Understanding the decisions of deep learning (DL) models is crucial for their acceptance in risk-sensitive applications. Class activation maps (CAMs) are commonly used in image analysis to visualize model reasoning by generating attention maps where signal intensities represent contributions to outputs. However, existing CAM algorithms focus on optimal weight design and salient feature layer selection, neglecting two key limitations in population-level interpretation: (1) lack of an appropriate intensity scale for proper interpretation and quantitative analysis, and (2) inability to identify which features within selected layers predominantly drive model reasoning. These gaps can lead to miscorrelations between CAMs and model outputs, causing erroneous interpretations, while restricting CAMs to case-specific visual inspections. We propose a framework to statistically analyze DL-extracted features at a population level, determinizing feature contributions for global intensity scaling and within-layer feature distinction. The global intensity scale standardizes CAMs, achieving high correlation coefficients (R) with model outputs. Within-layer feature distinction identifies overfitting, confounding factors, outliers, redundancies, and principal features. Applied to eight datasets, including five medical imaging datasets, this framework improved Rs between CAMs and outputs by 10.7–64.2%, achieving near 100% consistency. Furthermore, distinguishing principal features (5–25% in the selected layer) produced CAMs of equal quality while maintaining model output accuracy. This method, relying on minimal assumptions, enhances CAM-model consistency and broadens CAM applications by enabling standardized interpretation and deeper feature-level insights.

Pulmonary fissures are important landmarks for recognition of lung anatomy. In CT images, automatic detection of fissures is complicated by factors like intensity variability, pathological deformation and imaging noise. To circumvent this problem, we propose a derivative of stick (DoS) filter for fissure enhancement and a post-processing pipeline for subsequent segmentation. Considering a typical thin curvilinear shape of fissure profiles inside 2D cross-sections, the DoS filter is presented by first defining nonlinear derivatives along a triple stick kernel in varying directions. Then, to accommodate pathological abnormality and orientational deviation, a [Formula: see text] cascading and multiple plane integration scheme is adopted to form a shape-tuned likelihood for 3D surface patches discrimination. During the post-processing stage, our main contribution is to isolate the fissure patches from adhering clutters by introducing a branch-point removal algorithm, and a multi-threshold merging framework is employed to compensate for local intensity inhomogeneity. The performance of our method was validated in experiments with two clinical CT data sets including 55 publicly available LOLA11 scans as well as separate left and right lung images from 23 GLUCOLD scans of COPD patients. Compared with manually delineating interlobar boundary references, our method obtained a high segmentation accuracy with median F1-scores of 0.833, 0.885, and 0.856 for the LOLA11, left and right lung images respectively, whereas the corresponding indices for a conventional Wiemker filtering method were 0.687, 0.853, and 0.841. The good performance of our proposed method was also verified by visual inspection and demonstration on abnormal and pathological cases, where typical deformations were robustly detected together with normal fissures.

The intensity or gray-level derivatives have been widely used in image segmentation and enhancement. Conventional derivative filters often suffer from an undesired merging of adjacent objects because of their intrinsic usage of an inappropriately broad Gaussian kernel; as a result, neighboring structures cannot be properly resolved. To avoid this problem, we propose to replace the low-level Gaussian kernel with a bi-Gaussian function, which allows independent selection of scales in the foreground and background. By selecting a narrow neighborhood for the background with regard to the foreground, the proposed method will reduce interference from adjacent objects simultaneously preserving the ability of intraregion smoothing. Our idea is inspired by a comparative analysis of existing line filters, in which several traditional methods, including the vesselness, gradient flux, and medialness models, are integrated into a uniform framework. The comparison subsequently aids in understanding the principles of different filtering kernels, which is also a contribution of this paper. Based on some axiomatic scale-space assumptions, the full representation of our bi-Gaussian kernel is deduced. The popular γ-normalization scheme for multiscale integration is extended to the bi-Gaussian operators. Finally, combined with a parameter-free shape estimation scheme, a derivative filter is developed for the typical applications of curvilinear structure detection and vasculature image enhancement. It is verified in experiments using synthetic and real data that the proposed method outperforms several conventional filters in separating closely located objects and being robust to noise.

EMPIRE10 (Evaluation of Methods for Pulmonary Image REgistration 2010) is a public platform for fair and meaningful comparison of registration algorithms which are applied to a database of intrapatient thoracic CT image pairs. Evaluation of nonrigid registration techniques is a nontrivial task. This is compounded by the fact that researchers typically test only on their own data, which varies widely. For this reason, reliable assessment and comparison of different registration algorithms has been virtually impossible in the past. In this work we present the results of the launch phase of EMPIRE10, which comprised the comprehensive evaluation and comparison of 20 individual algorithms from leading academic and industrial research groups. All algorithms are applied to the same set of 30 thoracic CT pairs. Algorithm settings and parameters are chosen by researchers expert in the configuration of their own method and the evaluation is independent, using the same criteria for all participants. All results are published on the EMPIRE10 website (http://empire10.isi.uu.nl). The challenge remains ongoing and open to new participants. Full results from 24 algorithms have been published at the time of writing. This paper details the organization of the challenge, the data and evaluation methods and the outcome of the initial launch with 20 algorithms. The gain in knowledge and future work are discussed.