ATOMWISE
February 1, 2023
2023: The Year of The Artificial Intelligence
February 1, 2023PAIGE
Overview
Driven by a team of pathology, machine learning (ML), healthcare, and industry experts, Paige aims to help pathologists conduct their diagnostic work effectively and efficiently. They supply additional knowledge from digital slides and go beyond this conventional work to obtain new diagnostic information from pathology slides and patient samples that can support patients’ informed treatment pathways. Paige technology also allows organizations to analyze and design customized patient therapies.
Latest in Artificial Intelligence
Paige is powered by artificial intelligence. They use ML and deep learning algorithms on petabyte-scale datasets to discover new insights in pathology images and transform the pathologists’ workflow.
Computational Diagnostics Delivered via a Modern Imaging Platform
Paige is committed to the production of products that help pathologists make diagnoses that are knowledgeable and accurate. They develop digital diagnostics based on AI delivered via an interoperable enterprise imaging platform designed to drive diagnostic trust and a more efficient workflow in pathology. Life science companies are also motivated by Paige technologies to assess patient care options and create new biomarkers so that each patient gets specific treatment options.
Advancements in Computational Pathology
The technology used is the latest in computational pathology developed by the best minds in ML and pathology and driven by proprietary data from more than a decade of research at one of the most trusted cancer care centers in the world.
Trusted by Leading Organizations in Medicine and Science
They harbor strategic collaborations and alliances, including Invicro and Philips, with leading organizations assisting pathologists.
Validated by Regulatory Bodies and Independent Studies
They provide FDA-cleared products, CE-marked, and validated by independent research that understands computational pathology’s strength and potential.(1)
IT Overview
Paige’s ML experts have a decade of experience building large-scale ML systems for Computational Pathology. The system is powered by dedicated high-performance compute clusters with hundreds of GPUs. Moreover, the company counts with a unique, scanner-neutral slide viewer allowing real-time access to Paige’s capabilities.
Paige brings the latest computational pathology technology developed with proprietary data for custom development programs so their partners can:
- Fast-track the biomarkers discovery process to support pipeline programs.
- Discover novel biomarkers.
- Create consistent digital analogs to existing biomarker assays.
- Characterize tissue to inform clinical trial design.
Paige’s computational pathology products help to allow standardized detection and quantitation of tumors pre or post-treatment to inform correct treatment and other trial options to each patient at the point of care. Furthermore, it facilitates remote global collaboration between Paige’s life sciences partners, contract research organizations, and clinical trial sites.
FullFocus viewer is FDA-cleared and CE-marked. It was designed to deploy new algorithms at scale, which allows for the rapid distribution of novel biomarkers to expedite enrollment and assist in identifying treatment pathways and trial options. It also includes anywhere access for pathologists and research teams and data management, a storage solution that is secure, scalable, and cost-effective.
Paige’s ML experts have a decade of experience building large-scale ML systems for Computational Pathology. The system is powered by dedicated high-performance compute clusters with hundreds of GPUs. Moreover, the company counts with a unique, scanner-neutral slide viewer allowing real-time access to Paige’s capabilities.
Paige brings the latest computational pathology technology developed with proprietary data for custom development programs so their partners can:
- Fast-track the biomarkers discovery process to support pipeline programs.
- Discover novel biomarkers.
- Create consistent digital analogs to existing biomarker assays.
- Characterize tissue to inform clinical trial design.
Paige’s computational pathology products help to allow standardized detection and quantitation of tumors pre or post-treatment to inform correct treatment and other trial options to each patient at the point of care. Furthermore, it facilitates remote global collaboration between Paige’s life sciences partners, contract research organizations, and clinical trial sites.
FullFocus viewer is FDA-cleared and CE-marked. It was designed to deploy new algorithms at scale, which allows for the rapid distribution of novel biomarkers to expedite enrollment and assist in identifying treatment pathways and trial options. It also includes anywhere access for pathologists and research teams and data management, a storage solution that is secure, scalable, and cost-effective.
