ABSTRACT
Background: Breast cancer is a prevalent malignancy globally, necessitating accurate diagnostic tools for early detection and intervention. Ultrasound imaging plays an important role in breast lesion assessment, with various morphological features serving as key indicators of malignancy. Objective: This study aimed to employ logistic regression analysis to investigate ultrasound indices for effectively distinguishing between benign and malignant breast masses. Methods: A careful retrospective analysis was conducted on a dataset comprising 388 pathologically confirmed breast masses retrieved from 364 female patients, of which 142 were identified as malignant and 246 as benign. The primary outcome measures included the aspect ratio of breast masses, clarity of mass boundaries, and presence of spiculations and angularity, which were assessed through ultrasound imaging and analyzed using multifactorial logistic regression. Results: The analysis demonstrated that the aspect ratio, clarity of boundaries, and presence of spiculations and angularity were significant independent risk factors for identifying malignant breast masses (P < .001). Multifactorial logistic regression revealed age (OR=1.183, 95% CI 1.119-1.252, P < .001), mass size (OR=1.087, 95% CI 1.036-1.140, P = .001), marginal spiculation (OR=8.296, 95% CI 2.325-29.598, P = .001), defined borders (OR=5.500, 95% CI 1.765-14.140, P = .003), aspect ratio (OR=5.830, 95% CI 1.742-19.505, P = .004), margin angularity (OR=5.183, 95% CI 1.910-14.063, P = .001), and marginal microtubules (OR=9.180, 95% CI 2.307-36.523, P = .002) significantly influenced mass benignity. Conclusions: The aspect ratio, boundary clarity, and presence of spiculation and angularity serve as crucial predictive indicators for distinguishing between benign and malignant breast lumps. Moreover, the utilization of a multifactorial logistic regression model significantly enhances the identification and differentiation of the benign and malignant nature of breast lumps. Continued research in this area is essential for further refining diagnostic approaches and enhancing overall breast cancer management.
