Breast Cancer Awareness Month is mostly focused on making sure women recognize the signs and go to their annual mammography check. I think this is not enough and, once again, putting all the responsibility on women. How about making all the stakeholders aware of what they should know about breast cancer? The good practices, the latest available tools, the possible improvements of the screening programs.
The history of women’s imaging shows that it is not given the highest priority, despite mammography screening having proven its massive impact on survival. For instance, mammography was the last radiology modality to be digitized.
On some occasions though, industry leaders not only bring new technology, but also make sure that once proven efficient, both medical practitioners and patients know about it and it is being used rapidly and widely. Tomosynthesis, or 3D mammography, is one good example of that.
Breast cancer screening programs involve many stakeholders, actually an entire ecosystem, and this makes it complex to improve. Double reading of mammograms has long proven its efficiency everywhere outside the US. The detection rate is just above 5 for 1,000 screen exams in the US when it is above 6 in Europe which represents 20% more cancers being detected for the same number of exams!
Over the last 4 years, the world of medical imaging has started a revolution: the adoption of Artificial Intelligence (AI). Few highly specialized companies have developed and validated algorithms. The decision-support softwares using these algorithms, when used in combination with the radiologists, have proven to improve the cancer detection rate and thus been clear for use by the authorities (FDA). This is the opportunity to make it up for women on the various levels mentioned above.
For once, a technological innovation, AI, has quickly translated into products that benefit specifically women: more cancers found, earlier, with quicker results.
The industry providing these tools and the prime users of the tool, the radiologists, have worked together so that the latter do not fear being replaced, and understand that the soon-extinct species will be radiologists not using AI. This technology gives them the reassurance and confidence they don’t overlook something suspicious, they can spend more time on suspicious cases and less on a non-suspicious one, so overall decrease their risk of burnout.
Besides the performance of the algorithm, the AI revolution brings more efficient access to technology. This is pure software that doesn’t require specific or expensive infrastructure and hardware since it is hosted in a secure cloud environment. The installation and training have been designed to be the least disruptive of the radiologists’ workflow, and can be performed remotely so almost no local IT resource is needed. Last but not least, the Software-as-a-Service (SaaS) business model is the most commonly offered way to gain access to this technology, which means no upfront investment, and fair pricing, all-inclusive and based only on usage.
So, given the need (to reduce the number of missed cancers, to find cancer earlier), the additional benefits (less anxiety for women as results can be delivered quicker, and less stress for the radiologist who fears missing something) and the ease of adoption and implementation, why is it still so rare that women are even offered to be screened with the help of AI?
I posit that the awareness is here but the priorities in decision-making are not right.
Radiologists are rapidly convinced that they should evaluate these tools in their practice. But, it is not just the radiologist that needs to be on board with utilizing this new, useful, and successful software. The entire organization must see the value and understand that AI can make a difference in lives, and from a business perspective as well. The evaluation is made simple and easy with the cloud. This could be seen as ‘yet another project’ by the IT department and just added to the queue whereas it is indeed a rare opportunity to directly make a difference in women’s life. Additionally, the financial aspects are very straightforward, with multiple positive impacts on the various lines of the P&L, ending always with a positive business case. Finally, there are many decisions to be taken that do have a complex decision pathway, if it is at all defined, before the AI can be available to radiologists. Every effort has been made to simplify these various steps, so it is the sort of project that should be given priority because of its low impact on the organization and a big impact on all stakeholders, women being paramount.
During this month of awareness, I am calling on all stakeholders in breast cancer detection to take 10 minutes:
- To speak to colleagues about how AI could be offered to their patients
- To get a demo of an AI tool and see what it can do
- To speak to me and ask any questions you might have on that topic!
Matthieu Leclerc-Chalvet, CEO