Architectural distortion is among the non-palpable findings associated with early-stage breast cancer. Early detection improves the prognosis of the patient. In breast MRI, which is employed as a problem-solving method, evaluating more parameters based on evidence with the Kaiser score provides greater clinical guidance. In this study, the diagnostic accuracy of the Kaiser score was determined to be 78.05%. The area under the curve (AUC) was calculated to be 0.833 according to the ROC analysis, as shown in Fig. 2. This indicates a reasonably good discriminatory power for the diagnostic accuracy of the Kaiser score.
Wang et al. reported a high diagnostic accuracy of the Kaiser score in their study, with an area under the curve (AUC) of 0.958 [7]. We speculate that the disparity in findings could be attributed to our study's focus solely on architectural distortions and the exclusion of lesions classified in the BI-RADS5 category.
In addition, it is believed that the difference in accuracy observed in our study compared to the literature may be attributed to the inclusion of high-risk breast lesions (ADH, RS, LCIS, etc.) with a Kaiser score above 5 in the benign group. However, it has been observed that surgical excision is appropriate in patients diagnosed with high-risk breast lesions due to the accompanying risk of malignancy and the discordance between radiology and pathology findings.
There is no other study in the literature that specifically evaluates the Kaiser score in relation to only architectural distortions and high-risk breast lesions.
While the Kaiser score has not been incorporated into the flowchart yet, it is suggested that including diffusion-weighted imaging findings in the scoring system would offer an additional quantitative advantage. According to the scoring system, lesions with a Kaiser score above 4 can be downgraded to 4 points if they exhibit high apparent diffusion coefficient (ADC) values [6]. Nevertheless, if the quality of the ADC map is deemed to be low, it is recommended to exclude ADC measurements from the scoring system to avoid the potential risk of false negativity. Due to the limited availability of DWI in our patient cohort, ADC values were not included in the analysis. However, in two of our patients who were diagnosed with high-risk lesions (ADH and RS), DWI was performed, and high signal intensity was observed in the corresponding area on ADC maps. When their Kaiser scores were downgraded accordingly, they dropped below 5, and it was observed that the utilization of DWI findings would further enhance diagnostic accuracy.
In addition, there is a suggestion to add 2 points to the scoring system when a low Kaiser score is observed in low-grade ductal carcinoma in situ cases accompanied by suspicious microcalcifications [6]. Wengert et al. demonstrated that the use of the Kaiser score in suspicious microcalcifications can lead to a reduction of unnecessary biopsies in 63.5% of cases [12].
There is currently no specific study available on the MRI Kaiser Score specifically focused on high-risk breast lesions. By incorporating the Kaiser score into a radiomics model, it is possible to develop a predictive model to estimate the probability of malignancy in high-risk breast lesions. Performing radiological follow-up in appropriate patients with high-risk breast lesions, who have a low Kaiser score and radiology-pathology compatibility, can help reduce unnecessary surgeries.
Marino et al. demonstrated that the variability between the Kaiser scoring system and the readers, which arises from differences in reader experience, can be minimized [13]. Furthermore, since the parameters utilized in the scoring system are independent of the imaging protocols, it proves to be a valuable tool in daily practice, particularly during radiology training and the early stages of breast imaging implementation.
The most significant limitations of this study are its retrospective design and the relatively small number of patients. Furthermore, due to the absence of DWI in some patients, ADC maps could not be included in the statistical analysis. More comprehensive results can be obtained through future studies that incorporate DWI, the Kaiser score, and artificial intelligence models, employing prospective planning with larger patient cohorts.
The combined utilization of mammography findings and the evidence-based MRI Kaiser score in suspected architectural distortions provides more accurate results in the differential diagnosis of breast cancer.
The Kaiser score can serve as a valuable tool in treatment decision-making and follow-up planning, helping to mitigate potential discrepancies arising from varying levels of experience.