Accuracy of the Fischer scoring system and the Breast Imaging Reporting and Data System in identification of malignant breast lesions

Abstract

BACKGROUND AND OBJECTIVES:Fischer developed a scoring system in 1999 that made identifying malignant lesions much easier for inexperienced radiologists. Our study was performed to assess whether this scoring system would help beginners to accurately diagnose breast lesions on magnetic resonance (MR) imaging and to assess the correlation between the magnetic resonance mammography Breast Imaging Reporting and Data System (MRM BI-RADS) grade and the final diagnosis.PATIENTS AND METHODS:The lesion morphology and contrast kinetics of 63 masses in 41 patients were evaluated on MRI and accorded a MRM BI-RADS final assessment category using the Fischer scoring system. The accuracy was evaluated after the final diagnosis was obtained by tissue sampling and follow-up imaging.RESULTS:There were 25 malignant and 30 benign lesions. Eight lesions were seen by MRI only and we could not verify their pathology since we did not have MR-guided biopsy facilities at the time of the study. On MR mammography, the proven carcinomatous lesions were characterized as BI-RADS category V in 16 (64%), category IV in 7 (28%), and category III in 2 (8%) lesions. Benign lesions were graded as category V in 3 (10%), category IV in 6 (20%), and category III in 3 (10%), category II in 10 (33%) and category I in 8 (27%) lesions. The MRM BI-RADS category accurately predicted malignancy in 92% and a benign pathology in 70% of the lesions. The overlap between the MRM features of chronic inflammatory lesions and carcinomas resulted in a lower accuracy in diagnosing benign as compared to malignant lesions.CONCLUSION:The MRM BI-RADS lexicon using the Fischer scoring system is useful and has a high predictive value, especially for malignant breast lesions, and is easy to apply. Overlapping features between benign inflammatory and malignant lesions might yield a reduced accuracy in inflammatory pathologies.

Enhanced Masses on Contrast‐Enhanced Breast: Differentiation Using a Combination of Dynamic Contrast‐Enhanced MRI and Quantitative Evaluation with Synthetic MRI

“…The BI‐RADS category 4 assessment covers a wide range of malignancy probabilities (>2% but <95% likelihood of malignancy) 9 . The rate of false‐positive cases can be considerable, especially for BI‐RADS 4 lesions in both this study and previous studies 11,28,29 . This means that the current policy for management of breast lesions has been adjusted based on a conservative set point to decrease the rate of missed cancers at the cost of performing unnecessary invasive procedures 30 .…”

“…DCE‐MRI is a sensitive tool for the detection of breast cancer, with a reported sensitivity of 94–100% 10,11 . Diagnostic criteria that combine morphologic and hemodynamic characteristics based on the BI‐RADS are effective for the evaluation of breast lesions.…”

“…Diagnostic criteria that combine morphologic and hemodynamic characteristics based on the BI‐RADS are effective for the evaluation of breast lesions. However, the specificity of DCE‐MRI has been reported to be as low as 37–97% 2,10,11 . This variable specificity is a limitation that can result in unnecessary biopsies, which can result in increased healthcare costs, patient anxiety, and delays in the treatment of breast cancer 12 .…”

Value of breast MRI for patients with a biopsy showing atypical ductal hyperplasia (ADH)

“…The other two studies were retrospective. Linda et al 15 used the Fischer score, which is known to correlate well with BI-RADS assessment, 21 defining I-III as nonsuspicious and IV-V as suspicious. Londero et al 18 used a simple assessment of the presence or absence of enhancement.…”

Evaluasi Tiga Jenis Algoritme Berbasis Pembelajaran Mesin untuk Klasifikasi Jenis Tumor Payudara

“…Sedangkan metode mammography memiliki akurasi 68%--79% [6]. MRI memiliki tingkat akurasi 70% untuk mendeteksi benign (tumor jinak payudara) dan 92% untuk malignant (tumor ganas payudara) [7]. Sedangkan FNA memiliki range akurasi yang lebih luas, yaitu 65%--98% [8].…”