Here Is the Perfect Solution if You Want to Bring Bioimage Analysis Into Your Lab

Did you know that you can save over 90% of time spent on data analysis by moving from manual to automated segmentation?

In a study evaluating variability in manual and automatic segmentation of neuroblastic tumors it was shown that time spent segmenting images could be reduced from 56 minutes of manual drawing to 4 minutes of verification and tweaking of computer generated selections (Veiga-Canuto et al. Cancers 2022).

And on top of that the end results were better too.

Challenges with bioimage analysis

Here is the problem you face: your lab has started capturing images as primary sources of data, but you are struggling to extract quantitative information from them.

This is not surprising — image analysis is a complex computer science domain in its own right. Furthermore, bioimage analysis comes with its own set of challenges including:

Inconsistencies during image capture
Managing huge volumes of data
The noisy nature of biological samples

Frustratingly the images are often readily interpreted by a trained biologist — but how do you scale that? And how do you mitigate human bias?

If you are struggling with any of the challenges above then I might be able to help!

Bringing bioimage analysis expertise into your lab

My bioimage analysis services provides you with:

Rapid feedback on preliminary data so that you can optimise image capture for computational analysis — meaning that you will save time, effort and money avoiding collecting images that are difficult or impossible to analyse.
Regular progress reports so you know what is going on at all times — one less thing to worry about.
Scripted analysis to minimise human bias so that you can trust the results — and avoid embarrassment and false claims.
Automation so you can process more images, meaning that you will have higher impact results for your grant applications and manuscripts.
Access to data analysis scripts so your lab can reuse the analysis and its constituent parts — building the foundations for your lab to upskill and grow.
The opportunity to get a trusted partner for your scientific computing needs so that you can design better experiments and write more compelling grant applications meaning your lab can become even more successful.

Background

My name is Tjelvar Olsson and I’m a biochemistry PhD. However, my diverse experience in both academia and the pharma industry has led me to open my own business — Cyborg Interfaces Ltd. I offer all kinds of support for researchers working at the interface between science and computing.

And one such support service just happens to be image analysis!

Depending on the scale at which you are working you may also be happy to hear that I have extensive experience of data management, high-performance computing, and software engineering as well. More details on my professional background can be found on LinkedIn, and you can see my publication record on Google Scholar.

Briefly — in the context of bioimage analysis — here are three papers that feature my pragmatic approach.

Duncan, S., Olsson, T. S. G., Hartley, M., Dean, C., & Rosa, S. (2016). A method for detecting single mRNA molecules in Arabidopsis thaliana. Plant Methods, 12(1). 10.1186/s13007-016-0114-x.

This paper contains a fairly standard analysis where tissue had to be segmented into cells for spot counts to be reported per cell. In this case the benefit came from the automation as there were many images, with many cells, and many many many spots to count.

**Figure 5 from Duncan et al, 2016.** Automated image analysis of PP2A mRNA. a Representative maximum projection image of cell files labeled with PP2A mRNA probes (red). DNA labeled with DAPI (blue). b, c Screen shots showing sequential detection steps used to determine positive mRNA signals. d Cell segmentation, where a false-color is rendering individual cells. e Output image indicating the number of mRNA signals detected on each cell segmented in (d). Scale bar = 10 μm

Yang, H., Berry, S., Olsson, T. S. G., Hartley, M., Howard, M., & Dean, C. (2017). Distinct phases of Polycomb silencing to hold epigenetic memory of cold in Arabidopsis. Science, 357(6356), 1142-1145 10.1126/science.aan1121

This study included confocal microscopy data of root tissues, where the intensity of a fluorescent probe needed to be quantified per cell. This required roots to be segmented into cells in 3D. However, analysis of preliminary data revealed that the microcopy strategy was ill suited for 3D segmentation. In this case the value of my work came from suggesting an alternative microscopy strategy resulting in more z-slices, along with custom Python scripts for 3D segmentation allowing the quantification of fluorescent probe signal per cell.

