Spatial biology can map gene expression within tissue architecture, giving researchers context that single-cell methods lose. Despite strong scientific interest, the field faces persistent barriers to clinical use: software that requires deep bioinformatics expertise, a lack of standardised workflows, high instrument costs, and a shortage of trained users. As of 2026, most spatial platforms remain research tools rather than clinical diagnostics
Spatial biology promises unprecedented insight into biology and disease – but is the field living up to its potential?
In this in-depth conversation, Dr Yuliia Shymko sit down with Dr. Giovanni Russo (Gio) to explore what’s truly holding spatial biology back. From the long-standing tension between academia and industry to persistent challenges in software, data interpretation, and training, this discussion goes beyond the hype to examine why clinical adoption remains so difficult.
Gio brings a rare cross-functional perspective, drawing on experience across research, field service, sales, and commercialization. Together, we unpack where spatial technologies are delivering real value today, which platforms are closest to clinical impact, and what must change over the next five years for the field to move forward.
Topics covered in our Spatial Biology talk include:
- Why spatial isn’t new – and why it’s treated like it is
- Academia vs. industry: barriers to commercialization
- Software and data analysis as the true bottlenecks
- Training bias and misinterpretation of complex datasets
- Why most spatial technologies stay in academia
- Revenue drivers, sales challenges, and real-world adoption
- What actually makes spatial worth the hype
Can’t watch right now? Read the full transcript of our conversation with Dr. Giovanni Russo below.
Dr. Yuliia Shymko: Hey guys, this is the first video we are filming dedicated to our guests in life sciences, multi-omics, spatial biology, and other exciting topics. Today we have our first guest, Dr. Giovanni Russo, and we’ll be talking about spatial biology.
Hello Giovanni — really a big pleasure to have you here. Thank you for travelling all the way to Amsterdam. Today we’ll be speaking about spatial biology, why it is so exciting to work in this field, and what this field is actually about. I feel like it’s quite new — ten years ago I don’t think people talked about it that much, at least in Europe. Let’s start with an introduction. Tell us who you are.
Dr. Giovanni Russo: Thank you for inviting me. I am Giovanni, but everybody calls me Gio — it’s shorter. I am Italian and I have been living in Germany for more than eight years. I am a former scientist — I did my PhD in neuroscience, working a lot with imaging, and then I moved to industry because I didn’t enjoy academia that much. I wanted to remain part of the world of science but in a different way.
I have worked across many markets: imaging, omics, spatial biology — always in customer-facing roles. I started as a field service engineer, building instruments and training customers, then moved to application specialist, and now sales. What I like about my job is that I still live and breathe science and work with scientists, but without the pressure of writing grants, writing papers, or constantly looking for money. I am also exposed to many different projects, different kinds of science, and new technology arriving in the market. For me, science is not only a job — it is also a passion.
What Is Spatial Biology — And Why Is It Not Actually New?
Dr. Yuliia Shymko: Let’s dig into the world of spatial biology. What is so special about it?
Dr. Giovanni Russo: Spatial biology is something that was always around. We just changed the name multiple times. If you look at how people examine samples, we have always put a piece of tissue under a microscope and tried to see what is inside. What changed is that the technology got better. Instead of visualising one single target, we can now label hundreds of thousands of targets — or an entire genome, or an entire proteome. This means we can extract far more information than before.
This also means the way we handle data is changing — it is becoming more interesting and more complicated. Previously we analysed data by hand. A pathologist would look at a sample and say “this is a cancer cell, this is not.” Now we have software that can do that. Previously microscopes were manual. Now they are automated. Previously you changed buffers and solutions by hand. Now we have liquid handlers.
Spatial biology is the combination of an old technique with many new things coming from automation, software, better detectors, and better probes for labelling structures.
Dr. Yuliia Shymko: You have been on different sides of the industry — as an engineer, application scientist, and in sales. Academia is under enormous pressure to publish, and academics are often the first users of a new instrument or platform. Do you sometimes see resistance from scientists when a new instrument comes to market — perhaps a fear that if everything is automated, what is left for them to do?
Dr. Giovanni Russo: There are several things to consider. First, the people. Some people are very open to innovation. Some are not. In my experience, Europe is very conservative. The US is much more open to early adoption. But once a technology is well established, Europe is actually where you have more opportunity for success. If you have a product still in development, it is easier to find early adopters in the US. But if you want to be really successful long-term, you need to come to Europe — to the big KOLs and key institutions.
If you look at the spatial biology market, the main players are American companies. They nearly all come from California. They started development, co-developed with scientists, and then came to Europe to engage with the key opinion leaders.
