PathNarratives: Data annotation for pathological human-AI collaborative diagnosis

Abstract

Pathology is the gold standard of clinical diagnosis. Artificial intelligence (AI) in pathology becomes a new trend, but it is still not widely used due to the lack of necessary explanations for pathologists to understand the rationale. Clinic-compliant explanations besides the diagnostic decision of pathological images are essential for AI model training to provide diagnostic suggestions assisting pathologists practice. In this study, we propose a new annotation form, PathNarratives, that includes a hierarchical decision-to-reason data structure, a narrative annotation process, and a multimodal interactive annotation tool. Following PathNarratives, we recruited 8 pathologist annotators to build a colorectal pathological dataset, CR-PathNarratives, containing 174 whole-slide images (WSIs). We further experiment on the dataset with classification and captioning tasks to explore the clinical scenarios of human-AI-collaborative pathological diagnosis. The classification tasks show that fine-grain prediction enhances the overall classification accuracy from 79.56 to 85.26%. In Human-AI collaboration experience, the trust and confidence scores from 8 pathologists raised from 3.88 to 4.63 with providing more details. Results show that the classification and captioning tasks achieve better results with reason labels, provide explainable clues for doctors to understand and make the final decision and thus can support a better experience of human-AI collaboration in pathological diagnosis. In the future, we plan to optimize the tools for the annotation process, and expand the datasets with more WSIs and covering more pathological domains.

Get Real Get Better: A Framework for Developing Agile Program Management in the U.S. Navy Supported by the Application of Advanced Data Analytics and AI

“…Such collaborative efforts, where the human-AI team outperforms the individual agents acting separately, are also called hybrid intelligence [32,33]. Examples of such effective human-AI collaboration include knowledge work [34], data annotation [35], medical diagnosis [36], mental health [37], and computationally informed mission planning [38]. Furthermore, the new developments in the theory of organizational decision-making [39] also suggest effective ways in which the decisions of organizational members can be combined with AI-based decisions for successful human-AI teaming and collaboration in the project management area.…”

Promoting higher-valent pediatric combination vaccines in China: challenges and recommendations for action

“…Fourth, low awareness and high price are other important hurdles behind the low uptake of the higher-valent combination vaccines [ 38 ]. A recent study showed general low awareness of non-NIP vaccines, and awareness of pediatric pentavalent vaccines was lowest [ 39 ]. Based on a study conducted in the well-developed Dongcheng District of Beijing, among 183 parents, only 21.31% (39) had heard of pentavalent vaccine and only 10.93% knew that higher-valent pediatric combination vaccines could substitute for program vaccines [ 40 ].…”

Upregulated lncRNA LINC01128 in colorectal cancer accelerates cell growth and predicts malignant prognosis through sponging miR-363-3p

“…CRC is the second leading cause of cancer deaths in the United States, and more than 34% of CRC patients died in 2023, according to a new report (Siegel et al 2023 ). The new incidence of CRC in China was 12.2% by 2020, and the incidence in urban population was much higher than that in rural areas (Wang et al 2023b ; Zhang et al 2022 ). Currently, it is believed that family heredity, dietary habits, lifestyle and environment may be the predisposing factors of CRC, but the specific etiology has not been clarified (Zhou et al 2022b ).…”