Source author record

Martin Wagner

Martin Wagner appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

cond-mat.mes-hall Computer Vision cond-mat.mtrl-sci cond-mat.str-el eess.IV Machine Learning

Catalog footprint

What is connected

7works

6topics

4close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2022arXiv

Robust deep learning-based semantic organ segmentation in hyperspectral images

Semantic image segmentation is an important prerequisite for context-awareness and autonomous robotics in surgery. The state of the art has focused on conventional RGB video data acquired during minimally invasive surgery, but full-scene semantic segmentation based on spectral imaging data and obtained during open surgery has received almost no attention to date. To address this gap in the literature, we are investigating the following research questions based on hyperspectral imaging (HSI) data of pigs acquired in an open surgery setting: (1) What is an adequate representation of HSI data for neural network-based fully automated organ segmentation, especially with respect to the spatial granularity of the data (pixels vs. superpixels vs. patches vs. full images)? (2) Is there a benefit of using HSI data compared to other modalities, namely RGB data and processed HSI data (e.g. tissue parameters like oxygenation), when performing semantic organ segmentation? According to a comprehensive validation study based on 506 HSI images from 20 pigs, annotated with a total of 19 classes, deep learning-based segmentation performance increases, consistently across modalities, with the spatial context of the input data. Unprocessed HSI data offers an advantage over RGB data or processed data from the camera provider, with the advantage increasing with decreasing size of the input to the neural network. Maximum performance (HSI applied to whole images) yielded a mean DSC of 0.90 ((standard deviation (SD)) 0.04), which is in the range of the inter-rater variability (DSC of 0.89 ((standard deviation (SD)) 0.07)). We conclude that HSI could become a powerful image modality for fully-automatic surgical scene understanding with many advantages over traditional imaging, including the ability to recover additional functional tissue information. Code and pre-trained models: https://github.com/IMSY-DKFZ/htc.

preprint2021arXiv

Heidelberg Colorectal Data Set for Surgical Data Science in the Sensor Operating Room

Image-based tracking of medical instruments is an integral part of surgical data science applications. Previous research has addressed the tasks of detecting, segmenting and tracking medical instruments based on laparoscopic video data. However, the proposed methods still tend to fail when applied to challenging images and do not generalize well to data they have not been trained on. This paper introduces the Heidelberg Colorectal (HeiCo) data set - the first publicly available data set enabling comprehensive benchmarking of medical instrument detection and segmentation algorithms with a specific emphasis on method robustness and generalization capabilities. Our data set comprises 30 laparoscopic videos and corresponding sensor data from medical devices in the operating room for three different types of laparoscopic surgery. Annotations include surgical phase labels for all video frames as well as information on instrument presence and corresponding instance-wise segmentation masks for surgical instruments (if any) in more than 10,000 individual frames. The data has successfully been used to organize international competitions within the Endoscopic Vision Challenges 2017 and 2019.

preprint2020arXiv

Robust Medical Instrument Segmentation Challenge 2019

Intraoperative tracking of laparoscopic instruments is often a prerequisite for computer and robotic-assisted interventions. While numerous methods for detecting, segmenting and tracking of medical instruments based on endoscopic video images have been proposed in the literature, key limitations remain to be addressed: Firstly, robustness, that is, the reliable performance of state-of-the-art methods when run on challenging images (e.g. in the presence of blood, smoke or motion artifacts). Secondly, generalization; algorithms trained for a specific intervention in a specific hospital should generalize to other interventions or institutions. In an effort to promote solutions for these limitations, we organized the Robust Medical Instrument Segmentation (ROBUST-MIS) challenge as an international benchmarking competition with a specific focus on the robustness and generalization capabilities of algorithms. For the first time in the field of endoscopic image processing, our challenge included a task on binary segmentation and also addressed multi-instance detection and segmentation. The challenge was based on a surgical data set comprising 10,040 annotated images acquired from a total of 30 surgical procedures from three different types of surgery. The validation of the competing methods for the three tasks (binary segmentation, multi-instance detection and multi-instance segmentation) was performed in three different stages with an increasing domain gap between the training and the test data. The results confirm the initial hypothesis, namely that algorithm performance degrades with an increasing domain gap. While the average detection and segmentation quality of the best-performing algorithms is high, future research should concentrate on detection and segmentation of small, crossing, moving and transparent instrument(s) (parts).

