confocal dSTORM that improves reconstruction speed and accuracy by combining deep learning and compressed sensing.

The paper provides a detailed structural analysis of the anterior sensilla in C. elegans dauer larvae using FIB-SEM, revealing previously unknown neuronal structures.

The paper develops a method for 3D-dSTORM imaging of mouse brain sections to analyze presynaptic active zones, finding differences in Bassoon cluster size and density across hippocampal regions.

The paper investigates presynaptic homeostatic potentiation (PHP) in Drosophila melanogaster, finding that AZs compact rather than enlarge, increasing protein density without changing Brp copy numbers.

The paper introduces Haruspex, a convolutional neural network that automatically annotates Cryo-EM maps for protein secondary structures and RNA/DNA, achieving high recall and precision on test data.

The paper introduces an automated method to classify synaptic vesicles in C. elegans electron tomograms using machine learning, finding increased DCVs and greater distance from AZs in dauer larvae.