Periodic Reporting for period 1 - AIOIA (AI Organoid Image Analysis)
Période du rapport: 2023-06-01 au 2024-11-30
Organoids are microscopically small patient-derived 3D organs that can be cultivated in the laboratory over months mimicking human organs and their functions in vitro. These mini-organs have a human genetic background, maintain disease traits in vitro and are currently available for almost every human organ. Due to these advantages, organoid research and its commercial applications are rapidly evolving and increasingly used. Given the high variability of these complex 3D structures, classical cell culture-based image quantification tools techniques do not accurately capture organoids in microscope images. This has resulted in much image quantification at our laboratory - like in many others - being performed using manual time-consuming tools with a high researcher variability. In sum, there is a global challenge to in a standardized manner quantify organoid experiments. The overall objective of this project was to develop the first Software as a Service (SaaS) toolbox explicitly tailored to organoid imaging, building upon powerful AI-based algorithms for cutting-edge image quantification and independent of the underlying microscope and culture system hardware. The SaaS will provide a user-friendly approach to upload brightfield and immunofluorescence microscope images and videos of organoid cultures, which can then automatically be quantified by AI-models generating a range of organoid metrics. Overall, the provided automation envisioned would drastically reduces analysis time, promotes accurate phenotypical organoid analyses and ensures standardization of results between different researchers, culture conditions, and imaging hardware.
To facilitate organoid research we build on AI-based tools for object detection in images that are increasingly used. Our initial tool at the start of POC was able to analyze organoids in bright field images that can be used for growth analyses of organoids. These initial applications resulted in time reduction and facilitated the upscaling of organoid analyses and studies in science and commercial applications. Building on these features and our network we planned to further develop the AI model to analyse more complex images and further develop the tool into SaaS application widely available for scientific and commercial purposes. In pursuit the following Workpackages were performed:
Work Package 1: Customer discovery and needs assessment: In WP 1 we have expanded the visibility of GOAT and made an assessment of the needs of users of organoid systems.
Work Package 2: Technology adaptation and validation: Building on the recommendation of users of organoids in WP2, we further tailored the tool to the needs of the organoid users.
Work Package 3: Optimization of SaaS platform: WP3 aimed at improving access to the SaaS platform. This includes assessment of pricing of usage of the SaaS platform as well as user-friendliness of the SaaS platform.
Work Package 4: Upscaling business plan: With support of the ERC POC, we have been able to recruit experts in business and marketing that provided support for pitching the project to investors. Furthermore, we have pitched the commercialization of the tool to investors.
At the end of the project, it was decided that the tool developed will not be commercialized, but potentially developed as an open access tool.
Work Package 1: Customer discovery and needs assessment: In WP 1 we have expanded the visibility of GOAT and made an assessment of the needs of users of organoid systems.
Work Package 2: Technology adaptation and validation: Building on the recommendation of users of organoids in WP2, we further tailored the tool to the needs of the organoid users.
Work Package 3: Optimization of SaaS platform: WP3 aimed at improving access to the SaaS platform. This includes assessment of pricing of usage of the SaaS platform as well as user-friendliness of the SaaS platform.
Work Package 4: Upscaling business plan: With support of the ERC POC, we have been able to recruit experts in business and marketing that provided support for pitching the project to investors. Furthermore, we have pitched the commercialization of the tool to investors.
At the end of the project, it was decided that the tool developed will not be commercialized, but potentially developed as an open access tool.
The tool developed will not be commercialized, but potentially developed as an open access tool