Computational miniature mesoscope for large-scale 3D fluorescence imaging
2022
Online
Hochschulschrift
Zugriff:
Fluorescence imaging is indispensable to biology and neuroscience. The need for large-scale imaging in freely behaving animals has further driven the development in miniaturized microscopes (miniscopes). However, conventional microscopes and miniscopes are inherently constrained by their limited space-bandwidth-product, shallow depth-of-field, and inability to resolve 3D distributed emitters such as neurons. In this thesis, I present a Computational Miniature Mesoscope (CM2) that leverages two computation frameworks to overcome these bottlenecks and enable single-shot 3D imaging across a wide imaging field-of-view (FOV) of 7~8 mm and an extended depth-of-field (DOF) of 0.8~2.5 mm with a high lateral (7 um) and axial resolution (25 um). The CM2 is a novel fluorescence imaging device that achieves large-scale illumination and single-shot 3D imaging on a compact platform. This expanded imaging capability is enabled by computational imaging that jointly designs optics and algorithms. In this thesis, I present two versions of CM2 platforms and two 3D reconstruction algorithms. In addition, pilot studies of in vivo imaging experiments using a wearable CM2 prototype are conducted to demonstrate the CM2 platform's potential applications in large-scale neural imaging. First, I present the CM2 V1 platform and a model-based 3D reconstruction algorithm. The CM2 V1 system has a compact lightweight design that integrates a microlens array (MLA) for 3D imaging and an LED array for excitation on a single compact platform. The model-based 3D deconvolution algorithm is developed to perform volumetric reconstructions from single-shot CM2 measurements, achieving 7 um lateral and 200 um axial resolution across a wide 8 mm FOV and 2.5 mm DOF in clear volumes. This mesoscale 3D imaging capability of CM2 is validated on various fluorescent samples, including resolution target, fibers, and particle phantoms in different geometry. I further quantify the effects of bulk scattering and background fluorescence in phantom experiments. Next, I investigate and improve the CM2 V1 system for both the hardware and the reconstruction algorithm. Specially, the low axial resolution (200 um), insufficient excitation efficiency (24%), and heavy computational cost of the model-based 3D deconvolution hinder CM2 V1's biomedical applications. I present and demonstrate an upgraded CM2 V2 platform augmented with a deep learning-based 3D reconstruction framework, termed CM2Net, to address the above limitations. Specially, the CM2 V2 design features an array of freeform illuminators and hybrid emission filters to achieve 3 times higher excitation efficiency (80%) and 5 times better suppression of background fluorescence, compared to the V1 design. The multi-stage CM2Net combines ideas from view demixing, lightfield refocusing and view synthesis to account for the CM2’s multi-view geometry and achieve reliable 3D reconstruction with high axial resolution. Finally, trained purely on simulated data, I show that the CM2Net can generalize to experimental measurements. A key element of CM2Net's generalizability is a 3D Linear Shift Variant (LSV) model of CM2 that simulates realistic measurements by accurately incorporating field varying aberrations. I experimentally validate the CM2 V2 platform and CM2Net achieve faster, artifact-free 3D reconstructions across a 7 mm wide FOV and 800 um DOF with 25 um axial and 7 um lateral resolution in phantom experiments. Compared to the CM2 V1 with model-based deconvolution, the CM2Net achieves a 10 times better axial resolution at 1400 times faster reconstruction speed without sacrificing the imaging FOV or lateral resolution. The new system design of CM2 V2 with the LSV-embedded CM2Net provides an intriguing solution to large-scale fluorescence imagers with a small form factor. Built from off-the-shelf and 3D printed components, I envision that this low-cost and compact computational imaging system can be adopted in various biomedical and neuroscience labs. The CM2 systems and the developed computational tools can have impact in a wide range of large-scale 3D fluorescence imaging applications.
Titel: |
Computational miniature mesoscope for large-scale 3D fluorescence imaging
|
---|---|
Autor/in / Beteiligte Person: | Xue, Yujia |
Link: | |
Veröffentlichung: | 2022 |
Medientyp: | Hochschulschrift |
Schlagwort: |
|
Sonstiges: |
|