Cancer-related fatigue classification based on heart rate variability signals from wearables

FIGURE 1. Cancer-related fatigue analysis procedure.

FIGURE 2. Plot of data distribution between BFI and (LF/HF) disorder ratio in sleep phase; circles represent cases with mild fatigue (BFI < 4), and triangles represent cases with moderate fatigue (BFI ≥ 4). The dotted line illustrates a boundary line with cutoff point, LHDS = 0.5.

FIGURE 3. Plot of data distribution between BFI and average (LF/HF) ratio in sleep phase; circles represent cases with mild fatigue (BFI < 4), and triangles represent cases with moderate fatigue (BFI ≥ 4). The dotted line illustrates a boundary line with cutoff point, LHP = 1.9.

FIGURE 4. Plot of data distribution between BFI and (LF/HF) disorder ratio in active phase; circles represent cases with mild fatigue (BFI < 4), and triangles represent cases with moderate fatigue (BFI ≥ 4). The dotted line illustrates a boundary line with cutoff point, LHDA = 0.2.

FIGURE 5. Classification ratio and p-value obtained from t-test between mild and moderate categories for varied (LF/HF) disorder ratio in sleep phase. “Acc” stands for classification accuracy.

FIGURE 6. Classification ratio and p-value obtained from t-test between mild and moderate categories for varied average (LF/HF) ratio in sleeping phase. “Acc” stands for classification accuracy.

FIGURE 7. Classification ratio and p-value obtained from t-test between mild and moderate categories for varied (LF/HF) disorder ratio in the active phase. “Acc” stands for classification accuracy.