Histopathology is considered the gold standard in determining the presence and nature of tumors. Technological advances in automated high-speed and high-resolution whole-slide imaging have laid the foundation for a digital revolution in microscopy. Digital histopathological images can be analyzed efficiently with image analysis and machine learning techniques. These techniques have shown great potential for extracting sub-visual, quantitative, and valuable features from whole-slide images to characterize tumors. Besides histopathological images, other data modalities, such as radiological images and multi-omics data, are also used to assist the decision-making process for cancer diagnosis, treatment, and prognosis. At present, it is not clear how these macroscopic, microscopic, and molecular features are related. There is also a need to integrate multimodal data in order to achieve precision medicine.
Computational pathology is an integrative approach that leverages multiple sources of data not only from histopathology in the narrow sense, but also from radiology, genetics, genomics, and electronic medical records, to make the best possible medical decisions. Such integrative analyses require techniques and innovations from radiomics, bioimage informatics, genomics, bioinformatics, computer vision, and machine learning.
The aim of this Research Topic is to highlight the latest developments in computational pathology for improved clinical decision making that use either classical image analysis or state-of-the-art deep learning solutions. Topics of interest include, but are not limited to:
• Novel methods to quantify tumor heterogeneity or tumor microenvironment from histopathological images
• Discovery of diagnostic, prognostic, and predictive biomarkers from histopathological images
• Associations between histopathological image features and radiomic features (e.g., CT, MRI, ultrasound)
• Associations between histopathological image features and genomic/genetic features
• Integrative analysis of multimodal data (e.g., histological images, omics data, radiological images) for making diagnosis and for predicting prognosis and treatment response
Histopathology is considered the gold standard in determining the presence and nature of tumors. Technological advances in automated high-speed and high-resolution whole-slide imaging have laid the foundation for a digital revolution in microscopy. Digital histopathological images can be analyzed efficiently with image analysis and machine learning techniques. These techniques have shown great potential for extracting sub-visual, quantitative, and valuable features from whole-slide images to characterize tumors. Besides histopathological images, other data modalities, such as radiological images and multi-omics data, are also used to assist the decision-making process for cancer diagnosis, treatment, and prognosis. At present, it is not clear how these macroscopic, microscopic, and molecular features are related. There is also a need to integrate multimodal data in order to achieve precision medicine.
Computational pathology is an integrative approach that leverages multiple sources of data not only from histopathology in the narrow sense, but also from radiology, genetics, genomics, and electronic medical records, to make the best possible medical decisions. Such integrative analyses require techniques and innovations from radiomics, bioimage informatics, genomics, bioinformatics, computer vision, and machine learning.
The aim of this Research Topic is to highlight the latest developments in computational pathology for improved clinical decision making that use either classical image analysis or state-of-the-art deep learning solutions. Topics of interest include, but are not limited to:
• Novel methods to quantify tumor heterogeneity or tumor microenvironment from histopathological images
• Discovery of diagnostic, prognostic, and predictive biomarkers from histopathological images
• Associations between histopathological image features and radiomic features (e.g., CT, MRI, ultrasound)
• Associations between histopathological image features and genomic/genetic features
• Integrative analysis of multimodal data (e.g., histological images, omics data, radiological images) for making diagnosis and for predicting prognosis and treatment response
