back

AI instead of chemistry

Start-up ViSPAGI turns cancer diagnostics on its head

2024/10/28 by

Cancer is one of the most frightening diagnoses for many people. According to statistics from the Robert Koch Institute (RKI), around half of the German population will develop the disease in the course of their lives. There are around 500,000 new cases in Germany every year. But there is hope: more than half of those affected survive, thanks in part to improved diagnostics. Young etit scientists are revolutionising cancer diagnosis with their start-up ViSPAGI.

Dr Özdemir Cetin and Ahmed Elshamanhory are co-founders of the biotechnology start-up ViSPAGI

The start-up ViSPAGI, founded by etit junior scientists Dr Özdemir Cetin and Ahmed Elshamanhory from the Self-Organising Systems Lab, is revolutionising cancer diagnosis with AI technology. ViSPAGI enables precise diagnoses within minutes instead of weeks. This dramatically reduces psychologically stressful waiting times for results and enables earlier and more effective treatments. Supported by TU Darmstadt and Frankfurt University Hospital, the company has demonstrated its potential in a pilot study and is about to be launched on the market. The aim: faster, more precise results and better chances of recovery for patients. In this interview, the two founders talk about their project.

Özdemir Cetin: If doctors discover abnormal tissue changes during examinations – whether by palpation or imaging procedures – they take a biopsy. These samples must be analysed to determine whether they are cancerous. Traditionally, this is a time-consuming procedure, as the tissue samples have to be chemically stained several times in order to recognise the cancer cells under the microscope. The colour contrast allows pathologists to better distinguish the different structures and cell types in the tissue.

This procedure is time-consuming and very stressful for patients. ViSPAGI, on the other hand, uses artificial intelligence to generate the contrast digitally. Our technology uses virtual colouring to highlight the necessary structures. This eliminates the need for physical contrast agents or additional chemical processes. The result is a digital image that mimics the visual effect of different staining methods without the use of physical reagents. ViSPAGI can apply different virtual stains sequentially to the same digital tissue image, allowing pathologists to obtain multiple diagnostic information from one image. This enables a reliable diagnosis within minutes instead of weeks. And it saves enormous laboratory costs.

First (pink) image: H&E staining in the laboratory, second (blue) image: CD8 staining in the laboratory, last image: CD8 again, but generated with ViSPAGI.
First (pink) image: H&E staining in the laboratory, second (blue) image: CD8 staining in the laboratory, last image: CD8 again, but generated with ViSPAGI.

Ahmed Elshamanhory: Unlike other AI-based systems that are trained to recognise cancer or tumours as such from images, virtual staining as we use it is not a diagnostic tool in itself. ViSPAGI helps pathologists interpret tissue images more efficiently and reach a diagnosis faster by reducing the need to perform many physical stains in the lab. In certain cases, however, traditional staining methods may still play a role. The final diagnosis is therefore made by experts who carefully analyse these virtually stained images.

Özedemir Cetin: The AI mainly works with images – i.e. whole slide images (WSIs), which are digital scans of tissue samples. We receive these images from the pathology departments of the respective hospitals. They are used to assess the morphology of the tissue, and the AI algorithm applies virtual staining to create contrasts similar to those achieved by traditional chemical staining. The diagnosis is therefore based solely on the visual information of these scanned images and no additional reports or data input is required.

Ahmed Elshamanhory: Not at all. We only receive the images of anonymised tissue samples. We know no names, no medical reports, no findings. We don't need that for our technology. We are one hundred per cent GDPR-compliant and guarantee the highest standards in terms of data protection and data security.

Özdemir Cetin: Doctors often find it difficult to trust AI because they cannot understand how the results are achieved by the technology. As AI acts like a ‘black box’, there is a lack of transparency, which makes trust difficult.

We are working on making our software more transparent so that doctors can better understand how the results are achieved by the AI. We are also taking into account the uncertainties in the model to ensure that the results are comprehensible and trustworthy.

We are currently purchasing tissue images to train our AI. When we are ready to go to market, we will make ViSPAGI available to doctors as software. No images or other data will leave the clinic. On their own conventional laptops – even in the operating theatre – the doctors enter the tissue images they have taken themselves and have them stained virtually. This provides transparency and security in the procedure.

Özdemir Cetin: A friend of mine works in the pathology department at Frankfurt University Hospital and told me about the complex diagnostic procedures for cancer. I know from family members who have cancer that it is psychological hell to wait weeks for a diagnosis.

I am an electrical engineer and, as an AI enthusiast, I am convinced that things can be better. So I set about developing a corresponding technology. I showed the first result to a pathologist friend of mine. I can tell you that he was blown away and convinced that our technology will herald a paradigm shift in cancer diagnostics.

Ahmed Elshamanhory: ViSPAGI is a real game changer with huge application potential worldwide. That's why we have now founded a start-up. And we believe it will be very successful because we are not just throwing a product on the market and then waiting to see if anyone needs it. We are offering a revolutionary solution that addresses a major problem in pathology by changing workflows and significantly reducing time in the laboratory. Nevertheless, implementation requires training and adaptation to existing clinical structures.

Özdemir Cetin: With the help of the staff at the HIGHEST innovation and start-up centre, we got in touch with hessian.AI and other players in the start-up ecosystem, such as the ryon accelerator and Kai Gehring, the member of the Bundestag responsible for innovation. Together, we received around one million euros from the federal government's Distr@l funding programme, which strengthens digitalisation and research transfer, Hessian.AI's LAISIF programme and SOS. This enabled us to finalise a prototype, finance a feasibility study and hire additional employees.

Ahmed Elshamanhory: We receive coaching as part of the funding programmes, take part in training sessions and develop a business model for ViSPAGI. We see ourselves as founders who are able to solve problems quickly and efficiently from different perspectives. We have studied and researched here in Germany, understand the needs of the market and have a clear vision of how to improve cancer diagnosis. We are on the verge of bringing a breakthrough technology to the market with the aim of creating a future where cancer diagnosis is smarter, safer and more accurate, worldwide.

Always up-to-date with exciting news from the etit campus: Follow us on Instagram!

Recommended external content

We have selected external content from Instagram for you and would like to show it to you right here. To do this, you must reveal it with one click. You can hide the external content at any time with another click.

I agree to external content from X being shown to me. This may result in personal data being transmitted to third-party platforms. You can find more information in our Privacy Policy.