Spaces:
Paused
Paused
| import numpy as np | |
| import pandas as pd | |
| import matplotlib.pyplot as plt | |
| from sklearn.model_selection import train_test_split | |
| from sklearn.preprocessing import StandardScaler | |
| from sklearn.linear_model import LogisticRegression | |
| import gradio as gr | |
| # --------------------------------------------------------- | |
| # 1. Core Logic & Data Processing (Identical to raw code) | |
| # --------------------------------------------------------- | |
| np.random.seed(42) | |
| num_patients = 1000 | |
| data = { | |
| 'age': np.random.randint(18, 70, size=num_patients), | |
| 'hypertension': np.random.choice([0, 1], size=num_patients, p=[0.75, 0.25]), | |
| 'heart_disease': np.random.choice([0, 1], size=num_patients, p=[0.85, 0.15]), | |
| 'avg_glucose_level': np.round(np.random.uniform(55.0, 280.0, size=num_patients), 1), | |
| 'bmi': np.round(np.random.uniform(18.0, 48.0, size=num_patients), 1), | |
| 'stroke': np.random.choice([0, 1], size=num_patients, p=[0.90, 0.10]) | |
| } | |
| df = pd.DataFrame(data) | |
| X = df.drop(columns=['stroke']) | |
| y = df['stroke'] | |
| X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42, stratify=y) | |
| scaler = StandardScaler() | |
| X_train_scaled = scaler.fit_transform(X_train) | |
| X_test_scaled = scaler.transform(X_test) | |
| model = LogisticRegression() | |
| model.fit(X_train_scaled, y_train) | |
| # Generate Feature Importance Plot | |
| def get_feature_importance_plot(): | |
| importance = model.coef_.flatten() | |
| feature_names = ['Age', 'Hypertension', 'Heart Disease', 'Avg Glucose Level', 'BMI'] | |
| feature_importance = pd.DataFrame({ | |
| 'Clinical Feature': feature_names, | |
| 'Statistical Weight': importance | |
| }) | |
| feature_importance['Absolute Weight'] = feature_importance['Statistical Weight'].abs() | |
| feature_importance = feature_importance.sort_values(by='Absolute Weight', ascending=True) | |
| fig, ax = plt.subplots(figsize=(8, 4)) | |
| ax.barh(feature_importance['Clinical Feature'], feature_importance['Statistical Weight'], color='#1f77b4') | |
| ax.axvline(0, color='black', linestyle='--', linewidth=1) | |
| ax.set_title('Neuro-Vascular Feature Importance (Model Weighting)', fontsize=12, fontweight='bold') | |
| ax.set_xlabel('Predictive Influence (Coefficient Value)', fontsize=10) | |
| ax.set_ylabel('Clinical Metrics', fontsize=10) | |
| ax.grid(axis='x', linestyle=':', alpha=0.6) | |
| plt.tight_layout() | |
| return fig | |
| # --------------------------------------------------------- | |
| # 2. Evaluation / Prediction Handler | |
| # --------------------------------------------------------- | |
| def analyze_patient(patient_age, hypertension, heart_disease, blood_glucose, patient_bmi): | |
| has_hypertension = 1 if hypertension == "Yes" else 0 | |
| has_heart_disease = 1 if heart_disease == "Yes" else 0 | |
| custom_patient = np.array([[patient_age, has_hypertension, has_heart_disease, blood_glucose, patient_bmi]]) | |
| custom_patient_scaled = scaler.transform(custom_patient) | |
| prob = model.predict_proba(custom_patient_scaled)[0][1] * 100 | |
| report = [] | |
| report.append("==========================================================") | |
| report.append(" DYNAMIC CLINICAL ANALYSIS REPORT ") | |
| report.append("==========================================================") | |
| report.append("📋 PATIENT BIOMETRICS PROFILE:") | |
| report.append(f" • Age: {patient_age} years old") | |
| report.append(f" • Hypertension History: {hypertension}") | |
| report.append(f" • Cardiovascular Disease: {heart_disease}") | |
| report.append(f" • Avg Blood Glucose: {blood_glucose} mg/dL") | |
| report.append(f" • Body Mass Index (BMI): {patient_bmi} kg/m²") | |
| report.append("----------------------------------------------------------") | |
| report.append(f"🎯 CALCULATED CEREBROVASCULAR STROKE RISK: {prob:.2f}%") | |
| report.append("----------------------------------------------------------") | |
| if prob >= 15.0: | |
| report.append("🚨 WARNING: ELEVATED VASCULAR RISK CATEGORY") | |
| report.append("👉 DPT Action: Check baseline BP/HR. Avoid prolonged dependent positions or maximum manual resistance.") | |
| report.append("👉 Objective: Focus on submaximal aerobic training, functional mobility, and reactive balance strategy.") | |
| else: | |
| report.append("✅ CLEARANCE: BASELINE STABLE RISK PROFILE") | |
| report.append("👉 DPT Action: Safe to push metabolic capacity and neuroplasticity boundaries.") | |
| report.append("👉 Objective: High-intensity task-specific training and advanced dynamic gait coordination.") | |
| report.append("==========================================================") | |
| fig = get_feature_importance_plot() | |
| return "\n".join(report), fig | |
| # --------------------------------------------------------- | |
| # 3. Gradio Web Interface | |
| # --------------------------------------------------------- | |
| with gr.Blocks(title="DPT Neurological Clinical Appraisal") as demo: | |
| gr.Markdown("# 🧠 DPT Neurological Clinical Appraisal & Stroke Risk Calculator") | |
| gr.Markdown("Adjust patient biometrics below to calculate cerebrovascular risk and generate clinical DPT guidance.") | |
| with gr.Row(): | |
| with gr.Column(): | |
| age_input = gr.Slider(minimum=18, maximum=100, value=67, step=1, label="Patient Age") | |
| htn_input = gr.Radio(choices=["No", "Yes"], value="Yes", label="Hypertension History") | |
| hd_input = gr.Radio(choices=["No", "Yes"], value="No", label="Heart Disease History") | |
| glucose_input = gr.Slider(minimum=50.0, maximum=300.0, value=185.4, step=0.1, label="Avg Blood Glucose Level (mg/dL)") | |
| bmi_input = gr.Slider(minimum=15.0, maximum=50.0, value=31.2, step=0.1, label="Patient BMI (kg/m²)") | |
| submit_btn = gr.Button("Calculate Risk Profile", variant="primary") | |
| with gr.Column(): | |
| output_report = gr.Textbox(label="Clinical Report Output", lines=18, max_lines=25) | |
| output_plot = gr.Plot(label="Model Feature Weights") | |
| submit_btn.click( | |
| fn=analyze_patient, | |
| inputs=[age_input, htn_input, hd_input, glucose_input, bmi_input], | |
| outputs=[output_report, output_plot] | |
| ) | |
| demo.load( | |
| fn=analyze_patient, | |
| inputs=[age_input, htn_input, hd_input, glucose_input, bmi_input], | |
| outputs=[output_report, output_plot] | |
| ) | |
| if __name__ == "__main__": | |
| demo.launch() |