Convolutional neural network algorithm trained on lumbar spine radiographs to predict outcomes of transforaminal epidural steroid injection for lumbosacral radicular pain from spinal stenosis

Table 2 TFESI model detail.

Layer (type)	Output Shape	Param #
anteroposterior img (InputLayer)	[(None, 384, 384, 3)]	0
lateral img (InputLayer)	[(None, 384, 384, 3)]	0
left oblique img (InputLayer)	[(None, 384, 384, 3)]	0
right oblique img (InputLayer)	[(None, 384, 384, 3)]	0
efficientnetv2-s (Functional)	(None, None, None, 1280)	2,033,136
global_average_pooling2d (GlobalAveragePooling2D)	(None, 1280)	0
global_average_pooling2d_1 (GlobalAveragePooling2D)	(None, 1280)	0
global_average_pooling2d_2 (GlobalAveragePooling2D)	(None, 1280)	0
global_average_pooling2d_3 (GlobalAveragePooling2D)	(None, 1280)	0
concatenate (Concatenate)	(None, 5120)	0
dense (Dense)	(None, 1024)	5,243,904
batch_normalization (BatchNormalization)	(None, 1024)	4,096
Dropout (Dropout)	(None, 1024)	0
Dense_1 (Dense)	(None, 1024)	1,049,600
batch_normalization_1 (BatchNormalization)	(None, 1024)	4,096
dense_2 (Dense)	(None, 1)	1,025
Total params: 26,634,081 (101.60 MB) Trainable params: 26,476,113 (101.00 MB) Non-trainable params: 157,968 (617.06 KB)