Spaces:
Sleeping
Sleeping
File size: 13,298 Bytes
f8a22ea | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 | import gradio as gr
import numpy as np
import plotly.graph_objects as go
from pathlib import Path
from scipy.integrate import trapezoid
import scipy.signal as signal
# =============================================================================
# CONFIGURATION
# =============================================================================
OUTPUT_IEEG = Path("consolidated_ieeg.npz")
OUTPUT_LCMV = Path("consolidated_lcmv.npz")
RUN_MAP = {"c": "eyes_closed", "o": "eyes_open", "l": "left_hand", "r": "right_hand"}
# PSD Parameters
SFREQ_DEFAULT = 500.0
PSD_WINDOW_SEC = 2.0
FMAX = 50
FREQ_BANDS = {
'Delta': (1, 4), 'Theta': (4, 8), 'Alpha': (8, 12),
'Low_Beta': (12, 20), 'High_Beta': (20, 30),
'Low_Gamma': (30, 50), 'High_Gamma': (50, 100),
}
# Patterns
STN_PATTERNS = ["STN-L", "STN-R", "STN_L", "STN_R", "Left-STN", "Right-STN"]
GPI_PATTERNS = ["GPi-L", "GPi-R", "GPi_L", "GPi_R", "pGP-lh", "pGP-rh", "L-GPi", "R-GPi", "GPI-L", "GPI-R"]
M1_L_PATTERNS = ["ECOG-8-9-L", "ECOG-10-11-L", "M1-L", "Left-M1"]
M1_R_PATTERNS = ["ECOG-8-9-R", "ECOG-10-11-R", "M1-R", "Right-M1"]
ATLAS_LABELS = {
"STN": "STN (DiFuMo-223)",
"L_GPi": "L-GPi (GT pGP-lh)",
"R_GPi": "R-GPi (GT pGP-rh)",
}
COLORS = {
"IEEG": "#1f77b4",
"LCMV": "#d62728",
"STN": "#ff7f0e",
"L_GPi": "#2ca02c",
"R_GPi": "#9467bd",
}
# Global Data Handles
ALL_IEEG_DATA = None
ALL_LCMV_DATA = None
# =============================================================================
# CORE LOGIC
# =============================================================================
def compute_psd(time_series, sfreq=SFREQ_DEFAULT, fmax=FMAX):
ts = np.real(time_series).astype(np.float64)
window_size = int(PSD_WINDOW_SEC * sfreq)
if len(ts) < window_size:
window_size = max(int(len(ts)*0.8), 100)
nyq = sfreq * 0.5
if nyq <= 0.5: nyq = 0.51
b, a = signal.butter(4, 0.5 / nyq, btype='high')
filtered = signal.filtfilt(b, a, ts)
freqs, psd = signal.welch(filtered, fs=sfreq, window='hann', nperseg=window_size,
noverlap=window_size // 2, detrend='constant')
mask = (freqs >= 1.0) & (freqs <= fmax)
freqs, psd = freqs[mask], psd[mask]
if len(freqs) == 0:
return np.array([1, 10]), np.log10(np.array([1e-10, 1e-10]) + 1e-12)
psd_log = np.log10(psd + 1e-12)
return freqs.astype(np.float32), psd_log.astype(np.float32)
def load_data():
global ALL_IEEG_DATA, ALL_LCMV_DATA
if ALL_IEEG_DATA is None or ALL_LCMV_DATA is None:
if not OUTPUT_IEEG.exists() or not OUTPUT_LCMV.exists():
raise FileNotFoundError("Consolidated files missing. Please run consolidation first.")
