Table 8 Performance criteria with MFCCs feature
From: Performance evaluation of lung sounds classification using deep learning under variable parameters
Parameter combination | Performance criteria (%) | Stage | ||||
|---|---|---|---|---|---|---|
Sensitivity (CI) | Specificity (CI) | Accuracy (CI) | Precision (CI) | F1_score (CI) | ||
Lwin = 64 OP = 25% | 59.3 (57.1–61.5) | 59.1 (56.7–61.1) | 59.2 (56.9–61.3) | 59.2 (56.9–61.3) | 59.2 (57.0–61.4) | Training |
53.3 (50.1–55.2) | 49.5 (47.2–52.0) | 52.3 (49.2–54.7) | 74.7 (67.8–86.1) | 62.2 (57.6–67.3) | Test | |
Lwin = 64 OP = 50% | 62.6 (59.3–64.7) | 62.1 (60.1–65.2) | 62.5 (59.6–64.8) | 86.9 (71.2–88.1) | 72.8 (64.7–74.6) | Training |
55.8 (54.1–58.1) | 49.5 (47.2–51.8) | 54.1 (52.3–56.6) | 74.9 (74.4–79.4) | 64.0 (62.6–67.1) | Test | |
Lwin = 64 OP = 75% | 65.0 (62.9–67.3) | 69.1 (67.3–72.1) | 66.1 (64.2–68.2) | 85.2 (82.1–91.3) | 73.7 (71.2–77.5) | Training |
57.4 (54.8–59.6) | 50.1 (48.2–52.3) | 55.4 (52.6–57.7) | 75.3 (67.9–78.0) | 65.1 (60.7–67.6) | Test | |
Lwin = 128 OP = 75% | 73.7 (70.9–75.6) | 69.9 (67.1–72.3) | 72.2 (69.2–74.4) | 79.0 (72.7–82.7) | 76.2 (71.8–79.0) | Training |
64.3 (62.1–66.5) | 65.0 (62.9–67.4) | 64.5 (62.5–66.7) | 82.1 (62.6–87.7) | 72.1 (62.3–75.6) | Test | |
Lwin = 256 OP = 75% | 76.7 (74.2–78.9) | 73.1 (70.0–75.3) | 75.5 (72.6–77.8) | 85.1 (80.1–87.9) | 80.7 (77.0–83.2) | Training |
72.1 (69.8–74.9) | 64.5 (62.1–66.5) | 69.3 (66.9–72.2) | 77.7 (75.3–82.5) | 74.8 (72.4–78.5) | Test | |