Computational Imaging
Our Research
Our research focuses on optical imaging techniques that leverage computational, mathematical, and algorithmic methods to enhance or simplify conventional imaging paradigms. We develop advanced algorithms to correct aberrations, increase resolution, introduce novel contrast mechanisms, and elevate overall image quality.
- Holographic Optical Coherence Tomography
© Sarvesh Thakur / Dierck Hillmann What if you could see beneath the surface of living tissues, mapping their hidden layers with microscopic resolution, all without a single incision? Optical Coherence Tomography (OCT) achieves exactly that with something as simple as a Michelson interferometer. OCT utilizes low-coherence interferometry and Fourier-transform-based signal processing to reconstruct 3-dimensional volumetric images of biological tissues, such as the skin or a living retina, in microscopic detail.
Holographic OCT extends the capabilities of OCT by integrating methods from Digital Holography. This synergy enables advanced physics-based post-processing to digitally manipulate the focus and correct imaging aberrations.
In our research group, we employ a holographic Full-Field Fourier-Domain OCT (FF-FD-OCT) setup to investigate the function of various retinal neuronal layers. Our OCT system leverages a high-speed camera capable of capturing interferograms at 65,000 fps, making it 100x faster than traditional point-scanning OCT systems. Holographic OCT records interferograms of the wavefield at the sample plane as the laser sweeps across different wavelengths of infrared light. Once the wavefield is recovered, we harness the power of computational imaging to correct optical aberrations, such as defocus, in post-processing.
D. Hillmann, H. Spahr, C. Hain, H. Sudkamp, G. Franke, C. Pfäffle, C. Winter, and G. Hüttmann, “Aberration-free volumetric high-speed imaging of in vivo retina,” Sci. Rep., vol. 6, no. 1, 2016, doi: 10.1038/srep35209.
- Optoretinography
© Dierck Hillmann The retina is structured like an onion, with different layers harboring distinct functions, cells, and neurons. The photoreceptor layer is the first to respond to a light stimulus by absorbing photons. From there, the signal is transmitted through several layers of neurons and ultimately to the brain via the optic nerve. With holographic OCT, we can observe the activity of individual layers in the retina using a technique called optoretinography.
OCT signals encode information about optical path-length delays. As we stimulate specific regions of the retina, we can monitor how each layer responds to the stimulus by changing its (optical) thickness due to various physiological processes. In the future, this capability could prove invaluable in medical diagnostics for identifying dysfunctional cells or predicting the onset of certain chronic diseases.
D. Hillmann, H. Spahr, C. Pfäffle, H. Sudkamp, G. Franke, and G. Hüttmann, “In vivo optical imaging of physiological responses to photostimulation in human photoreceptors,” Proc. Natl. Acad. Sci. U.S.A., vol. 113, no. 46, pp. 13138–13143, 2016, doi: 10.1073/pnas.1606428113.
- Computational Imaging and Algorithm Development
© / Dierck Hillmann Essentially, we replace complex hardware solutions with sophisticated algorithmic software. To achieve this, we develop high-performance computational methods for image reconstruction and inverse problem-solving. Our work integrates Fourier-transform techniques, non-linear optimization, and deep learning, utilizing tools such as Python (NumPy, CuPy, PyTorch) and lower-level languages like C++ and CUDA.
