Skip to main content

Global integrative framework for Computational Bio-Imaging

Periodic Reporting for period 3 - GlobalBioIm (Global integrative framework for Computational Bio-Imaging)

Reporting period: 2019-10-01 to 2021-03-31

One needs vast amounts of acquired data (Fourier samples for MRI, projections for x-ray imaging) to reconstruct high-quality medical and biological images. For efficiency purposes, the classical strategy is to proceed by batch, leaving the consolidation of the batch-based reconstructions to a later stage. Focusing instead on quality and resolution, we propose an alternative strategy in which we consider the data jointly, which in turn requires the precise reallocation of the measurements in 3D space/time before all the available information can be integrated; then only will the underlying complete signal be reconstructed. This joint reconstruction will result from a very large-scale optimization that exploits the redundancy of the signal (sparsity, spatiotemporal correlations) to improve the solution. Due to recent advances in the field, we are arguing that such a big-data integration is now within reach and that our team is ideally qualified to lead the way. A successful outcome will have a profound impact on the design of future bioimaging systems.

We are proposing a unifying framework for the development of such next-generation reconstruction algorithms. It involves a clear separation between the physical (forward model) and signal-related (regularization, incorporation of prior constraints) aspects of the problem. The three pillars of our formulation are: an operator algebra with a corresponding set of fast linear solvers; an advanced statistical framework for the principled derivation of reconstruction methods; learning schemes for the optimization and self-tuning of parameters. These core technologies are incorporated into a modular software library (GlobalBioIm) that features key components to implement and test iterative reconstruction algorithms. We are applying our framework to improve upon the state-of-the-art in the following modalities:
1) phase-contrast x-ray tomography in full 3D;
2) structured-illumination microscopy;
3) single-particle analysis in cryo-electron tomography;
4) a novel multipose fluorescence microscopy;
5) real-time MRI;
6) a new multimodal digital microscope.
In all instances, we are working in close collaboration with the imaging scientists who are in charge of the instrumentation.

In short, the two primary objectives of GlobalBioIm are
(i) to improve the resolution and quality of images obtained from a wide range of modalities using data integration and advanced processing techniques,
(ii) to provide imaging scientists with a unified software framework that facilitates and speeds-up the development of state-of-the-art image-reconstruction algorithms.

Due to the widespread use of imaging in medicine and biology, the project should contribute to the advancement of science and to the improvement of image-based diagnostics (e.g. radiology).
The work is divided into two main parts: a conceptual/enabling component that takes the form of a unified software library (GlobalBioIm) which is common to all sub-projects; and an application-oriented part that is devoted to the development of specialized reconstruction algorithms for a variety of imaging modalities.

During this term, we have set the foundations of the common software framework (GlobalBioIm) and made it available to the community at
The leading idea of our system is to decouple the physical aspect of the problem, which is application-specific, from the so-called “regularization” of the solution. Regularization dictates how to properly constrain the reconstruction and is often the same across modalities. Once equipped with these two components, the user has to specify a suitable cost functional, together with an adequate method of optimization. The desired image-reconstruction algorithm will then be automatically assembled by GlobalBioIm. Since the critical part of this approach is the choice of the regularization, we have also performed a theoretical investigation of its effect.
Finally, we have proposed an alternative approach to image reconstruction that uses convolutional neural networks and that is capable of outperforming the state-of-the-art.

On the application side, we have already implemented
1) a high-end reconstruction method for x-ray computer tomography with parallel rays;
2) a new algorithm for structured-illumination microscopy that uses computational sectioning;
3) a compressed-sensing scheme for scanning electron microscopy that reduces radiation exposure;
4) a new multiresolution reconstruction method for cryo-EM tomography;
5) a method to improve the axial resolution of fluorescence microscopy through the combination of multiple views of particle replicates that differ by their orientation;
6) a novel method to retrieve the optical refractive index of a specimen based on the solution of the inverse scattering problem.
Our results clearly demonstrate the great potential of machine-learning techniques (in particular, neural networks) for image reconstruction. This is undoubtedly a most promising direction of research that needs to be further explored and applied/tested with more diverse modalities. The groundwork we performed during the first term of the project left us ideally positioned to pursue this endeavor, and we aim at demonstrating that advanced data processing can dramatically increase image quality. Equivalently, we aim at producing acceptable images from fewer (or noisier) measurements, which translates into a corresponding reduction of the radiation dosage.

The other important topic is dynamic MRI. There, we have encouraging preliminary indications that our framework will result in good progress, too.