A Systematic Interaction Analysis of Distance Measures and Feature Representations in Mammography

Iqbal, Muhammad; Butt, Talal Ashraf; Shareef, Zahid; Siddique, Abubakar; Browne, Will N

doi:10.1109/access.2026.3699018

A Systematic Interaction Analysis of Distance Measures and Feature Representations in Mammography

Files

Journal article

Size: 2.8 MB, File format: Adobe PDF

Date

2026-06-01

Authors

Iqbal, Muhammad

Butt, Talal Ashraf

Shareef, Zahid

Siddique, Abubakar

Browne, Will N

Item type

Journal Article

Publisher

Institute of Electrical and Electronics Engineers (IEEE)

Abstract

Distance-based classifiers such as k -nearest neighbors remain widely used in mammographic computer-aided diagnosis and image retrieval, yet distance measure selection is rarely evaluated systematically. This paper investigates how the interaction between feature representation and distance formulation shapes mammographic lesion classification, evaluating 20 distance measures across 13 feature sets (handcrafted, deep, hybrid), three datasets (MIAS, CBIS-DDSM, INbreast), and two classifiers ( k -NN, nearest mean-distance), spanning 13,260 configurations under 5-fold cross-validation. Five findings emerge. First, distance choice significantly affects performance (Friedman χ²=1901 , p≈0 ; best–worst AUC gap = 0.079). Second, the optimal measure depends on feature type: handcrafted features favor Learned Mahalanobis, deep features favor Adaptive Weighted Euclidean, and hybrid features favor the proposed CCFCD (AUC = 0.712). Third, measure rankings are dataset-specific (cross-dataset ρ as low as −0.10). Fourth, k -NN and NMD agree at the configuration level ( ρ=0.94 ) but not at the measure level ( ρ=0.077 ). Fifth, increasing k improves AUC while reducing sensitivity from 0.40 to 0.14; 11.6% of configurations achieve accuracy above 80% with sensitivity below 10%, showing that AUC and accuracy alone are insufficient for clinical assessment. A computational complexity analysis shows that per-pair costs range from O(d) for most measures to O(d²) for Mahalanobis, with the top-performing adaptive measures remaining tractable. These findings demonstrate that distance-based mammographic classification is governed by a coupled interaction among feature type, distance formulation, classifier, neighborhood size, and evaluation metric. Adaptive and class-aware measures consistently outperform conventional defaults when matched to the feature space.

Keywords

46 Information and Computing Sciences, 40 Engineering, 08 Information and Computing Sciences, 09 Engineering, 10 Technology, Mammography, breast cancer classification, distance measures, similarity measures, nearest neighbor classification, k-nearest neighbors, feature representation

Source

IEEE Access, ISSN: 2169-3536 (Print); 2169-3536 (Online), Institute of Electrical and Electronics Engineers (IEEE), 14, 85033-85057. doi: 10.1109/access.2026.3699018