preprint2015arXiv

Graphene-Based Platform for Infrared Near-Field Nanospectroscopy of Water and Biological Materials in an Aqueous Environment

Scattering scanning near-field optical microscopy (s-SNOM) has emerged as a powerful nanoscale spectroscopic tool capable of characterizing individual biomacromolecules and molecular materials. However, applications of scattering-based near-field techniques in the infrared (IR) to native biosystems still await a solution of how to implement the required aqueous environment. In this work, we demonstrate an IR-compatible liquid cell architecture that enables near-field imaging and nanospectroscopy by taking advantage of the unique properties of graphene. Large-area graphene acts as an impermeable monolayer barrier that allows for nano-IR inspection of underlying molecular materials in liquid. Here, we use s-SNOM to investigate the tobacco mosaic virus (TMV) in water underneath graphene. We resolve individual virus particles and register the amide I and II bands of TMV at ca. 1520 and 1660 cm$^{-1}$, respectively, using nanoscale Fourier transform infrared spectroscopy (nano-FTIR). We verify the presence of water in the graphene liquid cell by identifying a spectral feature associated with water absorption at 1610 cm$^{-1}$.

preprint2015arXiv

Phase transition in bulk single crystals and thin films of VO2 by nano-infrared spectroscopy and imaging

We have systematically studied a variety of vanadium dioxide (VO2) crystalline forms, including bulk single crystals and oriented thin films, using infrared (IR) near-field spectroscopic imaging techniques. By measuring the IR spectroscopic responses of electrons and phonons in VO2 with sub-grain-size spatial resolution (~20 nm), we show that epitaxial strain in VO2 thin films not only triggers spontaneous local phase separations but also leads to intermediate electronic and lattice states that are intrinsically different from those found in bulk. Generalized rules of strain and symmetry dependent mesoscopic phase inhomogeneity are also discussed. These results set the stage for a comprehensive understanding of complex energy landscapes that may not be readily determined by macroscopic approaches.

preprint2014arXiv

Ultrafast and Nanoscale Plasmonic Phenomena in Exfoliated Graphene Revealed by Infrared Pump-Probe Nanoscopy

Pump-probe spectroscopy is central for exploring ultrafast dynamics of fundamental excitations, collective modes and energy transfer processes. Typically carried out using conventional diffraction-limited optics, pump-probe experiments inherently average over local chemical, compositional, and electronic inhomogeneities. Here we circumvent this deficiency and introduce pump-probe infrared spectroscopy with ~20 nm spatial resolution, far below the diffraction limit, which is accomplished using a scattering scanning near-field optical microscope (s-SNOM). This technique allows us to investigate exfoliated graphene single-layers on SiO2 at technologically significant mid-infrared (MIR) frequencies where the local optical conductivity becomes experimentally accessible through the excitation of surface plasmons via the s-SNOM tip. Optical pumping at near-infrared (NIR) frequencies prompts distinct changes in the plasmonic behavior on 200 femtosecond (fs) time scales. The origin of the pump-induced, enhanced plasmonic response is identified as an increase in the effective electron temperature up to several thousand Kelvin, as deduced directly from the Drude weight associated with the plasmonic resonances.

preprint2010arXiv

Analysis of Thermoelectric Properties of Scaled Silicon Nanowires Using an Atomistic Tight-Binding Model

Low dimensional materials provide the possibility of improved thermoelectric performance due to the additional length scale degree of freedom for engineering their electronic and thermal properties. As a result of suppressed phonon conduction, large improvements on the thermoelectric figure of merit, ZT, have been recently reported in nanostructures, compared to the raw materials' ZT values. In addition, low dimensionality can improve a device's power factor, offering an additional enhancement in ZT. In this work the atomistic sp3d5s*-spin-orbit-coupled tight-binding model is used to calculate the electronic structure of silicon nanowires (NWs). The Landauer formalism is applied to calculate an upper limit for the electrical conductivity, the Seebeck coefficient, and the power factor. We examine n-type and p-type nanowires of diameters from 3nm to 12nm, in [100], [110], and [111] transport orientations at different doping concentrations. Using experimental values for the lattice thermal conductivity in nanowires, an upper limit for ZT is computed. We find that at room temperature, scaling the diameter below 7nm can at most double the power factor and enhance ZT. In some cases, however, scaling does not enhance the performance at all. Orientations, geometries, and subband engineering techniques for optimized designs are discussed.