Scientists are under enormous pressure to publish, and to publish you need something innovative. But companies are under a different kind of pressure — they need to produce something that delivers novelty while also being stable. If you have a machine that doesn’t work, scientists are not publishing. It is not complicated to build a machine that works for one specific lab and one specific purpose. The challenge for a company is to build something flexible enough to serve multiple purposes while still delivering good data. This is why early adopters are so important. You start with ambitious researchers, and then you need the big players on board to establish yourself in the market.
From a hardware perspective, there is actually not that much left to innovate. The machines that exist are pretty good. The difference comes from chemistry — which probes are you using, how do those probes bind to the structure of interest, and how does that relate to data quality. And then of course there is software.
Why Software, Not Hardware, Is the Real Bottleneck
Dr. Yuliia Shymko: Why software?
Dr. Giovanni Russo: Because we are now generating terabytes of data. Analysing this manually means either recruiting an army of PhD students or spending years to process a single well plate. Software for data analysis is where companies are fighting hardest right now. You need software that produces reliable data efficiently.
It is also interesting to watch the shift happening between wet lab and computational lab. When I was doing my PhD, I was doing my own data analysis — sitting at a chair for hours with a lot of coffee. Now we have software that does this almost automatically with a little supervision. Cell segmentation, spatial correlation analysis — these tools matter because the beauty of spatial biology is that we are adding a layer of information to standard genomics or proteomics. When you sequence something, you know which genes are expressed, but you do not know where those genes are located and how they relate spatially to each other. You do not know whether certain RNA or protein patterns indicate a pathological condition versus a healthy one. This is very difficult to determine manually. This is why software is so critical.
Training Bias and Data Interpretation in Spatial Biology
Dr. Yuliia Shymko: With AI and computing power evolving every day — do we already have the tools to understand this data? Or are we limited by computational capacity, and we need quantum computers before we can get real answers? Or is the real issue less about new platforms and more about focusing on data quality, proper labelling, and a standardised approach that scientists can all agree on?
Dr. Giovanni Russo: Getting two scientists to agree on something — that is very difficult. I partially disagree with the premise. We already have far more powerful tools than before. The new generation of GPUs can process terabytes of data in parallel, which allows us to extract information much faster. The problem is that when you look at the data, you still need to give it an interpretation. That is where the human element plays its role.
Machine learning and artificial intelligence will speed up this process. There are already companies moving in that direction — analysing images and providing hypotheses about tumour grade, metastatic status, whether tissue is infiltrative or not. Is it perfect? No. And it will never be perfect. But we will reach an excellent level of approximation.
What this requires is more initiative from people coming together and sharing data — motivated by the desire for knowledge, not the desire to be the best. If that becomes the guiding principle for both users and companies, then I can see data analysis becoming much faster, interpretation improving, and better diagnostic tools emerging. Some of that is already happening today.
Dr. Yuliia Shymko: What about training bias in data interpretation? When images are labelled by scientists or doctors who make mistakes, those mistakes get propagated through the training data — similar to how gender bias in language propagated through language models because it was embedded in centuries of literature. Is this a real bottleneck in spatial biology?
Dr. Giovanni Russo: The mechanism is real. Bias will always be there. Even the best-trained artificial intelligence or machine learning tool will make mistakes — that is normal. What is possible is to minimise it, and the way to do that is to combine multiple sources of evidence.
If you look at the market, many companies in the genomics and sequencing space are now entering spatial biology, then proteomics, then lipidomics — there is a new omics every day. Multi-omics. That is actually a good thing, because if you combine multiple layers of evidence from the same sample, you get a better interpretation of the data. A monodimensional point of view leads to bias.
The historical example is clear. In the past, a single doctor looked at a glass slide and gave a diagnosis. Then they introduced a second doctor reviewing the same slide, and that massively reduced mistakes. Two sets of eyes are better than one. Now imagine not just two sets of eyes, but multiple layers of evidence combined — spatial, genomic, proteomic. That reduces the chance of error significantly. Mistakes in this field are always made in good faith, but the more data we can combine from the same sample, the better the diagnosis or interpretation becomes.
Why Spatial Biology Is Still Stuck in Academia
Dr. Yuliia Shymko: Companies seem to be racing to build instruments that cover as many dimensions as possible. How far are we from one instrument that can do genomics, proteomics, transcriptomics, and lipidomics?
Dr. Giovanni Russo: I am going to be provocative here: why do we need one instrument?
Yes, it would save cost — one instrument instead of three, less space, less logistics, one training process. But do scientists actually need it? When you buy a coffee machine, you look for the kind of coffee you prefer. I am Italian, so I want something that makes excellent espresso. I do not particularly care if it also makes cappuccino. Scientists look for the technology that best fits their specific requirements.
