How Roche Uses ML to Boost Their Digital Pathology Capabilities

Pharma company leverages ML to augment and support pathologists in assessing the toxicity of drugs better and faster than ever, increasing the transparency of the quality assurance process and elevating patient care.

The Client
Personalized Healthcare

Main Challenge

Preclinical studies generate high volumes of pathological data, making pathological analysis time-consuming. To comply with rigorous drug safety standards, Roche analyzes thousands of whole slide images as part of the toxicity assessment process.

In each study, pathologists must find and score subtle lesions on approximately 1600 slides, each measuring 200 million pixels (i.e. 23 meters by 23 meters). However, approx 70% of the slides do not contain any lesions. So most of the time pathologists are looking at normal slides.

How we helped

We developed a Machine Learning model that identifies healthy slides and discards them automatically using the latest computer vision techniques and digital pathology data. The model provides heatmaps on tissue slides to highlight lesion-prone areas for expert pathologists, making their analysis more accurate and efficient.

  • Safely discards true negative slides (containing no lesion)

  • Assists pathologists in prioritizing the focus areas for the remaining slides

  • Reduces the time needed to find lesions

The Impact

With a more efficient way to analyze slides, the company can reduce by 50% the number of slides to review. It also lowers the risk of submitting false-positive reports that can significantly delay the drug development process and reduce future drug competitiveness. The newly acquired AI capabilities can potentially speed up clinical trials and shorten time-to-market. Finally, it allows the Roche to better leverage its data while developing its technical capabilities and understanding of data science solutions. 


Roche can reduce in half the number of slides to review, leading to more efficient workflows and time savings in the analysis process.


The strategic use of data is set to have long-term benefits, enhancing understanding of data science solutions and fueling innovative drug development. 


The deployment of the AI model can fast-track the clinical trial phase, reducing the overall drug development timeline and increasing speed-to-market.

Check this video for more information about the project.