ALL_IEEG_DATA = np.load(OUTPUT_IEEG, allow_pickle=True)
ALL_LCMV_DATA = np.load(OUTPUT_LCMV, allow_pickle=True)
def get_consolidated_ieeg(subj_id, run_code):
global ALL_IEEG_DATA
meta_key = f"meta_{subj_id}_{run_code}"
if meta_key not in ALL_IEEG_DATA.files:
return None, None
meta = ALL_IEEG_DATA[meta_key].item()
channels = {}
prefix = f"{subj_id}_{run_code}_"
for key in ALL_IEEG_DATA.files:
if key.startswith(prefix) and key != meta_key:
channels[key.replace(prefix, "")] = ALL_IEEG_DATA[key]
return channels, meta
def get_consolidated_lcmv(subj_id):
global ALL_LCMV_DATA
meta_key = f"meta_{subj_id}"
if meta_key not in ALL_LCMV_DATA.files:
return None, None
meta = ALL_LCMV_DATA[meta_key].item()
rois = {}
prefix = f"{subj_id}_"
for key in ALL_LCMV_DATA.files:
if key.startswith(prefix) and key != meta_key:
rois[key.replace(prefix, "")] = ALL_LCMV_DATA[key]
return rois, meta
def find_channel(channels_dict, patterns):
if channels_dict is None:
return None, None
for pattern in patterns:
if pattern in channels_dict:
return pattern, channels_dict[pattern]
for key in channels_dict.keys():
if pattern.lower() in key.lower():
return key, channels_dict[key]
return None, None
def create_interactive_plot(roi_name, ieeg_signal, ieeg_sfreq, ch_used,
source_signal, source_sfreq, source_label, source_color,
subject_id, run_id):
freqs_ieeg, psd_ieeg = compute_psd(ieeg_signal, sfreq=ieeg_sfreq)
freqs_src, psd_src = compute_psd(source_signal, sfreq=source_sfreq)
fig = go.Figure()
fig.add_trace(go.Scatter(
x=freqs_ieeg, y=psd_ieeg,
mode='lines', name=f'iEEG ({ch_used})',
line=dict(color=COLORS["IEEG"], width=3),
hovertemplate=f'<b>iEEG</b><br>Freq: %{{x:.2f}} Hz<br>PSD: %{{y:.2f}}<extra></extra>'
))
fig.add_trace(go.Scatter(
x=freqs_src, y=psd_src,
mode='lines', name=source_label,
line=dict(color=source_color, width=3, dash='dash'),
hovertemplate=f'<b>{source_label}</b><br>Freq: %{{x:.2f}} Hz<br>PSD: %{{y:.2f}}<extra></extra>'
))
shapes = []
n_bands = len(FREQ_BANDS)
band_colors = [f"rgba(31, 119, 180, {0.1 + (i/n_bands)*0.2})" for i in range(n_bands)]
for i, (band, (fmin, fmax)) in enumerate(FREQ_BANDS.items()):
band_low = max(fmin, min(freqs_ieeg))
band_high = min(fmax, max(freqs_ieeg))
if band_low < band_high:
shapes.append(dict(
type="rect", xref="x", yref="paper",
x0=band_low, x1=band_high, y0=0, y1=1,
fillcolor=band_colors[i], opacity=0.3, layer="below", line_width=0
))
title_text = f"{subject_id} | Run: {run_id} | ROI: {roi_name}<br><sup>{source_label} vs iEEG</sup>"
fig.update_layout(
title=dict(text=title_text, font=dict(size=14, family="Arial")),
xaxis_title="Frequency (Hz)",
yaxis_title="PSD (logโโ)",
xaxis=dict(range=[1, FMAX], type="linear"),
yaxis_type="linear",
hovermode="x unified",
legend=dict(x=0, y=1, bgcolor="rgba(255,255,255,0.8)"),
shapes=shapes,
template="plotly_white",
height=600,
margin=dict(l=50, r=50, t=60, b=50)
)
return fig
def generate_all_plots(subj_id, run_code):
"""Generates all valid plots for a subject/run and returns a dictionary."""
