Skip to main content

Table 4 Kruskal–Wallis tests were performed to assess the relationships between patient and lesion characteristics, and performance metrics

From: Stratified assessment of an FDA-cleared deep learning algorithm for automated detection and contouring of metastatic brain tumors in stereotactic radiosurgery

Characteristics

p-Value

Lesion-wise DSC

Lesion-wise AVD [mm (%)]

FP Count

Sensitivity (%)

Patient level

Sex

0.546

0.771 (0.726)

0.459

0.146

Age

0.118

0.899 (0.226)

0.838

0.566

Race

0.529

0.618 (0.418)

0.433

0.446

Primary histology

0.186

0.616 (0.160)

0.519

0.613

Tumor count

0.085

0.207 (0.858)

0.835

0.030

Lesion level

Tumor size

1.64e−31

0.484 (2.54e−29)*

N/A**

5.05e−09

  1. *AVD in terms of absolute distance (mm) is not correlated with tumor size. There is a correlation between relative AVD (%) and tumor size, because relative AVD decreases as tumor size increases
  2. **Statistical result cannot be assessed because there is no ground-truth size for false positives