Periodic Reporting for period 2 - AIPACA (AI-powered pathological cancer diagnostics)
Okres sprawozdawczy: 2021-08-01 do 2022-07-31
Cancer diagnostics today is a time-consuming process which is still performed manually by pathologists. Mindpeak’s products democratize cancer diagnostics, making it accessible to everyone globally. This helps solving the increasing demand-supply gap in cancer diagnosis. We develop solutions based on advanced artificial intelligence (AI) to achieve this. Our AI analyses histopathological tissue slides within seconds. It finds tumour regions and quantifies biomarkers. Pathologists stay in full control and can accept, correct, or even reject suggestions made by our AI. Patients benefit from increased accuracy and a faster turnaround time as well as reduced costs, while pathologists are relieved from tedious manual work. The key USP of our innovation is its reliability in the clinical context: our AI is robust to the significant variations within and across labs. It can easily be integrated in all clinical environments and works with scanner hardware of all major providers. Through feedback from the pathologist, our AI learns over time and continuously improves its accuracy. In contrast to human analysis, its results are not influenced by external factors and always reproducible.
Our innovative AI solutions lie at the heart of Mindpeak. We are a start-up that builds support tools for cancer diagnostics. The fundamental technology of these tools is our AI. Our first product detects and quantifies tumour cells and hit the market in October 2019. It is the first commercially available product for tumour quantification using AI. Up to now Mindpeak is providing several AI software products for cancer diagnosis and other diseases. As a result, our experience and knowledge base has grown tremendously and will allow us to further develop relevant and innovative AI software to help physicians make treatment decisions and to improve the quality of care. In addition, our first commercial steps with international partners have successfully taught us effective go-to-market strategies and how to involve customers in product development so that in the end both benefit from a successful cooperation. Gaining user acceptance is key to disruptive technologies like our innovative AI. Understanding how to convince and work together with pathologists to use our products puts us in a unique position to become a global AI analysis provider for pathology with high penetration in the international market of digital pathology. So, on the one hand, we have realized high-quality AI software in AIPACA on the technological side, and on the other hand, the commercial side, several resulting products have been clinically CE-validated and brought to market through cooperation with pathologists in AIPACA. As a result, Mindpeak has significantly increased its competitiveness of AI software developed at AIPACA. Due to the current lack of AI applications in diagnostics that are as high quality as Mindpeak's products, we see an opportunity to become a leading international provider of highly accurate and robust AI software for clinical use.
Our innovative AI solutions lie at the heart of Mindpeak. We are a start-up that builds support tools for cancer diagnostics. The fundamental technology of these tools is our AI. Our first product detects and quantifies tumour cells and hit the market in October 2019. It is the first commercially available product for tumour quantification using AI. Up to now Mindpeak is providing several AI software products for cancer diagnosis and other diseases. As a result, our experience and knowledge base has grown tremendously and will allow us to further develop relevant and innovative AI software to help physicians make treatment decisions and to improve the quality of care. In addition, our first commercial steps with international partners have successfully taught us effective go-to-market strategies and how to involve customers in product development so that in the end both benefit from a successful cooperation. Gaining user acceptance is key to disruptive technologies like our innovative AI. Understanding how to convince and work together with pathologists to use our products puts us in a unique position to become a global AI analysis provider for pathology with high penetration in the international market of digital pathology. So, on the one hand, we have realized high-quality AI software in AIPACA on the technological side, and on the other hand, the commercial side, several resulting products have been clinically CE-validated and brought to market through cooperation with pathologists in AIPACA. As a result, Mindpeak has significantly increased its competitiveness of AI software developed at AIPACA. Due to the current lack of AI applications in diagnostics that are as high quality as Mindpeak's products, we see an opportunity to become a leading international provider of highly accurate and robust AI software for clinical use.
In AIPACA, we have developed AI software for diagnostic assessment of cancer progression and characteristics based on immunohistochemistry (IHC). Our solutions are based on current medical guidelines and have been CE validated for use in clinical routine settings. Diagnostic accuracy with support of our AI software for the biomarkers Ki67, ER, PR and HER2 in breast tissue is on par with human pathologists. We have developed fully automated AI architectures and workflows to enable instant visualization of all detected and classified cells. Mindpeak's user interface, AI Console, has been fully developed within AIPACA in cooperation with scientists and pathologists from all over the world. AI Console is a stand-alone software used for interactive visual examination of digital images of histopathological slides in combination with supplementary information. It allows the user to interactively change image display magnification, pan and rotate within the histopathological image and draw and save regions of interest for diagnosis and for traceability and display clinical scores. AI Console has undergone CE-validation and has passed successfully usability testing studies with experienced and renowned pathologists to prove its user-friendliness and its usage advantages. Exploitation: All of our in AIPACA developed and clinically validated products are already on the market and our solutions are integrated into the daily routine workflows of our customers. In addition, partner labs from Germany, the US, Norway, Poland, UK and other countries are continuously testing our AI software and AI Console in terms of usability, accuracy and robustness and workflow integration. Dissemination: We frequently demonstrate our solutions at pathology conferences and exhibitions in Europe as well as in the US. All our AI software products developed and validated within AIPACA are presented on our website and disseminated via LinkedIn posts and regular press releases to the scientific and digital pathology community. We work hand in hand with partner vendors of our AI software.
In AIPACA, Mindpeak has developed and brought to market several AI software products, all of them being pathological AI-systems for image-based detection and quantification of breast cancer cells to overcome the disadvantages of the current manual process. Image recognition in pathology is still extremely demanding and can only be solved with the latest methods of artificial intelligence and deep learning. Mindpeak has significantly increased its competitiveness by commercialising AI software products. Due to the current lack of AI applications in diagnostics that work on the high quality accuracy level of Mindpeak’s products, we see the opportunity to become an international leader in the provision of highly accurate and robust AI software for clinical use. In the clinical studies for CE-validation of our AI software products, we were able to show that cancer assessment with the support of Mindpeak's software is on par with human pathologists agreement (interobserver variance).
Mindpeak has received CE-IVD Mark for Ki67, ER, PR and HER2 detection and quantification in breast tissue, which makes the software available to clinical pathology practices across Europe and helps to overcome the challenge of an increasing demand for cancer diagnostics. Mindpeak is the first company to receive approval for a deep learning solution which can distinguish between tumorous and non-tumorous structures on a single-cell level.
The impact of Mindpeak’s algorithms developed in AIPACA is multifaceted. Cancer patients will receive more proper and effective therapies due to more accurate diagnostic results and due to the fact that AI assists medical professionals with tedious and repetitive standard tasks, so that pathologists can focus on the really important cases. AI technology delivers independent and unbiased results always based on the same analysis. Both the reduction in the workload of healthcare professionals and the improved care and medical treatment of cancer patients can provide lasting relief to the ailing healthcare system.
Mindpeak has received CE-IVD Mark for Ki67, ER, PR and HER2 detection and quantification in breast tissue, which makes the software available to clinical pathology practices across Europe and helps to overcome the challenge of an increasing demand for cancer diagnostics. Mindpeak is the first company to receive approval for a deep learning solution which can distinguish between tumorous and non-tumorous structures on a single-cell level.
The impact of Mindpeak’s algorithms developed in AIPACA is multifaceted. Cancer patients will receive more proper and effective therapies due to more accurate diagnostic results and due to the fact that AI assists medical professionals with tedious and repetitive standard tasks, so that pathologists can focus on the really important cases. AI technology delivers independent and unbiased results always based on the same analysis. Both the reduction in the workload of healthcare professionals and the improved care and medical treatment of cancer patients can provide lasting relief to the ailing healthcare system.