Label-free optical microscopy with artificial intelligence: a new paradigm in pathology
Chiho Yoon, Eunwoo Park, Donggyu Kim, Byullee Park, Chulhong Kim
Abstract
Significance: Pathological examination is essential for diagnosing diseases in tissues such as cancer but involves labor-intensive and time-consuming processes. Label-free optical microscopy has emerged as a promising alternative that offers the ability to visualize tissue structures without the need for histochemical staining. Further, the integration of artificial intelligence (AI) into label-free microscopy has the potential to streamline the overall pathological diagnostic process. Aim: We aim to review the use of AI-assisted label-free optical microscopy in revolutionizing pathological workflows. Approach: We examine the integration of AI with label-free optical microscopy techniques and assess its overall impact on the pathological workflow. We evaluate how AI enhances each stage of label-free pathology, including specimen preparation, label-free imaging, virtual staining, and diagnostic analysis. Results: Label-free optical microscopy with AI has significantly improved the entire pathological workflow. AI assists specimen preparation with high efficiency, enhances label-free imaging with high resolution and speed, and enables cost-effective virtual staining with high throughput and automatic diagnostic analysis with high accuracy. Conclusions: AI-aided label-free optical microscopy enhances diagnostic speed, accuracy, and specimen preservation, offering a transformative approach that could redefine traditional pathology workflows and improve clinical outcomes.