As deep learning is widely used in the radiology field, the explainability of Artificial Intelligence (AI) models is becoming increasingly essential to gain clinicians’ trust when using the models for diagnosis. In this research, three experiment sets were conducted with a U-Net architecture to improve the disease classification performance while enhancing the heatmaps corresponding to the model's focus through incorporating heatmap generators during training. All experiments used the dataset that contained chest radiographs, associated labels from one of the three conditions [“normal”, “congestive heart failure (CHF)”, and “pneumonia”], and numerical information regarding a radiologist's eye-gaze coordinates on the images. The paper that introduced this dataset developed a U-Net model, which was treated as the baseline model for this research, to show how the eye-gaze data can be used in multi-modal training for explainability improvement and disease classification. To compare the classification performances among this research's three experiment sets and the baseline model, the 95% confidence intervals (CI) of the area under the receiver operating characteristic curve (AUC) were measured. The best method achieved an AUC of 0.913 with a 95% CI of [0.860, 0.966]. “Pneumonia” and “CHF” classes, which the baseline model struggled the most to classify, had the greatest improvements, resulting in AUCs of 0.859 with a 95% CI of [0.732, 0.957] and 0.962 with a 95% CI of [0.933, 0.989], respectively. The decoder of the U-Net for the best-performing proposed method generated heatmaps that highlight the determining image parts in model classifications. These predicted heatmaps, which can be used for the explainability of the model, also improved to align well with the radiologist's eye-gaze data. Hence, this work showed that incorporating heatmap generators and eye-gaze information into training can simultaneously improve disease classification and provide explainable visuals that align well with how the radiologist viewed the chest radiographs when making diagnosis.

To fully and rapidly develop a real-time early warning judgment system for slope failure at the time of heavy rains including overseas, it is necessary to predict water movement in the soil at the time of rainfall. In addition, to apply the system to a place where insufficient geotechnical and geological data have been amassed, it is necessary to evaluate the risk of slope failure based on physical properties obtained from a simple soil test. Therefore, in this study, the authors set Gogoshima Island in Ehime Prefecture as a study site and evaluated the water movement over time in the soil during heavy rain using a simple prediction equation of rainfall seepage process. Soil properties were determined through simple in-situ and laboratory tests. As a result, it was found that the factor of safety for slope failure in the head and wall of a valley dissecting the hillside slope composed of granodiorite in which weathering has progressed can be planarly evaluated using the simple prediction equation.

In recent years, sediment disaster has frequently been caused by heavy rainfall and has cost many human lives and great property losses. To estimate such risks, Wakai et al. [1] proposed a simplified prediction method to calculate the variation of groundwater levels in natural slopes both at the time of rainfall in wide areas and in real time. To calculate the variation of groundwater levels using this method, the slope conditions (such as material constant and initial conditions) must be determined in advance. This study takes the 2017 heavy rainfall in Northern Kyushu as an example to analyze surface layer thickness, one of the slope conditions that most significantly influences slope stability, over wide areas. The findings reveal that the prediction of slope failure distribution differs depending on how the surface layer thickness and sliding surface are determined.

No abstract

Measuring the amount of rainfall is essential for a wide-area evaluation of the risk of landslide disaster using a real-time simulation. In Thailand, located in Monsoon Asia, point observation is conducted using a rain gauge. Interpolation calculation is crucial for obtaining the planar rainfall intensity for the wide-area analysis from scattered point observation data. In this study, to accurately calculate rainfall intensity using the inverse distance weighting (IDW) method, the parameters affecting the results are examined. Additionally, using obtained rainfall data, a simple prediction calculation of groundwater level fluctuation by Wakai et al. [1] and Ozaki et al. [2] is performed. Finally, the relationship between the rainfall intensity and the fluctuation of groundwater level will be discussed.