Table 1 Data overview.

From: Intraoperative discrimination of native meningioma and dura mater by Raman spectroscopy

Meningioma subtype

Patients

Patient’s (n; %)

Measuring points (n; %)

Training set

Validation set

Training set

Validation set

Pure tumor tissue

Transitional meningioma

13

10 (76.9%)

3 (23.1%)

171 (74.7%)

58 (25.3%)

Meningiothelial meningioma

19

15 (78.9%)

4 (21.1%)

197 (78.5%)

54 (21.5%)

Atypical meningioma

4

4 (100%)

0

62 (100%)

0

Fibrous meningioma

6

6 (100%)

0

69 (100%)

0

Meningioma, not otherwise specified

2

2 (100%)

0

20 (100%)

0

Secretory meningioma

1

1 (100%)

0

14 (100%)

0

Subtotal “MGM”

 

23 (88.5%)

3 (11.5%)

336 (85.3%)

58 (14.7%)

Healthy dura mater

22

16 (72.7%)

6 (27.3%)

128 (79.0%)

34 (21.0%0

Tumor infiltration zone

Vide supra

Vide supra

 

371 (80.7%)

89 (19.3%)

Verification sample

1

1

108

Total

   

1268

  1. Three classes were defined: Unequivocal meningioma tumor tissue, healthy dura mater and tumor-infiltrated dura mater samples. The data was split up in a training and validation set. The percentages indicate the corresponding portions of samples and measuring points, subdivided in the different tissue types. Upper table part: Meningioma samples. Middle table part: Origin of the probed healthy dura mater samples. Lower table part: Meningioma-infiltrated dura mater samples. Closing line: Total numbers of patients, samples and measuring points. Notice that, the infiltration zone samples belong to patients that were already considered in the other classes. The kept out surgical-site-simulating verification sample was of atypical meningioma tissue origin and was used for the subsequent classifier mapping study.
  2. Significant values are in bold.