Doctor Isabelle Birtwisle Peyrottes
I enjoy being part of CERBAPATH, where our commitment to innovation and digitalization in pathology is unwavering. One of the challenges of the laboratory is currently integrating AI into the clinical workflow.
The digital transition at CERBAPATH began in 2022, and since then, we have made significant strides in automating our processes. From barcode systems to automated machines that engrave blocks and slides, we have streamlined our operations to handle high volumes of data and samples efficiently. Despite these advancements, there are always a few cases that require manual intervention, but these are minimal, perhaps five per day.
One of the tools we have been exploring is Owkin’s RlapsRiskⓇ BC. Intellectually, it is very appealing. The usability of RlapsRiskⓇ BC is straightforward; it involves entering a few data points into the application, which is not complex at all. What sets RlapsRiskⓇ BC apart is its ability to provide a visual control over what is being examined, unlike molecular biology techniques where everything is mixed in a test tube.
The integration of AI tools like RlapsRiskⓇ BC into our workflow has been relatively seamless. It fits well within our existing processes, allowing us to quickly adapt and utilize the tool effectively. However, there are challenges, such as the readability of the results. But with magnification, we can see the areas the algorithm has focused on, which is crucial for understanding the risk assessment.
The security of patient data is paramount to us. We are merely custodians of the material, ensuring its quality and conservation. There are strict regulations surrounding the handling and transportation of patient material, and we adhere to these guidelines rigorously.
Our laboratory faces daily challenges, including the management of high volumes of information and the need for precise and timely diagnoses. The digitalization of our processes has significantly improved our efficiency, but there is still room for further development. The objective of AI tools is to free up our time to focus on more complex cases.
The evolution towards hyper-specialization in medicine, driven by digitalization, has also facilitated remote collaboration among pathologists. This has been particularly beneficial in sharing expertise and improving the accuracy of diagnoses.
In conclusion, the integration of AI and digital tools in pathology is not just a trend but a necessity. It allows us to manage high volumes of data efficiently, improve the accuracy of diagnoses, and ultimately enhance patient care. I am excited about the future of pathology and the role that digitalization and AI will play in shaping it.
What I love about anatomical pathology is that it is the medical specialty that, for me, goes closest to the “truth"
Credit for photography and interview: Florent Tribalat