Benefits:
- Pathologist-centric user experience.
- Faster performance to load and view digital slides.
- Access to digital slides for multiple users from any internet-allowed location.
- Secure, scalable, and highly available storage.
- Cost-effective.
- Minimal ongoing maintenance needs.
- Designed for interoperability with most scanners, monitors, and LIS
Paige and Royal Philips announced a strategic collaboration to deliver clinical-grade AI applications to pathology laboratories. These AI technologies, starting with Paige Prostate, aim to help pathologists identify, quantify, and characterize cancer in tissue samples. Once digital images are available, the CE-marked Paige Prostate software is applied automatically to detect and localize prostate cancer. (2)
Paige's Financial Aspect
Paige announced on January 14, 2021, a Series C funding round of $100 million. Casdin Capital and Johnson & Johnson Innovation, the strategic venture arm for Johnson & Johnson, led the Series C financing. Existing investors and other funds are also participating in the round.
In addition to promoting the development of AI-based clinical applications, biomarkers, and diagnostics, Paige will also invest in sales and marketing efforts to rapidly scale product adoption by hospitals and labs. Accordingly, the company expanded by building out its engineering and commercial teams by hiring about 70 new employees in 2021, doubling the previous size. (3)
Clinical-Grade Computational Pathology Using Weakly Supervised DL on Whole Slide Images
Multiple instance learning (MIL) uses a substantial pathology dataset to train deep neural networks, which are then used in a recurrent neural network (RNN) to incorporate the information across the whole slide and make an accurate diagnosis.
44,732 slides from 15,187 patients were examined, with results showing that the performances of the models trained at 20× magnification on the respective test datasets had an area under the curve (AUC) higher than 0.98.
The model sorts the data and announces positive slides for rapid review in a screening scenario by the pathologists. (4)
Digital Pathology System Helps Remote Review during the COVID-19 pandemic
Digital pathology systems have accepted the digitization of glass slides to generate whole slide images and have the potential to allow pathologists to review and report pathology slides
remotely. The COVID-19 pandemic has impeded the natural workflow of pathologists due to safety measures, and thus this is a solution to ensure continuous pathology practice.
The Whole Side Image viewer software is a standard whole slide image viewer that includes tools such as thumbnail viewing, slide label viewing, zooming, panning, and co-registration of multiple slides.
Comparisons between the remote review and conventional microscopy revealed a concordance rate of 100%. (6)
Digital pathology systems have accepted the digitization of glass slides to generate whole slide images and have the potential to allow pathologists to review and report pathology slides
remotely. The COVID-19 pandemic has impeded the natural workflow of pathologists due to safety measures, and thus this is a solution to ensure continuous pathology practice.
The Whole Side Image viewer software is a standard whole slide image viewer that includes tools such as thumbnail viewing, slide label viewing, zooming, panning, and co-registration of multiple slides.
Comparisons between the remote review and conventional microscopy revealed a concordance rate of 100%. (6)
Novel AI System Increases the Detection of Prostate Cancer in Whole Slide Images of Core Needle Biopsies
Prostate cancer (PrCa) represents the second most common cause of cancer among men in the United States. A prostate needle core biopsy is necessary for detecting PrCa. For this study, researchers evaluated how pathologists interact with Paige Prostate Alpha (PPA), a PrCa detection system.
Three AP-board-certified pathologists evaluated 304 prostate needle core biopsies. Then, the pathologists classified each WSI as benign or cancerous. Four weeks later, the pathologists were asked to re-evaluate each WSI with the assistance of PPA. The system marked the area where the cancer was detected with the highest probability for each WSI. Without PPA, pathologists had a sensitivity of 74% and a specificity of 97%. With PPA, the average sensitivity for pathologists increased to 90%, with no change in specificity. (7)
References