Yisha Sun

Histopathological images stained with hematoxylin and eosin (H&E) are crucial for cancer diagnosis and prognosis. However, color variations caused by differences in tissue preparation and scanning devices can lead to data distribution discrepancies, adversely affecting the performance of downstream algorithms in tasks like classification, segmentation, and detection. To address these issues, stain normalization methods have been developed to standardize color distributions across images from various sources. Recent advancements in deep learning-based stain normalization methods have shown significant promise due to their minimal preprocessing requirements, independence from reference templates, and robustness. This review examines 115 publications to explore the latest developments in this field. We first outline the evaluation metrics and publicly available datasets used for assessing stain normalization methods. Next, we systematically review deep learning-based approaches, including supervised, unsupervised, and self-supervised methods, categorizing them by core technologies and analyzing their contributions and limitations. Finally, we discuss current challenges and future directions, aiming to provide researchers with a comprehensive understanding of the field, promote further development, and accelerate the progress of intelligent cancer diagnosis.

Forecasting the water supply of a water treatment plant is an important management and decision-making task in the water treatment system. Currently, most studies in this field do not consider the problems brought about by the time scale when selecting the input feature set. Therefore, this study explores how to select the relevant input feature set and model for the ultra-short-term water demand forecasting (UWDF) of water treatment plants from the perspective of time multi-scale features, in order to provide more accurate and effective water supply planning for water treatment plants. This study uses feature engineering to analyze the importance of relevant features affecting water supply, and compares and analyzes the predictive capabilities of seven deep learning models. The experimental results show that the time multi-scale features (TMFs) extracted from the historical water supply data play an important role in the prediction of ultra-short-term water supply, and the LSTM model performs better in the prediction task, with a mean absolute percentage error (MAPE) of 5.3530% and a goodness-of-fit (R²) of 0.9981. This study provides an effective method for intelligent operation and maintenance in the field of water treatment, and has broad application prospects.