try:
load_data()
except FileNotFoundError as e:
return {}, str(e)
cond = RUN_MAP.get(run_code, "unknown")
ieeg_ch, ieeg_meta = get_consolidated_ieeg(subj_id, run_code)
lcmv_rois, lcmv_meta = get_consolidated_lcmv(subj_id)
plots_dict = {}
logs = [f"Processing {subj_id} | Condition: {cond}"]
if ieeg_ch is None or lcmv_rois is None:
return plots_dict, f"No data found for {subj_id} (Run: {run_code})."
ieeg_sfreq = ieeg_meta.get('sfreq', SFREQ_DEFAULT)
lcmv_sfreq = lcmv_meta.get('sfreq', SFREQ_DEFAULT)
# Detect Electrodes
stn_l_ch, stn_l_sig = find_channel(ieeg_ch, STN_PATTERNS)
stn_r_ch, stn_r_sig = find_channel(ieeg_ch, [p.replace("-L","-R").replace("_L","_R") for p in STN_PATTERNS])
gpi_l_ch, gpi_l_sig = find_channel(ieeg_ch, GPI_PATTERNS)
gpi_r_ch, gpi_r_sig = None, None
if gpi_l_ch:
right_patterns = [gpi_l_ch.replace("L","R").replace("l","r").replace("lh","rh")]
right_patterns.extend([p.replace("-L","-R").replace("_L","_R") for p in GPI_PATTERNS])
gpi_r_ch, gpi_r_sig = find_channel(ieeg_ch, right_patterns)
m1_l_ch, m1_l_sig = find_channel(ieeg_ch, M1_L_PATTERNS)
m1_r_ch, m1_r_sig = find_channel(ieeg_ch, M1_R_PATTERNS)
def add_plot(name, sig, ch, roi_key, label, color):
if sig is not None and ch is not None and roi_key in lcmv_rois:
fig = create_interactive_plot(name, sig, ieeg_sfreq, ch, lcmv_rois[roi_key], lcmv_sfreq, label, color, subj_id, run_code)
key = f"{name} vs {label}"
plots_dict[key] = fig
logs.append(f"โ
Found: {key}")
# M1
add_plot("L_M1", m1_l_sig, m1_l_ch, f"L_M1_{cond}", "LCMV MNI voxel", COLORS["LCMV"])
add_plot("R_M1", m1_r_sig, m1_r_ch, f"R_M1_{cond}", "LCMV MNI voxel", COLORS["LCMV"])
# STN
if stn_l_sig is not None:
add_plot("L_STN", stn_l_sig, stn_l_ch, f"L_STN_{cond}", "LCMV MNI voxel", COLORS["LCMV"])
if f"STN_{cond}" in lcmv_rois:
add_plot("L_STN", stn_l_sig, stn_l_ch, f"STN_{cond}", ATLAS_LABELS["STN"], COLORS["STN"])
if stn_r_sig is not None:
add_plot("R_STN", stn_r_sig, stn_r_ch, f"R_STN_{cond}", "LCMV MNI voxel", COLORS["LCMV"])
if f"STN_{cond}" in lcmv_rois:
add_plot("R_STN", stn_r_sig, stn_r_ch, f"STN_{cond}", ATLAS_LABELS["STN"], COLORS["STN"])
# GPi (Fallback)
if gpi_l_sig is not None and stn_l_sig is None:
add_plot("L_GPi", gpi_l_sig, gpi_l_ch, f"L_GPi_{cond}", "LCMV MNI voxel (GPi)", COLORS["LCMV"])
if f"L_GPi_{cond}" in lcmv_rois:
add_plot("L_GPi", gpi_l_sig, gpi_l_ch, f"L_GPi_{cond}", ATLAS_LABELS["L_GPi"], COLORS["L_GPi"])
if gpi_r_sig is not None and stn_r_sig is None:
add_plot("R_GPi", gpi_r_sig, gpi_r_ch, f"R_GPi_{cond}", "LCMV MNI voxel (GPi)", COLORS["LCMV"])
if f"R_GPi_{cond}" in lcmv_rois:
add_plot("R_GPi", gpi_r_sig, gpi_r_ch, f"R_GPi_{cond}", ATLAS_LABELS["R_GPi"], COLORS["R_GPi"])
if not plots_dict:
logs.append("โ ๏ธ No matching electrode/ROI pairs found.")
