Deep Learning | MOL

Label-Free Intraoperative Imaging of Hemodynamics using Deep Learning

Wed, 18 Feb 2026 00:00:00 +0000

Computational Optics Wavefront Sensing and Phase Imaging Technology

Sun, 28 Dec 2025 00:00:00 +0000

On December 28, 2025, we were honored to host Prof. Liangcai Cao from the Department of Precision Instruments at Tsinghua University for an academic seminar titled “Computational Optics Wavefront Sensing and Phase Imaging Technology.”

Prof. Cao is a renowned expert in holographic display and computational imaging. During his presentation, he shared his group’s latest breakthroughs in integrating physical models with neural networks, specifically highlighting:

Propagation-Adaptive Holography: Utilizing physics-constrained spatial and Fourier neural operators to achieve high-fidelity 4K computer-generated holography.
Computational Wavefront Sensing: Implementing high-resolution wavefront characterization for complex wavefields in biological imaging.
Physics-Informed Deep Learning: Applications in adaptive aberration-corrected quantitative phase microscopy to overcome traditional imaging limits.

The seminar sparked extensive discussions among our group members regarding the intersection of random matrix theory and holographic reconstruction. We sincerely thank Prof. Cao for his profound insights and the valuable academic exchange.

Deep Holography, Quantum Imaging and Lensless 3D Endomicroscopy towards Advances in Biomedicine

Thu, 09 Oct 2025 00:00:00 +0000

On October 9, 2025, we were honored to host Prof. [cite_start]Juergen Czarske from the Dresden University of Technology (TUD), Germany, for an inspiring academic seminar titled “Deep Holography, Quantum Imaging and Lensless 3D Endomicroscopy towards Advances in Biomedicine”[cite: 1, 3].

[cite_start]Prof. Czarske is a Fellow of OPTICA, SPIE, and several other prestigious societies, and the recipient of the 2024 SPIE Dennis Gabor Award in Diffractive Optics[cite: 13, 15]. [cite_start]During the seminar, he presented a paradigm shift in biomedical imaging, moving from traditional bulky lens-based systems to lensless fiber endoscopes enabled by computational imaging[cite: 7, 8]. Key highlights included:

Physics-Informed Deep Learning: Exploiting speckle patterns to achieve high-fidelity 3D imaging and virtual staining of cancer cells[cite: 8, 9, 10].
[cite_start]Quantum Imaging Perspectives: Discussing the potential of using entangled photons and teleportation for advanced biomedical sensing[cite: 11].
[cite_start]Clinical Potential: How these advancements can detect and prevent disease development through early and gentle diagnostics[cite: 6].

[cite_start]The presentation sparked deep discussions on the intersection of deep holography and our group’s work in random matrix theory[cite: 12]. [cite_start]We thank Prof. Czarske for his insightful talk and for fostering international academic exchange between TUD, SIOM, and our laboratory[cite: 3, 5].

Snapshot Compressive Imaging: From Theory to Applications

Fri, 21 Mar 2025 00:00:00 +0000

On March 21, 2025, our group was honored to host Prof. Xin Yuan from the School of Engineering at Westlake University for an engaging academic seminar titled “Snapshot Compressive Imaging: From Theory to Applications.”

Prof. Yuan is a leading expert in the field of computational imaging and compressive sensing. During his presentation, he provided a comprehensive overview of his team’s significant contributions to Snapshot Compressive Imaging (SCI), focusing on:

Hardware-Algorithm Co-design: Innovative architectures for capturing high-dimensional data (video, hyperspectral) using 2D detectors through coded apertures.
Deep Learning for SCI: The transition from traditional optimization-based reconstruction to state-of-the-art deep learning models (such as PnP-ADMM and Transformers) for real-time, high-quality recovery.
Clinical and Industrial Applications: How SCI technology is revolutionizing ultra-high-speed photography and biomedical imaging, providing new tools for observing fast-evolving biological processes.

The seminar concluded with an interactive session where members of the Random Matrix Optics Research Group (RMORG) discussed the potential of combining SCI with statistical optics to tackle deep-tissue imaging challenges. We extend our gratitude to Prof. Yuan for sharing his visionary work.