Pathology artificial intelligence (AI) software developer Paige is highlighting a paper published July 15 in Nature Medicine that indicates the company's technology can be used to develop AI algorithms with "near-perfect accuracy" for analyzing pathology slides for prostate cancer, skin cancer, and breast cancer.
In the paper, Chief Scientific Officer Thomas Fuchs, PhD, of Memorial Sloan Kettering Cancer Center and colleagues describe how a series of deep-learning algorithms for clinical decision support in pathology were developed with an automated training and testing technique. Fuchs is the senior author on the paper, with his student Gabriele Campanella as the first author.
The deployment of clinical decision support for pathology has been hindered by the need to curate large, manually annotated datasets to test and train AI algorithms, the authors noted. Instead, Campanella et al present a system in which algorithms are trained using only the reported diagnoses. This eliminates the need for pixel-wise manual annotations.
The researchers tested the technique on a dataset of 44,732 whole slide images from 15,187 patients with no data curation used, applying the approach to prostate cancer, basal cell carcinoma, and breast cancer metastases to axillary lymph nodes. They found an area under the curve (AUC) of 0.98 for all cancer types.
In applying the technique to clinical applications, the technology could retain 100% sensitivity while allowing pathologists to exclude 65% to 75% of slides that do not have any signs of cancer. Furthermore, the system could be used to "train accurate classification models at unprecedented scale," they wrote, and enable the use of computer-based decision support in clinical pathology.
Indeed, Paige plans to commercialize several algorithms developed using the technique that can be built into clinical decision-support applications for pathologists. One such algorithm -- designed for an application other than the ones tested in the Nature Medicine paper -- is currently under review at the U.S. Food and Drug Administration through the agency's breakthrough device regulatory pathway.