Abstract
Background: There is a critical global shortage of pathologists, with North America having 50-65 pathologists per million people, Europe 26 per million, Asia 6.8 per million, and Africa just 4 per million. This shortage slows disease severity assessment, prognostics, and therapy decisions, particularly for cancer, impacting survival rates. In South Asia, approximately 60% of breast cancer (BC) cases are diagnosed at stage III/IV, leading to poor outcomes. Delays in cancer care exceed 30.7 weeks due to patient and system-related issues. Global cancer cases are expected to rise from 2.26 million in 2020 to 2.74 million in 2030. Traditional histological grading by pathologists is manual, time-consuming, and prone to intra-observer (0.85) and inter-observer (0.43) variability.
Methods: To address these challenges, DCS_PathIMS, an automated, AI-driven digital pathology platform is proposed. Whole Slide Imaging (WSI) scanners digitize entire biopsy slides into high-resolution pyramidal TIFF files, which are processed using AI-based deep learning models. DCS_PathIMS, has a specialized web-based imaging platform facilitates the storage, visualization, and annotation of WSI data, integrating AI to assist pathologists in diagnostic workflows.
Results: The proposed DCS_PathIMS platform has been validated with an end-to-end breast cancer histology grading diagnostics workflow. Thus, automation enhances diagnostic accuracy, consistency, and efficiency by reducing human bias and workload. AI-powered analysis identifies known and novel biomarkers, improves reproducibility, and delivers fast, quantified assessments. Pathologists can engage, evaluate, and collaborate remotely, leading to faster and more accurate decisions. The clinician-friendly UI in the proposed platform, designed and validated by clinicians, streamlines workflows and reduces stress.
Conclusion: AI-driven digital pathology addresses the shortage of pathologists and enhances diagnostic efficiency and accuracy. The proposed platform improves clinical decision-making, facilitates faster reporting, reduces false positives/negatives, and supports better patient outcomes through transparent and consistent evaluations. This approach minimizes medical-legal risks and lowers insurance costs, driving advancements in cancer care.