For a core facility, it is very difficult to justify acquiring one machine that does everything — because it is almost impossible for a single company to build a machine that is genuinely excellent at everything. Can we get a machine that is perfect for spatial transcriptomics and spatial proteomics and spatial lipidomics? Maybe eventually, but not soon. What we can get is a combination of machines.
Look at genomics. We have providers for short-read and long-read sequencing. There are companies that claim they can do both. But when you speak with the users, they say: no, not really — they are better at long reads, or better at short reads. And there is another question worth asking: would you trust a single machine for everything? I know researchers who run the same sample on two different machines and compare the results deliberately. If what I think is there is really there, I should see it in both.
What I think we actually need is not one machine but a defined level of quality that every machine must reach. Then I am very happy to run multiple machines together, analyse the data in different ways, and see if they all converge on the same answer.
Dr. Yuliia Shymko: With new technologies, there is always a gap before they move into clinical use — before they are validated for diagnosis or for predicting drug response. What are the bottlenecks for spatial biology entering the clinical space? And which technology do you think is closest?
Dr. Giovanni Russo: The lack of quality standards is the main thing limiting spatial biology’s entry into pharma and clinical applications.
I have had many conversations with people in pharma and pathology, and they all say the same thing: show me the immunohistochemistry. And you say: why IHC, we can do spatial biology. And they say: yes, but immunohistochemistry has 25 years of history behind it. When spatial biology has 25 years of evidence, maybe you will be right.
Is that a conservative approach? Yes. Is it wrong? No.
Think about a patient waiting for a biopsy result — a cancer diagnosis. Does that patient want the cool technology or the safe technology? They want the safe technology. Does the doctor delivering that diagnosis want to use a new technology with a chance of saying “you are perfectly healthy” when the patient is not? No.
The point is that we need to bring new technology to the same level of validation as the old technology. You first establish that you are as good as the existing standard, and then you show you are better. In my opinion, this is not happening consistently enough yet. Spatial biology is still mostly used in academia. There are some excellent initiatives to translate it into pathology — Denis Shapiro at Heidelberg is doing remarkable work in this direction, and Jean-Philippe Malvoisin is a great advocate for standardisation in clinical spatial biology. But the field as a whole still needs to establish that level of confidence.
This is not just about fulfilling the customer’s desire for safety. It is about a fundamental requirement: when we deliver a diagnosis, that diagnosis must be correct. You want to know if you have cancer or not. At the moment, I do not see that as fully proven or fully tested in spatial biology. It will get there — but it is going to take time.
Imaging-Based vs Sequencing-Based Spatial Platforms: Which Is Better?
Dr. Yuliia Shymko: Which spatial biology platforms do you think are most promising?
Dr. Giovanni Russo: There are two main technology categories: sequencing-based or amplification-based, and imaging-based. Both are excellent — it depends entirely on what you want to do.
For small panels, imaging-based technologies are still the best. Excellent resolution, very low false positive rates. But when the panel gets too large, you encounter optical crowding — the signal quality degrades. So if you are looking at a small, well-defined set of RNA targets or proteins, imaging works brilliantly. But if you want to look at the full genome or full proteome, then amplification or sequencing-based approaches are better — you get the full picture.
This connects back to the single-instrument argument. Can you have one machine that does both well? No. But can you combine them intelligently? Yes. A core facility might run a sequencing-based experiment first to identify the most interesting RNA or protein targets, and then run an imaging-based experiment to go deeper on those specific targets. It is the same principle as starting with bright-field microscopy, moving to confocal, and then moving to super-resolution — each step provides deeper insight.
The Next Five Years: How Will Spatial Biology Change?
Dr. Yuliia Shymko: How do you see spatial biology changing in the next five years?
Dr. Giovanni Russo: I think we will see most small companies becoming part of larger ones, and the technology getting integrated into broader portfolios. I also hope the field will do more work on establishing quality standards — the equivalent of a minimum specification that any spatial biology platform must meet.
I also hope the field moves beyond its near-exclusive focus on cancer biology. Cancer is fascinating, but there is neuroscience, immunology, and the growing field of organoids, which is also being driven by FDA pressure to reduce animal and human experiments.
In terms of technology direction: high-plex platforms will be used for discovery and widespread research. Low-plex platforms will be used for clinical validation. Multi-omics — RNA and protein are already being combined, and I expect DNA and lipids will follow. Metabolomics will be controversial but will come eventually.
And finally, the establishment of quality standards — a defined minimum threshold for what spatial biology platforms must deliver to the customer. That is the key step the field needs to take.
Dr. Yuliia Shymko: Let’s go back to your career. You studied neuroscience, but spatial biology is mainly known for cancer research. And you started as a field service engineer before moving to sales. How was that transition?
