Background
Methods
Study cohort and image acquisition
Data analysis
Algorithm overview
CSF dynamics quantification
Peak flow latency calculation
Ventricle segmentation
Patient outcome
Statistical methods
Results
Hydrodynamic results pre-shunt NPH and control group
Area (mm2) | Caudal max velocity (cm/s) | Cranial max velocity (cm/s) | Caudal mean flow (mL/min) | Cranial mean flow (mL/min) | SV (μL) | |
---|---|---|---|---|---|---|
Aqueduct | ||||||
Pre-shunt (n = 26) | 8.0 [3.4]** (5.1–13.7) | 12.8 [8.3]* (5.3–21.4) | 8.4 [5.8]* (4.3–17.0) | 0.32 [0.23]** (0.12–0.73) | 0.26 [0.18]** (0.09–0.57) | 124.5 [94.5]** (37.1–275.0) |
Control (n = 10) | 4.9 [1.7] (3.1–7.4) | 7.9 [4.5] (5.3–15.0) | 5.9 [2.2] (3.8–8.2) | 0.12 [0.07] (0.06–0.19) | 0.09 [0.04] (0.05–0.15) | 49.7 [32.3] (23.1–90.5) |
Cistern | ||||||
Pre-shunt (n = 21) | 58.6 [63.8] (20.1–255.7) | 6.3 [3.5] (2.7–14.1) | 4.9 [3.5] (1.9–12.4) | 0.78 [0.31] (0.3–2.0) | 0.43 [0.31] (0.14–1.47) | 293.5 [157.6] (111.9–683.6) |
Control (n = 8) | 69.6 [22.0] (31.1–92.2) | 5.7 [1.8] (3.1–7.5) | 4.3 [1.9] (2.1–8.3) | 0.73 [0.35] (0.4–0.8) | 0.41 [0.22] (0.27–0.63) | 299.0 [171.5] (181.2–418.7) |
Impact of shunt treatment on hydrodynamic metrics
Area (mm2) | Caudal max velocity (cm/s) | Cranial max velocity (cm/s) | Caudal mean flow (mL/min) | Cranial mean flow (mL/min) | SV (μL) | |
---|---|---|---|---|---|---|
Aqueduct | ||||||
Pre-shunt (n = 9) | 8.2 [3.3]* (5.1–16.4) | 12.5 [9.7] (5.3–21.4) | 9.5 [5.6] (4.5–13.0) | 0.42 [0.33]** (0.12–0.70) | 0.31 [0.23]** (0.09–0.46) | 144.6 [127.6]** (37.1–249.6) |
Post-shunt (n = 9) | 7.8 [4.3] (3.1–9.0) | 10.9 [4.5] (8.4–15.0) | 6.6 [2.6] (5.4–11.9) | 0.19 [0.17] (0.10–0.36) | 0.15 [0.13] (0.09–0.39) | 76.8 [55.1] (39.4–177.6) |
Cistern | ||||||
Pre-shunt (n = 6) | 58.0 [80.1] (45.2–255.7) | 5.4 [4.1] (2.7–9.9) | 4.7 [2.1] (3.0–9.3) | 0.83 [0.30] (0.45–1.11) | 0.41 [0.31] (0.29–0.89) | 297.4 [99.4] (188.3–486.5) |
Post-shunt (n = 6) | 54.3 [78.7] (36.6–161.1) | 5.1 [4.8] (1.9–11.9) | 4.7 [5.2] (1.9–14.0) | 0.82 [0.64] (0.2–1.29) | 0.46 [0.41] (0.15–0.97) | 346.5 [309.3] (99.3–576.0) |
Peak flow latency
Ventricle volume
Stroke volume ratio
Patient outcome
Discussion
Impact of shunting
Peak flow latency
Percentage of low-correlated voxels removed from the ROI | |||||||
---|---|---|---|---|---|---|---|
Degree of fit | 1% | 10% | 25% | 50% | 75% | 90% | 95% |
No poly fit | 0.076 | 0.076 | 0.061 | 0.107 | 0.040 | 0.094 | 0.061 |
4 | 0.007 | 0.009 | 0.009 | 0.009 | 0.009 | 0.007 | 0.004 |
5 | 0.009 | 0.009 | 0.011 | 0.009 | 0.011 | 0.007 | 0.015 |
6 | 0.009 | 0.007 | 0.009 | 0.009 | 0.009* | 0.007 | 0.015 |
7 | 0.023 | 0.014 | 0.014 | 0.025 | 0.023 | 0.009 | 0.036 |
8 | 0.044 | 0.036 | 0.044 | 0.044 | 0.029 | 0.009 | 0.067 |
9 | 0.036 | 0.036 | 0.044 | 0.040 | 0.029 | 0.011 | 0.067 |
10 | 0.066 | 0.055 | 0.036 | 0.044 | 0.036 | 0.011 | 0.116 |