MRI Features and Apparent Diffusion Coefficient Histogram-Based Nomogram for Classifying MRI-Only Suspicious Breast Lesions.

Abstract

Purpose: This study aimed to develop and validate a nomogram integrating clinicoradiologic features and apparent diffusion coefficient (ADC)-based histogram parameters for MRI-only suspicious lesions.

Methods: Ninety patients with MRI-detected suspicious lesions, who underwent breast MRI between May 2017 and August 2023, were retrospectively included and randomly assigned to a training cohort (n = 62) and a validation cohort (n = 28). Clinical and MRI data for each patient were reviewed and analyzed. Mean ADC values were computed using small two-dimensional region of interest measurements from ADC maps, followed by histogram analysis of the ADC maps, yielding 17 extracted histogram parameters. Univariate and multivariate logistic regression analyses identified significant variables associated with malignancy, which were incorporated into the nomogram. The diagnostic performance of these variables and the nomogram was evaluated using the area under the receiver operating characteristic (ROC) curve (AUC) and DeLong's test.

Results: Univariate analysis revealed significant differences between malignant and benign groups in terms of margin, kinetic pattern, mean ADC, and four ADC histogram parameters (ADC energy, ADC entropy, ADC range, and ADC uniformity) (all P < .05). Multivariate analysis identified kinetic pattern (P = .005, odds ratio [OR] = 2.569) and ADC entropy (P = .003, OR = 6.687) as significant predictors of MRI-only suspicious lesion classification. The nomogram combining kinetic pattern and ADC entropy demonstrated a C-index of 0.820 (95% confidence interval [CI]: 0.714-0.927) in the training cohort and 0.728 (95% CI: 0.528-0.878) in the validation cohort.

Conclusions: This nomogram, integrating kinetic pattern and ADC entropy, provides a simple, noninvasive tool for classifying MRI-only suspicious lesions, offering superior performance compared to mean ADC values.