MedtronicProject/model.py at model · SamanyuDixit/MedtronicProject · GitHub

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
import joblib
import pandas as pd
from sklearn.preprocessing import MinMaxScaler
from sklearn.model_selection import train_test_split
import tensorflow as tf
from tensorflow import keras
from tensorflow.keras import layers
import matplotlib.pyplot as plt
from sklearn.metrics import mean_squared_error, r2_score


dataset = pd.read_csv('medtronic_data.csv')

dataset = dataset.drop(columns=['patient_id', 'cadence_0_weeks', 'cadence_8_weeks', 'cadence_24_weeks', 'cadence_52_weeks', 'step_length_0_weeks', 'step_length_8_weeks',
                                'step_length_24_weeks', 'step_length_52_weeks', 'step_width_0_weeks', 'step_width_8_weeks', 'step_width_24_weeks', 'step_width_52_weeks'])
dataset.to_csv('anon_medtronic_data.csv', index=False)


parameters = ['knee_ext_1_wk', 'knee_ext_2_wk', 'knee_ext_4_wk', 'knee_ext_8_wk', 'knee_ext_12_wk', 'knee_ext_24_wk',
              'knee_flx_1_wk', 'knee_flx_2_wk', 'knee_flx_4_wk', 'knee_flx_8_wk', 'knee_flx_12_wk', 'knee_flx_24_wk',
              'kin_180_acl_recon_4_wk', 'kin_180_acl_recon_8_wk', 'kin_180_acl_recon_12_wk', 'kin_180_acl_recon_24_wk',
              'kin_60_acl_recon_4_wk', 'kin_60_acl_recon_8_wk', 'kin_60_acl_recon_12_wk', 'kin_60_acl_recon_24_wk']


missing_columns = [col for col in parameters if col not in dataset.columns]
if missing_columns:
    print(f"Missing columns in the dataset: {missing_columns}")
else:

    scaler = MinMaxScaler()
    normalized_parameters = scaler.fit_transform(dataset[parameters])

    joblib.dump(scaler, 'scaler.pkl')


    normalized_dataset = pd.DataFrame(normalized_parameters, columns=parameters)


    weights = {
        'knee_ext_1_wk': 0.125, 'knee_ext_2_wk': 0.06328125, 'knee_ext_4_wk': 0.03203125, 'knee_ext_8_wk': 0.01640625, 'knee_ext_12_wk': 0.00859375, 'knee_ext_24_wk': 0.0046875,
        'knee_flx_1_wk': 0.125, 'knee_flx_2_wk': 0.06328125, 'knee_flx_4_wk': 0.03203125, 'knee_flx_8_wk': 0.01640625, 'knee_flx_12_wk': 0.00859375, 'knee_flx_24_wk': 0.0046875,
        'kin_180_acl_recon_4_wk': 0.0625, 'kin_180_acl_recon_8_wk': 0.0625, 'kin_180_acl_recon_12_wk': 0.0625, 'kin_180_acl_recon_24_wk': 0.0625,
        'kin_60_acl_recon_4_wk': 0.0625, 'kin_60_acl_recon_8_wk': 0.0625, 'kin_60_acl_recon_12_wk': 0.0625, 'kin_60_acl_recon_24_wk': 0.0625
    }


    normalized_dataset['Composite_Score'] = sum(normalized_dataset[col] * weight for col, weight in weights.items())
    dataset['Composite_Score'] = normalized_dataset['Composite_Score']


    X = dataset.drop(columns=['Composite_Score'])
    y = dataset['Composite_Score']


    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)


    model = keras.Sequential([
        layers.Dense(64, activation='relu', input_shape=(X_train.shape[1],)),
        layers.Dense(64, activation='relu'),
        layers.Dense(1)  # Regression output
    ])

    model.compile(optimizer='adam', loss='mean_squared_error', metrics=['mean_absolute_error'])


    history = model.fit(X_train, y_train, epochs=50, validation_split=0.2)


    test_loss, test_mae = model.evaluate(X_test, y_test)
    print(f"Test MAE: {test_mae}")


    model.save('knee_recovery_model.h5')


    # # Plot training & validation loss and MAE
    # plt.figure(figsize=(12, 4))

    # # Plot training & validation loss
    # plt.subplot(1, 2, 1)
    # plt.plot(history.history['loss'], label='Training Loss')
    # plt.plot(history.history['val_loss'], label='Validation Loss')
    # plt.title('Model Loss')
    # plt.ylabel('Loss')
    # plt.xlabel('Epoch')
    # plt.legend(loc='upper right')

    # # Plot training & validation MAE
    # plt.subplot(1, 2, 2)
    # plt.plot(history.history['mean_absolute_error'], label='Training MAE')
    # plt.plot(history.history['val_mean_absolute_error'], label='Validation MAE')
    # plt.title('Model Mean Absolute Error')
    # plt.ylabel('MAE')
    # plt.xlabel('Epoch')
    # plt.legend(loc='upper right')

    # plt.show()

    # Evaluate residuals
    y_pred = model.predict(X_test)
    residuals = y_test - y_pred.flatten()

    # plt.figure(figsize=(10, 5))
    # plt.scatter(y_test, residuals)
    # plt.axhline(0, color='r', linestyle='--')
    # plt.xlabel('Actual Composite Score')
    # plt.ylabel('Residuals')
    # plt.title('Residual Analysis')
    # plt.show()


    mse = mean_squared_error(y_test, y_pred)
    print(f"Mean Squared Error (MSE): {mse}")

    rmse = mean_squared_error(y_test, y_pred, squared=False)
    print(f"Root Mean Squared Error (RMSE): {rmse}")

    r2 = r2_score(y_test, y_pred)
    print(f"R-squared: {r2}")