Investing in next generation medical imaging will revolutionise tumour surgery
Usually, tumour surgery is not limited to a single procedure. Removing malignant tissue and at the same time retaining as much healthy tissue as possible – e.g. when treating breast cancer – is not easy. Furthermore, the level of success is not determined until afterwards in a pathology lab, where the periphery of the removed tissue is analysed and assessed.
‘This regularly reveals that an additional operation is required’, says Shaffer. ‘These operations are stressful to the patient and very expensive. Our Histolog Scanner offers a solution. We bring ultrafast digital confocal microscopy to the operating theatre. After dipping fresh tissue in a contrast medium, it can be put on the Histolog Scanner; there is no need to cut it or fix it to glass. Within one minute it is ready for imaging. Within the following minute, you have a detailed picture of the morphology with a subcellular resolution. This allows the surgeon to determine margins much more accurately during surgery, to work more precisely and avoid a follow-up operation. But our technology can also be applied more broadly, e.g. for biopsies or as part of the pathology workflow.’
The Histolog Scanner is a breakthrough in medical imaging by itself. It can scan and process tissue samples with a diameter of up to 8 cm almost in real time. This large field of view – typical tissue sections are 1.5 cm – offers considerable advantages. But it also requires a great deal in terms of data management. According to Shaffer, the second major challenge is the development of user applications. ‘The accuracy with which specialists interpret images depends on many factors, like their training level and experience. We want to help clients eliminate this subjectivity and speed up the analysis process. The first step here is to automate the initial rough analysis. We recently released an application for our Histolog Scanner, which independently identifies areas of interest for further inspection. We created this application together with the software and mathware specialists of Sioux.’
‘That tool is cutting-edge technology at the interface between data analysis and artificial intelligence’, underlines Robbert van Herpen, responsible for the Mathware division at Sioux Technologies. ‘The system needs to quickly, accurately and reliably recognize various patterns in a variety of tissue types. That forms the basis for the data model. During its development, we applied the latest insights in the field of deep learning. By feeding it the right data, optimizing the hyperparameters and deploying it in the SamanTree system, we made the leap towards a high-tech solution. Right now, the tool is very good at pinpointing areas of interest and is ready for use in operating theatres. The holy grail would be an autonomous system that is 100 percent correct and excludes human error. We are not there yet, but the current scanner and applications are already facilitating a revolution in tumour surgery.’
SamanTree Medical operates a fleet of eight Histolog Scanners for demonstration projects in various European medical centres in preparation of the commercial roll-out. Sioux also plays a key role in this. Shaffer: ‘When looking for a development and manufacturing company, we came across Sioux 3 years ago. Now in 2020, we are connected in many different ways. The Sioux Tech Fund joined up as an investor. Apart from being an extension for our R&D, Sioux also is our exclusive partner in industrialization and manufacturing. Production has been fully transferred; this helps us to focus on our customers. That market introduction will be a step-by-step process due to the revolutionary nature of our technology. But the masses will follow the early adopters, of that I am certain. Sioux’s engineering and manufacturing skills will also help us then in terms of further development, upscaling of produc-tion and cost-down operations.’