Thanks to the fast development in optical imaging technologies, whole-slide imaging is becoming a standard imaging tool in pathology. In the past decade, plenty of studies have demonstrated the remarkable success of AI, particularly deep learning, in processing whole slide images for digital pathology. The integration of computational pathology with other modalities including radiology and bioinformatics could make a breakthrough into the established workflows in research and clinical settings.
Most of the successful deep learning solutions are currently supervised learning. However, whole slide images are massive by scale and both time-consuming and costly for detailed annotations. In cancer pathology studies, the morphological heterogeneity and strong resemblance to numerous non-malignant conditions make interpretation and clinical decisions complex, even for experienced pathologists. The AI-based systems applying state-of-the-art computational algorithms with integration of different data modalities including genetic and radiology could assist more accurate clinical decision. Thus, the development of automated, reliable detection, segmentation, classification, and prediction algorithms has become an important research focus in medical image computing for pathology.
This Research Topic focuses on the advances in machine learning and deep learning-based methods using whole-slide imaging or integration with other data modalities including radiology, bioinformatics and clinical data for improving cancer diagnosis, prognosis and therapy. Prospective authors are invited to submit Original Research as well as Review articles on topics including but not limited to the following themes:
• Whole-slide image analysis
• Semi-supervised learning or self-supervised learning in computational pathology
• Graph networks for pathology applications Explainable AI systems for pathology
• Machine learning and deep learning tools for predictive and prognostic biomarkers and patient outcome in computational pathology applications.
• Multimodal domain adaptation and shifting among pathology, radiology and bioinformatics using machine learning and deep learning
• Applications of computational pathology in the clinic
Thanks to the fast development in optical imaging technologies, whole-slide imaging is becoming a standard imaging tool in pathology. In the past decade, plenty of studies have demonstrated the remarkable success of AI, particularly deep learning, in processing whole slide images for digital pathology. The integration of computational pathology with other modalities including radiology and bioinformatics could make a breakthrough into the established workflows in research and clinical settings.
Most of the successful deep learning solutions are currently supervised learning. However, whole slide images are massive by scale and both time-consuming and costly for detailed annotations. In cancer pathology studies, the morphological heterogeneity and strong resemblance to numerous non-malignant conditions make interpretation and clinical decisions complex, even for experienced pathologists. The AI-based systems applying state-of-the-art computational algorithms with integration of different data modalities including genetic and radiology could assist more accurate clinical decision. Thus, the development of automated, reliable detection, segmentation, classification, and prediction algorithms has become an important research focus in medical image computing for pathology.
This Research Topic focuses on the advances in machine learning and deep learning-based methods using whole-slide imaging or integration with other data modalities including radiology, bioinformatics and clinical data for improving cancer diagnosis, prognosis and therapy. Prospective authors are invited to submit Original Research as well as Review articles on topics including but not limited to the following themes:
• Whole-slide image analysis
• Semi-supervised learning or self-supervised learning in computational pathology
• Graph networks for pathology applications Explainable AI systems for pathology
• Machine learning and deep learning tools for predictive and prognostic biomarkers and patient outcome in computational pathology applications.
• Multimodal domain adaptation and shifting among pathology, radiology and bioinformatics using machine learning and deep learning
• Applications of computational pathology in the clinic
