Temporal subtraction is a technique that calculates a subtraction image of a pair of registered images acquired from the same patient at different time points. Past studies have shown that the technique is effective in visualizing pathological changes over time, and therefore, it has been expected to be a supporting tool for medical doctors in radiological interpretation. However, in case of thick-slice images (which are widely used in actual clinical practice), even with an accurate image registration, artifacts caused by partial volume effects have deteriorated quality of the subtraction image. In this study, we proposed a method that reduces such artifacts by considering gaps of discretizing position based on the slice thickness of the two input images in the calculations of the subtraction. In this paper, we evaluated the proposed method in the following two ways: 1) quantitative evaluation using synthetic data and 2) radiologists' subjective evaluation using clinical data. Our results show that, in terms of both quantitative evaluation and radiologists' subjective evaluation, the proposed method was superior to the conventional subtraction method.

View full abstract

To compensate for an insufficiency of the case images needed in the development of CAD (computer-aided diagnosis), work is underway to create artificial case images by embedding tumors and other such lesions into lesion-free images. Previously, the authors have demonstrated the effectiveness of creating artificial case images for hepatic and breast tumors and utilizing them in CAD development. Thus far, however, when training data comprising 50% or more artificial cases is used in CAD development, the resulting discrimination performance on unknown data has tended to be somewhat inferior compared to when training data comprises only actual cases. With the objectives of applying artificial case images to a greater range of sites and of using exclusively artificial cases to develop a high-performance discriminator, in this study, effectiveness verification was conducted that focused on breast cancer calcifications as a new target. Because the characteristics of calcification shadows differ substantially from those of the hepatic and breast cancer tumor shadows studied thus far, a new artificial image creation technique was developed. Artificial cases created using this technique were applied to CAD development. As a result, a discriminator trained with 100% artificial cases obtained detection performance equal to that of a discriminator trained with entirely actual cases.

View full abstract

Pathologists visually observe hematoxylin-eosin (HE) stained images under a microscope to diagnose the presence or absence of tumors. However, variations exist in stained shades because of different HE staining methods. Such differences limit improvements in diagnosis. Conventional machine learning techniques create classifiers with features designed by humans. However, designing effective features for identifying tumor tissue requires a lot of time. This study proposes a method of automatically identifying the presence or absence of a tumor in a pathological image. The method consists of three steps: 1. pathological images with different stained shades are automatically classified by performing axial transformation by principal component analysis, 2. a classifier is created with a convolutional neural network (CNN) for each image group, and 3. the presence or absence of a tumor is judged by using the classifier. The experimental results using digital images of pathological tissue specimens of gastric moderately differentiated adenocarcinoma show improved identification accuracy.

View full abstract