A groundbreaking AI tool developed by Kheiron Medical Technologies and Imperial College London has shown promising results in detecting breast cancer at an early stage. The tool, called Mia, outperformed human “readers” by identifying up to 13% more breast cancers during screenings. The study, titled “Prospective implementation of AI-assisted screen reading to improve early detection of breast cancer” and published in Nature Medicine, highlighted the potential of Mia to significantly improve breast cancer detection rates.
Breast cancer remains a significant global health issue, with 2.3 million women diagnosed with the disease in 2020. Screening programs play a crucial role in detecting cancer early and reducing the need for aggressive treatments. However, approximately 20% of breast cancers are estimated to be missed during initial screenings. This is where the use of artificial intelligence can make a significant impact.
Mia, a certified Class IIa AI tool, was utilized in a study conducted in Hungary. Over a period of two years, mammograms from 25,065 women were examined by human radiologists, and then Mia was used as an additional reader. Mia’s role was to identify subtle signs of cancerous tissue that might have been missed by humans. Any potential false negatives detected by Mia were then reviewed by a third human reader, who determined whether the woman should be recalled for further investigation.
The study’s findings revealed that Mia detected an additional 24 cancers, representing a relative increase of 7% compared to human readings alone. Furthermore, Mia resulted in 70 more women being recalled for further examination, signifying a relative increase of 0.28%. Importantly, 83% of the additional cancers detected by Mia were invasive, highlighting the tool’s ability to identify cancers requiring urgent attention.
Dr. Peter Kecskemethy, CEO of Kheiron, emphasized the significance of Mia’s impact on breast cancer detection and urged the adoption of the AI tool in clinical practice. Dr. Ben Glocker, a study co-author and machine learning researcher, praised the results, which surpassed expectations and demonstrated the potential of AI to reduce missed cancer cases.
While the study focused on a specific setting in Hungary, the researchers acknowledge the importance of replicating the results in other countries to account for variations in screening programs and populations. Additionally, longer follow-up studies are required to fully assess Mia’s impact on cancer detection and mortality rates.
With an ongoing study in the UK nearing completion and a planned roll-out in the US, the potential of AI-assisted breast cancer screening appears promising. Mia’s ability to augment human readings and enhance early cancer detection could revolutionize breast cancer care, providing better outcomes for patients worldwide.
FAQ
1. What is Mia?
Mia is an AI tool developed by Kheiron Medical Technologies and Imperial College London. It is designed to assist in the early detection of breast cancer by identifying subtle signs of cancerous tissue in mammograms.
2. How does Mia improve breast cancer detection?
Mia works alongside human radiologists during breast cancer screenings. It analyzes mammograms and flags potential false negatives that may have been missed by human readers. These flagged cases are then reviewed by a third human reader, allowing for a more comprehensive assessment and increased early detection rates.
3. How effective is Mia compared to human readers?
The study found that Mia detected up to 13% more breast cancers than human readers alone. This demonstrated Mia’s potential to significantly improve early breast cancer detection rates during screenings.
4. What are the implications of Mia’s findings?
The results of the study highlight the importance of integrating AI tools like Mia into breast cancer screening programs. By reducing the number of missed cancers and facilitating early detection, Mia has the potential to improve patient outcomes and reduce cancer-related mortality.
5. Is Mia currently being used in clinical practice?
While more research and validation are needed, Mia is showing promising results. Ongoing studies in different countries, including the UK and the US, aim to confirm the benefits of Mia’s AI-assisted reading workflow. If successful, Mia could become a vital tool in breast cancer screening worldwide.