Stefan Ploner

Stefan Ploner, M. Sc.

Researcher

Department of Computer Science
Chair of Computer Science 5 (Pattern Recognition)

Room: Room 09.130
Martensstr. 3
91058 Erlangen

Office hours

Research Interest: Advanced Image Fusion in Optical Coherence Tomography and OCT Angiography

VISTA showing neovascularization having slow flow in PDR case
VISTA revealing slower blood flow in neovascularization (top-left of FAZ) in a motion-corrected OCTA scan of a 30 y/o PDR patient.

Since its recent transfer from research to the clinic, OCT Angiography (OCTA) is revolutionizing the field of ophthalmic Optical Coherence Tomography (OCT). Whereas OCTA provides detailed information on vasculature today, to take full advantage, novel image processing algorithms are yet to be developed. For example, algorithms like Variable Interscan Time Analysis (VISTA) could add the whole new dimension of blood flow speed to OCT data. An ongoing research topic of our group is the correction of scanning artifacts, including patient motion and varying illumination. Robust and accurate motion and illumination correction is essential for high quality reconstructions, however, current algorithms do not model the scanned imaging precisely, are challanged by self-overlapping OCTA scan patterns, and often have calculation times that are impractical for clinical workflows.

The goal of my PhD is to fill these gaps and to bring these promising advancements to the clinic. To make this possible, I’m in close collaboration with the Biomedical Optical Imaging and Biophotonics Group at the Massachusetts Institute of Technology, the group who built the first and still develops most advanced ophthalmic OCT systems to date. Our developments are transferred to clinical application at the New England Eye Center, the first clinic that used OCT technology in patient care. To meet workflow requirements, I jointly consider algorithmic complexity as well as recent hardware architectures to provide most accurate results in outstanding runtimes.

Academic CV

Since 07/2017: PhD Student, Pattern Recognition Lab, Friedrich-Alexander University Erlangen-Nürnberg

  • Research focus on motion correction and enhanced signal reconstruction in the field
    of Optical Coherence Tomography (OCT) and OCT angiography (OCTA)
  • In collaboration with the Biomedical Optical Imaging and Biophotonics Group, MIT
  • Published articles on motion correction at MICCAI and in Biomedical Optics Express (among the ~5% highlighted as Editor’s pick)
  • Journal reviewer for BOE, OE, IOVS (“exceptional reviewer”), MedPhys, JBio, JBO, Optics Letters, TVST, and others

10/2014 – 05/2017: Master of Science with distinction in Computer Science, Friedrich-Alexander University Erlangen-Nürnberg

  • Focus on medical image processing, pattern recognition and high-performance computing
  • Master thesis: Improving 3D OCT motion correction (oral presentation at the ARVO annual meeting)

10/2015 – 05/2016: Visiting student, Massachusetts Institute of Technology, Cambridge, USA

04/2011 – 03/2015: Bachelor of Science in Computer Science, Friedrich-Alexander University Erlangen-Nürnberg

  • Focus on pattern recognition, computer architecture, discrete optimization and computer graphics
  • Student research assistant at the Pattern Recognition Lab, research on false
    color visualization in the multispectral imaging software framework Gerbil

09/2001 – 06/2010: University entrance diploma (Abitur), Emil-von-Behring-Gymnasium, Spardorf

Teaching Experience

  • 10/2014 – 01/2015: Student teaching assistant, Friedrich-Alexander University Erlangen-Nürnberg
    • Exercises in Theory of Computation and Formal Languages
  • 10/2011 – 09/2012: Student teaching assistant, Friedrich-Alexander University Erlangen-Nürnberg
    • Exercises in Algorithms and Data Structures

Projects

2022

  • Temporally resolved 3-D retinal blood flow quantification using advanced motion correction and signal reconstruction in optical coherence tomography angiography

    (Third Party Funds Single)

    Term: since November 15, 2022
    Funding source: DFG-Einzelförderung / Sachbeihilfe (EIN-SBH)