return plots_dict, "\n".join(logs)
def get_available_subjects():
if not OUTPUT_LCMV.exists():
return []
data = np.load(OUTPUT_LCMV, allow_pickle=True)
subjects = set()
for key in data.files:
if key.startswith("meta_"):
subjects.add(key.replace("meta_", ""))
return sorted(list(subjects))
# =============================================================================
# GRADIO INTERFACE
# =============================================================================
# Note: 'theme' parameter removed from constructor for Gradio 5.0+ compatibility
with gr.Blocks(title="Interactive iEEG-LCMV Viewer") as demo:
gr.Markdown("# Interactive iEEG & LCMV Viewer")
gr.Markdown("Select a subject and condition to generate available comparisons. Then choose specific plots to visualize.")
# State to store generated plots for the current selection
current_plots_state = gr.State({})
with gr.Row():
with gr.Column(scale=1):
gr.Markdown("### 1. Select Data")
btn_refresh = gr.Button("๐ Refresh Subjects")
subject_dropdown = gr.Dropdown(label="Subject", choices=[], interactive=True)
run_dropdown = gr.Dropdown(
label="Condition",
choices=["c", "o", "l", "r"],
value="c",
info="c: Eyes Closed, o: Eyes Open, l: Left Hand, r: Right Hand"
)
btn_generate = gr.Button("๐ Find Available Plots", variant="primary")
gr.Markdown("### 2. Choose Visualization")
plot_selector = gr.Dropdown(label="Select Plot to View", choices=[], interactive=True)
gr.Markdown("### Log")
val_log = gr.Textbox(label="Status", lines=6, interactive=False)
with gr.Column(scale=3):
gr.Markdown("### PSD Comparison")
plot_display = gr.Plot(label="Interactive Plot", show_label=False)
# Event Handlers
def refresh_subjects():
subs = get_available_subjects()
return gr.Dropdown(choices=subs, value=subs[0] if subs else None)
def process_and_update_dropdown(subj, run):
"""Generates plots, updates state, log, dropdown options, and shows the first plot."""
if not subj:
return {}, "Please select a subject.", gr.Dropdown(choices=[], value=None), None
plots_dict, log_msg = generate_all_plots(subj, run)
choices = list(plots_dict.keys())
if not choices:
return plots_dict, log_msg, gr.Dropdown(choices=[], value=None), None
initial_val = choices[0]
initial_fig = plots_dict[initial_val]
return plots_dict, log_msg, gr.Dropdown(choices=choices, value=initial_val), initial_fig
def on_plot_selection(plots_dict, selected_key):
"""Updates only the plot when dropdown changes."""
if not plots_dict or not selected_key:
return None
return plots_dict.get(selected_key)
# Wire up events
btn_refresh.click(fn=refresh_subjects, inputs=[], outputs=[subject_dropdown])
demo.load(fn=refresh_subjects, inputs=[], outputs=[subject_dropdown])
# When Generate is clicked: Update State, Log, Dropdown, AND Plot
btn_generate.click(
fn=process_and_update_dropdown,
inputs=[subject_dropdown, run_dropdown],
outputs=[current_plots_state, val_log, plot_selector, plot_display]
)
# When Dropdown changes: Update Plot Display only
plot_selector.change(
fn=on_plot_selection,
inputs=[current_plots_state, plot_selector],
outputs=[plot_display]
)
if __name__ == "__main__":
# Note: 'theme' parameter moved to launch() for Gradio 5.0+
demo.launch(theme=gr.themes.Soft()) |