Can we expect microscopy go 3D instead of 2D...not only light but electron microscopy

Can anyone tell someone working on microscopy in depth from light microscopy to electron microscopy to atomic force microscopy… And making this microscopy into 3D for better understanding?? Please … anyone?

Not sure if I completely understand your question, but 3D (even 4D, 3D over time) has been happening for a long time and continues to improve! Cheers.

There’s an overview of some 3D electron microscopy techniques for bioimaging here: https://www.sciencedirect.com/science/article/pii/S0968432814000250

1 Like

The short answer is that you have many good options now. Imaris software (for example, no commercial interest) has long specialized in 3D fluorescence measurement, and I believe there are great tools for reconstructing block face SEM serial sections into cells and organelles.

However I think you hit an under-appreciated point here. 3D analysis is hard! You can’t just hand-draw boundaries like you do with 2D images, and intensity/shape-based structure detection doesn’t always work. I suspect light sheet never quite took off as as a technology at least in part because the average user has a hard time with volume data. In my view one of the most exciting things we could see in the next few years would be an application that uses machine learning to automatically segment 3D structures. I’ve seen some promising reports in the primary literature, but so far no one has anything you can buy with a user manual and vendor support. Some day, I hope.

Hi @Tim_F

There are many efforts in that direction. In addition to Imaris, In the commercial space, Airvia by DRVision, apparently has Deep Learning and handles terrabyte sized data (no commercial interest).

In the open source space there is Care for deep learning, Paintera for visualization and 3D annotation, and a whole family of technologies Big Stitcher, SciView, Big Data Viewer and others that integrate with Imagej and can be used for stitching, segmentation and visualization of big data.

In terms of support for the open source efforts, there is a sister forum to this one, image.cs, https://forum.image.sc/, which has 30 open source projects using the forum as their support channel. The forum is essentially the “vendor support” for open source. In addition several commercial companies post there as well. So it’s probably the best source to get further information about the state-of-the-art in ND image processing.

Thanks Brian!! I am excited to try that out.