    Die optische Kohärenztomographie (OCT) erzeugt volumetrische 3-D-Bilder von Gewebe mit Mikrometerauflösung, indem sie einen Laserstrahl zum Scannen verwendet und die Amplitude und Zeitverzögerung von zurückgestreutem Licht misst. Die OCT hat einen großen Einfluss auf die Augenheilkunde und wurde zu einer Standard-Bildgebungsmodalität für die Diagnose, die Überwachung des Krankheitsverlaufs und das Ansprechen auf die Behandlung sowie für die Untersuchung der Pathogenese von Krankheiten wie diabetischer Retinopathie, altersbedingter Makuladegeneration und Glaukom. Die jüngste Entwicklung der OCT-Angiographie (OCTA) hat die grundlegende und klinische Forschung dramatisch beschleunigt. OCTA führt eine tiefenaufgelöste (3-D) Bildgebung der retinalen Mikrovaskulatur durch, indem es wiederholt die gleiche Netzhautposition abbildet und den Bewegungskontrast von sich bewegenden Blutzellen erkennt. Im Vergleich zu herkömmlichen Ansätzen, die auf injizierten Kontrastmitteln basieren, hat OCTA den Vorteil, dass es nicht invasiv ist, sodass die Bildgebung bei jedem Patientenbesuch durchgeführt werden kann, was Längsschnittstudien ermöglicht. Allerdings hat OCTA auch einige Einschränkungen. Da eine wiederholte Bildgebung erforderlich ist, um den Blutfluss zu erkennen, sind die Aufnahmezeiten lang und die Daten können durch Augenbewegungen und Bildartefakte verzerrt werden, was eine quantitative Längsschnittanalyse erschwert. OCTA-Algorithmen können das Vorhandensein eines Blutflusses erkennen, sind jedoch nur begrenzt in der Lage, subtile Veränderungen des Flusses aufzulösen, die frühe Anzeichen einer Krankheit sein können. Zeitliche Schwankungen des Flusses, die durch den Herzzyklus oder die funktionelle Reaktion der Netzhaut verursacht werden, sind schwer zu untersuchen. Wir schlagen vor, ein neues Framework für OCTA zu entwickeln, das eine Bewegungskorrektur auf Kapillarebene ermöglicht, Blutflussgeschwindigkeiten differenziert und eine Analyse auf mehreren Zeitskalen ermöglicht (4-D OCTA). Die Fähigkeit, über die Visualisierung der Mikrovaskulatur hinauszugehen und den Fluss und seine zeitlichen Schwankungen zu beurteilen, ermöglicht die Beurteilung subtiler Beeinträchtigungen der mikrovaskulären Perfusion sowie des Herzzyklus und der Reaktion auf funktionelle Stimulation. In Kombination mit der vaskulären strukturellen Bildgebung versprechen diese Fortschritte, neue Krankheitsmarker in früheren Krankheitsstadien bereitzustellen, eine genauere Messung des Krankheitsverlaufs und des Ansprechens auf die Therapie in pharmazeutischen Studien zu ermöglichen und zur Aufklärung der Pathogenese bei Netzhauterkrankungen beizutragen.

2017

  • Joint Iterative Reconstruction and Motion Compensation for Optical Coherence Tomography Angiography

    (Third Party Funds Single)

    Term: August 1, 2017 - July 31, 2019
    Funding source: DFG-Einzelförderung / Sachbeihilfe (EIN-SBH)

    Optical coherence tomography (OCT) is a non-invasive 3-D optical imagingmodality that is a standard of care in ophthalmology [1,2]. Since the introduction of Fourier-domain OCT [3], dramatic increases in imaging speedbecame possible, enabling 3-D volumetric data to be acquired. Typically, aregion of the retina is scanned line by line, where each scanned lineacquires a cross-sectional image or a B-scan. Since B-scans are acquiredin milliseconds, slices extracted along a scan line, or the fast scanaxis, are barely affected by motion. In contrast, slices extractedorthogonally to scan lines, i. e. in slow scan direction, areaffected by various types of eye motion occurring throughout the full,multi-second volume acquisition time. The most relevant types of eyemovements during acquisition are (micro-)saccades, which can introducediscontinuities or gaps between B-scans, and slow drifts, which causesmall, slowly changing distortion [4]. Additional eye motion is caused by pulsatile blood flow,respiration and head motion. Despite ongoing advances in instrumentscanning speed [5,6] typical volume acquisition times havenot decreased. Instead, the additional scanning speed is used for densevolumetric scanning or wider fields of view [7]. OCT angiography (OCTA) [811] multiplies therequired number of scans by at least two, and even more scans are neededto accommodate recent developments in blood flow speed estimation whichare based on multiple interscan times [12,13]. As a consequence,there is an ongoing need for improvement in motion compensation especiallyin pathology [1416].

    We develop novel methods for retrospective motion correction of OCT volume scans of the anterior and posterior eye, and widefield imaging. Our algorithms are clinically usable due to their suitability for patients with limited fixation capabilities and increased amount of motion, due to their fast processing speed, and their high accuracy, both in terms of alignment and motion correction. By merging multiple accurately aligned scans, image quality can be increased substantially, enabling the inspection of novel features.

Publications

2023

Journal Articles

Conference Contributions

2022

Journal Articles

Conference Contributions

2021

Journal Articles

Conference Contributions

2020

Journal Articles

Conference Contributions

2019

Journal Articles

Conference Contributions

2018

Journal Articles

Book Contributions

Conference Contributions

2017

Journal Articles

Conference Contributions

2016

Journal Articles

Conference Contributions

Lectures

No matching records found.