Dr. Giovanni Russo: When people ask me what having a PhD means to me, I always say: I was taught how to learn. How to look at information, understand it, and use it. That made it easy for me to move from neuroscience to cancer biology, immunology, oncology, even plant biology — because some things are always the same. Biology is less complicated than people think. When you know the fundamentals, you can grasp more complex levels.
For my first job as a field service engineer — I was coming out of academia and I knew I did not understand business. But I have always liked building and repairing things. If you want to understand something, you need to know how it works. Field service is exactly that. And it gives you an enormous amount of time with customers. A salesperson might spend one hour at dinner with a customer. A field service engineer, depending on the instrument, spends up to two weeks with a customer. Two weeks — you become friends. You understand the machine deeply, and you understand the customer deeply.
Field service is also a great introduction to business. Within that role there is already a bit of sales and a bit of application work. From there you can move to applications, which is more technical and focused on the biological or chemical consequences of what the machine does. And then there is sales.
Why Biotech Sales Fails — And What Companies Get Wrong
Dr. Yuliia Shymko: What do most biotech sales people get wrong?
Dr. Giovanni Russo: I am going to be provocative. I see a lot of people coming out of university and going directly into very senior commercial positions. And then — I say this with sympathy — they fail. Because university does not teach you business. A PhD does not teach you how to deal with people. You may be excellent at dealing with lab animals, but keeping a customer happy is a different skill entirely.
People enter jobs without knowing what they are doing, sometimes thrown straight into the field without proper training. I have met people in product management who do not even know what their product looks like. How can you manage something you do not know?
And too often companies forget one very important thing: customers do not care about your brand or your business pressures. When a customer has a problem, they expect you to help them solve it. I am not saying a salesperson needs to be able to repair an instrument. But a salesperson must be able to point the customer toward a solution and work with the field service engineer and the applications specialist to make that happen.
Behind every machine that does not work or does not deliver, there is a PhD student who cannot finish their thesis, a postdoc who cannot get the data for their paper, and a professor on tenure track who needs that publication. When we send people who are not properly trained to serve those customers, we are setting everyone up to fail. It is why some companies are burning through sales staff faster than cigarettes.
A salesperson should be capable of using the machine, at least at a basic level. You gain credibility in front of the customer. And seriously — how can you sell something you do not understand?
This is why I started from what some people call the lowest rung, which is absolutely wrong. Field service engineers are fundamental to the success of every company. They are the people customers actually remember. The sales team is who customers call to complain. The applications team is who they call for scientific advice. The field service engineer is who they call when something needs fixing. All three are a unit and they need to work together. That collaboration is what builds reputation, and reputation is what generates introductions to new customers, which generates revenue.
Dr. Yuliia Shymko: Where does the real revenue come from in this industry?
Dr. Giovanni Russo: Look at the big established companies and how they actually make money. Most of the time it is not from selling capital equipment. It is from everything that goes around the capital equipment — services, consumables, training. At one point you need to deliver ongoing value, not just push expensive machines that may or may not work.
One company I worked for was making more money from service contracts than from instrument sales. There was more opportunity to offer a discount on the instrument itself than on the service contract or consumables. The level of investment in training the field service engineers and sales engineers was very high — because those people had to be capable and professional every single time, with every customer.
It is very difficult to find a customer who can buy a €400,000 machine every year. At some point that market dries up. Pure capital equipment companies do not last long — they generally get acquired. The smart companies diversify: consumables, services, training, and new instruments. The market is showing this right now. A new company opens every day and a company fails every day. Before starting a company, the main question needs to be: do I have a product someone wants to buy? And even if the answer is yes — what do I do after I have sold 100 machines? There are recent examples of companies that sold 1,000 machines and then got acquired because the market dried up.
Dr. Yuliia Shymko: Last question — what is your favourite thing about spatial biology?
Dr. Giovanni Russo: The first answer is the images. They are genuinely amazing.
But in reality, what I love is that spatial biology brings together the best of multiple technologies. High quality imaging, high quality chemistry, good software — the best of multiple worlds, combined. It allows you to see things that were unthinkable fifteen years ago. Visualising RNA, DNA, and protein simultaneously in the same sample — people would have told you that was impossible. Now we can do it in large panels.
This is also making life better for clinicians and academics. There was recently a beautiful paper on something called STAMP — using spatial biology for a new form of sequencing. That is a measure of how innovative this technology still is, how much value it can still add, and how much further there is to go.
Some technologies are so established that real innovation becomes very difficult. Spatial biology is still young — like a teenager. There is the opportunity to grow and develop into something far more significant. My home office wall has images from spatial biology platforms and providers on it, because they really are extraordinary.
Dr. Yuliia Shymko: Thank you very much, Gio. This was a fantastic conversation.
Dr. Giovanni Russo: Thank you. It was a pleasure.
