Abstract:Rationale and Objectives This study aims to develop and validate a new diagnostic model based on the Kaiser score for preoperative diagnosis of the malignancy probability of enhancing lesions on breast MRI. Materials and Methods This study collected consecutive inpatient data (including imaging data, clinical data, and pathological data) from two different institutions. All patients underwent preoperative breast Dynamic Contrast-Enhanced Magnetic Resonance Imaging (DCE-MRI) examinations and were found to have enhancing lesions. These lesions were confirmed as benign or malignant by surgical resection or biopsy pathology (all carcinomas in situ were confirmed by pathology after surgical resection). Data from one institution were used as the training set(284 cases), and data from the other institution were used as the validation set(107 cases). The Kaiser score was directly incorporated into the diagnostic model as a single predictive variable. Other predictive variables were screened using Least Absolute Shrinkage and Selection Operator (LASSO) regression. Multivariate logistic regression was employed to integrate the Kaiser score and other selected predictive variables to construct a new diagnostic model, presented in the form of a nomogram. Receiver operating characteristic (ROC) curve, DeLong test, net reclassification improvement (NRI), and integrated discrimination improvement (IDI) were adopted to evaluate and compare the discrimination of the diagnostic model for breast enhancing lesions based on Kaiser score (hereinafter referred to as the "breast lesion diagnostic model") and the Kaiser score alone. Calibration curves were used to assess the calibration of the breast lesion diagnostic model, and decision curve analysis (DCA) was used to evaluate the clinical efficacy of the diagnostic model and the Kaiser score. Results LASSO regression indicated that, besides the indicators already included in the Kaiser score system, "age", "MIP sign", "associated imaging features", and "clinical breast examination (CBE) results" were other valuable diagnostic parameters for breast enhancing lesions. In the training set, the AUCs of the breast lesion diagnostic model and the Kaiser score were 0.948 and 0.869, respectively, with a statistically significant difference (p < 0.05). In the validation set, the AUCs of the breast lesion diagnostic model and the Kaiser score were 0.956 and 0.879, respectively, with a statistically significant difference (p < 0.05). The DeLong test, NRI, and IDI showed that the breast lesion diagnostic model had a higher discrimination ability for breast enhancing lesions compared to the Kaiser score alone, with statistically significant differences (p < 0.05). The calibration curves indicated good calibration of the breast lesion diagnostic model. DCA demonstrated that the breast lesion diagnostic model had higher clinical application value, with greater net clinical benefit over a wide range of diagnostic thresholds compared to the Kaiser score. Conclusion The Kaiser score-based breast lesion diagnostic model, which integrates "age," "MIP sign", "associated imaging features", and "CBE results", can be used for the preoperative diagnosis of the malignancy probability of breast enhancing lesions, and it outperforms the classic Kaiser score in terms of diagnostic performance for such lesions.

Enhancing the Kaiser score for lesion characterization in unenhanced breast MRI

Development and Validation of a Diagnostic Model for Enhancing Lesions on Breast MRI: Based on Kaiser Score

Reliability of Kaiser Score in Assessing Additional Breast Lesions Identified on Staging MRI in Patients with Breast Cancer

Breast lesions on MRI in mass and non-mass enhancement: Kaiser score and modified Kaiser score + for readers of variable experience

Application of the Kaiser score on contrast-enhanced mammography in the differential diagnosis of breast lesions: comparison with breast magnetic resonance imaging

A Comparative Assessment of MR BI-RADS 4 Breast Lesions With Kaiser Score and Apparent Diffusion Coefficient Value

Diagnostic value of Kaiser score combined with breast vascular assessment from breast MRI for the characterization of breast lesions

Diagnostic performance of the Kaiser score for contrast-enhanced mammography and magnetic resonance imaging in breast masses: A Comparative Study

Diagnostic accuracy of the breast MRI Kaiser score in suspected architectural distortions and its comparison with mammography

Improved Differential Diagnosis Based on BI-RADS Descriptors and Apparent Diffusion Coefficient for Breast Lesions: A Multiparametric MRI Analysis as Compared to Kaiser Score

Evaluation of the differentiation of benign and malignant breast lesions using synthetic relaxometry and the Kaiser score

A Study of a Breast MRI Abbreviated Protocol for Breast Neoplasm Screening

Value of Scoring by Contrast-Enhanced Ultrasound Combined with Ultrasonic Elastography in Differential Diagnosis of BI-RADS 4 Lesions

Magnetic resonance imaging of fibroadenoma-like lesions and correlation with Breast Imaging-Reporting and Data System and Kaiser scoring system

Value of contrast-enhanced mammography combined with the Kaiser score for clinical decision-making regarding tomosynthesis BI-RADS 4A lesions

Pilot study: A simple CAD-based tool to detect breast cancer on MRI of the breast

Preoperative Staging in Breast Cancer: Intraindividual Comparison of Unenhanced MRI Combined With Digital Breast Tomosynthesis and Dynamic Contrast Enhanced-MRI

A risk prediction stratification for non-mass breast lesions, combining clinical characteristics and imaging features on ultrasound, mammography, and MRI

Comparison of Diagnostic Performance between Classic and Modified Abbreviated Breast MRI and the MRI Features Affecting Their Diagnostic Performance

Characterization of common breast MRI abnormalities: comparison between abbreviated and full MRI protocols

Artificial Intelligence-Based Classification of Breast Lesions Imaged With a Multiparametric Breast MRI Protocol With Ultrafast DCE-MRI, T2, and DWI