CEM DBT Preoperative Breast Cancer Staging Shows Promise
Hear’s a breakdown of the data presented in the table:
Table Overview
The table appears to be comparing the performance of different methods (likely diagnostic or imaging techniques) against pathology results. It focuses on concordance (agreement) with pathology and the ability to detect multifocal disease (disease present in multiple locations).
Columns
column 1 (Purple Background): Describes the metric being measured (e.g., “Concordance with pathology”, “Multifocal disease cases found”).
Columns 2-5 (Light Blue Background): Represent the results for four different methods or approaches. The numbers within these columns are the values for the corresponding metric.
Data Interpretation
Let’s look at the key findings:
Concordance with Pathology: The concordance with pathology increases as you move from left to right across the columns.
Method 1: 58.7% concordance
Method 2: 71.7% concordance
Method 3: 71.7% concordance
Method 4: 80.4% concordance
this suggests that Method 4 is the most accurate in aligning with the “gold standard” of pathology results.
Multifocal Disease Cases Found:
Method 1: 0%
Method 2: 0%
Method 3: 0%
Method 4: 0%
All methods appear to have a 0% detection rate for multifocal disease cases.This could indicate a limitation of all the methods tested, or it could mean that multifocal disease is rare in the population being studied.
AUC (Area Under the Curve): The AUC values also increase from left to right.
Method 1: 0.314
Method 2: 0.636
Method 3: 0.660
* Method 4: 0.811
An AUC of 1 represents perfect discrimination, while an AUC of 0.5 represents random chance. Method 4 has the highest AUC, indicating the best ability to distinguish between positive and negative cases.
overall Conclusion
Based on this data, Method 4 appears to be the most promising approach, demonstrating the highest concordance with pathology and the best discriminatory power (highest AUC). However, all methods struggle to identify multifocal disease cases.