Periodic Reporting for period 1 - BMAI (Empowering Radiologists in Cancer Diagnostics with Artificial Intelligence)
Reporting period: 2024-01-01 to 2024-12-31
The increasing incidence of cancer globally, combined with a growing shortage of radiologists, is creating serious challenges in the healthcare sector. Radiologists are burdened with repetitive and time-consuming tasks, such as manually measuring and classifying lesions on CT scans. These tasks, while crucial for cancer diagnosis and treatment planning, are contributing to burnout and misdiagnosis due to the overwhelming workload. With cancer cases expected to rise by 27.4% over the next decade, the pressure on healthcare systems is intensifying, making it harder to deliver timely and accurate diagnoses.
The Better Medicine AI (BMAI) project aims to tackle this challenge by developing a revolutionary AI-powered solution that automates key aspects of cancer diagnosis and tracking. Our tool is designed specifically for radiologists, offering a comprehensive, full-body cancer detection and monitoring system that significantly reduces their workload. Unlike current AI tools that focus on individual organs, our system takes a whole-body approach, allowing radiologists to detect and track metastatic spread across multiple regions of the body.
The core innovation of our solution lies in its ability to automate the data labelling process, which has historically been one of the most time-consuming steps in AI development. By using advanced deep learning techniques, our system can generate detailed, pixel-level pseudo-labels from high-level image annotations, eliminating the need for manual labelling. This breakthrough enables us to rapidly train AI models, improving the speed and accuracy of cancer detection and reducing the overall cost of AI development.
The expected impact of our project is profound. By reducing the time radiologists spend on manual tasks by up to 60%, we help alleviate the shortage of specialists and improve the efficiency of cancer diagnosis. This, in turn, leads to faster, more accurate diagnoses, ultimately saving lives. Additionally, our AI tool's ability to assess tumour dynamics over time allows for more precise monitoring of disease progression, giving clinicians a powerful tool to track treatment effectiveness and adapt patient care accordingly.
In summary, the BMAI project is positioned to have a transformative impact on the field of radiology. By providing a highly efficient, full-body cancer detection system powered by AI, we aim to improve healthcare outcomes and reduce the strain on radiologists, contributing to better patient care in the fight against cancer.
The Better Medicine AI (BMAI) project aims to tackle this challenge by developing a revolutionary AI-powered solution that automates key aspects of cancer diagnosis and tracking. Our tool is designed specifically for radiologists, offering a comprehensive, full-body cancer detection and monitoring system that significantly reduces their workload. Unlike current AI tools that focus on individual organs, our system takes a whole-body approach, allowing radiologists to detect and track metastatic spread across multiple regions of the body.
The core innovation of our solution lies in its ability to automate the data labelling process, which has historically been one of the most time-consuming steps in AI development. By using advanced deep learning techniques, our system can generate detailed, pixel-level pseudo-labels from high-level image annotations, eliminating the need for manual labelling. This breakthrough enables us to rapidly train AI models, improving the speed and accuracy of cancer detection and reducing the overall cost of AI development.
The expected impact of our project is profound. By reducing the time radiologists spend on manual tasks by up to 60%, we help alleviate the shortage of specialists and improve the efficiency of cancer diagnosis. This, in turn, leads to faster, more accurate diagnoses, ultimately saving lives. Additionally, our AI tool's ability to assess tumour dynamics over time allows for more precise monitoring of disease progression, giving clinicians a powerful tool to track treatment effectiveness and adapt patient care accordingly.
In summary, the BMAI project is positioned to have a transformative impact on the field of radiology. By providing a highly efficient, full-body cancer detection system powered by AI, we aim to improve healthcare outcomes and reduce the strain on radiologists, contributing to better patient care in the fight against cancer.
The project has made significant strides in both product development and regulatory approval, with the first AI tool for kidney cancer detection nearing market readiness.
In Product Development, two key areas have seen major progress:
-AI Model Development: We have created a novel approach that uses high-level labels, such as diagnosis codes, to automatically generate detailed, pixel-level information on medical images. This allows us to train AI models more efficiently, producing highly accurate results for detecting and segmenting pathologies without the need for manual labelling, which is typically slow and costly. This innovation accelerates the development of AI tools, enabling us to expand the capability of our models to cover multiple types of cancer.
-Tracking Disease Progression: We developed a module that tracks changes in a patient’s condition by comparing multiple scans over time. This tool automatically measures how a tumour or other pathology is progressing, providing radiologists with detailed reports that can help them assess treatment outcomes or disease progression. This functionality is essential for monitoring long-term cancer patients and optimizing treatment plans based on accurate, automated data.
In Clinical Validation and Market Preparation, our efforts have focused on two critical areas:
-Regulatory Approval: We have completed the necessary steps to meet the requirements of the European Medical Device Regulation (MDR). We are now on the verge of receiving the official certification that will allow our AI tool for kidney cancer to be used in clinical settings. Achieving this regulatory milestone is a significant step toward making our technology available to radiologists and hospitals, ensuring that it meets the highest standards of safety and effectiveness.
-Building Expert Support: We have established strong relationships with Key Opinion Leaders (KOLs) in the field of radiology. These leading experts have endorsed our AI solution and will work with us to promote its adoption in clinical settings. Their involvement is crucial as we move into the commercialization phase, where their expertise and influence will help demonstrate the value of our product to the wider medical community.
In summary, the project has made great progress toward developing a cutting-edge AI tool for kidney cancer detection. With technical validation complete and regulatory approval imminent, we are well on track to bring our solution to the market, where it will help radiologists diagnose and monitor cancer more efficiently and accurately.
In Product Development, two key areas have seen major progress:
-AI Model Development: We have created a novel approach that uses high-level labels, such as diagnosis codes, to automatically generate detailed, pixel-level information on medical images. This allows us to train AI models more efficiently, producing highly accurate results for detecting and segmenting pathologies without the need for manual labelling, which is typically slow and costly. This innovation accelerates the development of AI tools, enabling us to expand the capability of our models to cover multiple types of cancer.
-Tracking Disease Progression: We developed a module that tracks changes in a patient’s condition by comparing multiple scans over time. This tool automatically measures how a tumour or other pathology is progressing, providing radiologists with detailed reports that can help them assess treatment outcomes or disease progression. This functionality is essential for monitoring long-term cancer patients and optimizing treatment plans based on accurate, automated data.
In Clinical Validation and Market Preparation, our efforts have focused on two critical areas:
-Regulatory Approval: We have completed the necessary steps to meet the requirements of the European Medical Device Regulation (MDR). We are now on the verge of receiving the official certification that will allow our AI tool for kidney cancer to be used in clinical settings. Achieving this regulatory milestone is a significant step toward making our technology available to radiologists and hospitals, ensuring that it meets the highest standards of safety and effectiveness.
-Building Expert Support: We have established strong relationships with Key Opinion Leaders (KOLs) in the field of radiology. These leading experts have endorsed our AI solution and will work with us to promote its adoption in clinical settings. Their involvement is crucial as we move into the commercialization phase, where their expertise and influence will help demonstrate the value of our product to the wider medical community.
In summary, the project has made great progress toward developing a cutting-edge AI tool for kidney cancer detection. With technical validation complete and regulatory approval imminent, we are well on track to bring our solution to the market, where it will help radiologists diagnose and monitor cancer more efficiently and accurately.