Mansfield, C., Newman, J. L., Olsson, T. S. G., Hartley, M., Chan, J., & Coen, E. (2018). Ectopic BASL Reveals Tissue Cell Polarity throughout Leaf Development in Arabidopsis thaliana. Current Biology, 28(16), 2638–2646.e4. 10.1016/J.CUB.2018.06.019

The image analysis pipeline created for this paper, required one step with input from a trained biologist. This resulted in challenges with human bias and manual labour. The value I offered to this study was to create tooling to mitigate human bias, minimise the number of manual button clicks, and keep track of the data and metadata generated by the button clicks of the expert biologist.

**Extract from Figure S1 Mansfield et al 2018.** (G) Image processing pipeline. (H) Raw confocal data is automatically segmented. (I) Individual cells are identified and position of centroid (cross) is extracted. Yellow arrows indicate cell rotation. (J)Individual cell is randomly rotated in one of 4 orientations. (K and L) BASL signal is identified from merged image of cell and separate colour channels for clearer visualisation. (M) BASL signal marked by hand (indicated by asterisk) to create vector, indicated by white arrow. (N) Cell and vector are rotated back into original position. (O) Process repeated for every segmented cell to produce a vector field for the leaf. White dashed line in H and O indicates leaf outline.

What others are saying

“It was great working with Tjelvar. His background in biochemistry made it easy to talk biology with him, and his computational expertise and analysis allowed me to extract key information from my bioimages. Further, Tjelvar's feedback on preliminary data helped me optimise image capture to ensure successful 3D segmentation of cells. Throughout the project Tjelvar provided regular updates to help me stay on top of progress. I'm happy to say that the analysis provided by Tjelvar contributed to my research being published in Science.”

— Dr Scott Berry, Group Leader, University of New South Wales

“Tjelvar was always helpful, seeking out solutions to analyse a particular set of our leaf microscopy images. He listened carefully to problems then made intelligent suggestions as to the best way forward and implemented them. A great colleague and a pleasure to work with.”

— Prof Enrico Coen, Group Leader, John Innes Centre

"During my PhD, I faced the challenge of quantifying single RNA transcripts in plant cells, but the sheer volume of microscopy data made manual analysis impractical. Tjelvar helped me overcome this by developing a bespoke automated image analysis workflow that accurately extracted RNA transcript counts from segmented cell areas."

"This pipeline has been an indispensable tool in my research, enabling quantitative insights into diverse gene regulatory mechanisms, including sense/antisense transcription and boron transport. Beyond his technical expertise, Tjelvar was a thoughtful and collaborative colleague who translated complex computational methods into accessible solutions for biologists."

— Dr Susan Duncan, Postdoctoral Scientist, John Innes Centre

“I have the pleasure to be working with Tjelvar Olsson on questions of scientific data management. Tjelvar has experience in working with people from a variety of backgrounds (science, engineering, experiment, theory, etc.), which helps him to quickly understand the core data management requirements when approaching him. I recommend him very highly.”

— Prof Lars Pastewka, Group Leader, University of Freiburg

“Tjelvar is friendly, very capable and dependable. He has an excellent understanding of the demands of biologists. He has helped me and my lab members on numerous occasions with our queries and finding creative solutions to safely store and retrieve our data cost-efficiently.”

— Dr Brande Wulff, Group Leader, John Innes Centre

“Not to be forgotten – Tjelvar dares asking the tough questions and is a good critic. Few things escape his scrutiny, which is also a vital ability in Science; it is better to have your colleagues find the errors than the peer reviewers!”

— Dr Mats Linder, Circular Economy expert and strategist

Pricing

The services I provide include:

Advice on experimental setup
Rapid feedback on preliminary data
Regular progress reports
Built-in reproducibility
Access to analysis scripts
Support writing up materials and methods as well as research grants (EU Horizon PIC registered)

Pricing will depend on the complexity of the project, but typical projects fall in the range of £4,000 to £8,000.

Bonus — 2 days of free training

To ensure continuity within your lab all projects over £4,000 come with 2 days of free online training in bioimage analysis and scientific computing. This can be used to help build up expertise within your lab.

Guarantee

If you have existing data I will carry out a feasibility study before the project begins. That way you can see what type of information you can expect to extract from your images before paying.

Reach out now to book a free consultation

If you are interested in getting support with your bioimage analysis I’d love to have a chat to learn about your images and research, and for us to work out if we are a good match.

Schedule a Free Consultation Now

Looking forward to hearing from you!

PS — Your success is my priority, so I will not take on more work than I can handle. Reach out now to secure a space.