Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

< Back to Article

Fig 1.

A T1-weighted (left) and a T2-weighted (right) traverse MRI images of the L3/L4 Intervertebral Disc of a patient are shown.

One marked difference in the two images is the cerebrospinal fluid (CSF) in the spinal canal that appears black on the T1-weighted image but as a brighter region on the T2-weighted image because of its low fat contents.

More »

Fig 1 Expand

Fig 2.

A mid-sagittal view of a lumbar spine MRI showing the intersection lines between the sagittal plane and the traverse planes that are shown in Fig 3.

The lines marked in red are the intersection lines of traverse planes that cut closest to the half-height of an IVD.

More »

Fig 2 Expand

Fig 3.

An example of nine traverse images of a lumbar spine.

Image 2, 5, and 8 are from the planes that cut closest to the half-height of L3/L4, L4/L5, and L5/S1 IVD, respectively.

More »

Fig 3 Expand

Fig 4.

A flowchart describing the traditional Transfer Learning approach of using Deep Convolutional Neural Network for medical image classification, where a) depicts the training process and b) depicts the inference step.

More »

Fig 4 Expand

Table 1.

The range of acquisition parameter values used during traverse MRI scans.

More »

Table 1 Expand

Fig 5.

Two example cases where the image registration process succeeded (left column) and failed (right column).

The top row shows the T1-weighted images, the middle row shows T2-weighted images, and the bottom row shows the resulting composite images after image registration.

More »

Fig 5 Expand

Table 2.

Dataset sizes (number of images).

More »

Table 2 Expand

Fig 6.

A flowchart describing the methodology used where a) depicts the feature extraction step, b) depicts the DR and FS modeling step, c) depicts the ML and FC training step and d) depicts the inference/classification step.

More »

Fig 6 Expand

Table 3.

The list of the DCNNs used and the summary of their architecture.

More »

Table 3 Expand

Fig 7.

A tree diagram depicting the method combination used in this study.

More »

Fig 7 Expand

Table 4.

The description of the hyperparameter optimization options and range of values for each ML learner.

More »

Table 4 Expand

Table 5.

The description of the hyperparameter optimization options and values for each FC learning optimization algorithm.

More »

Table 5 Expand

Table 6.

The feature length of each DCNN model and DR/FS method combination.

More »

Table 6 Expand

Table 7.

The best ML learner for each combination method and its average classification performance (using Accuracy metric).

More »

Table 7 Expand

Table 8.

The best ML learner for each combination method and its average classification performance (using Precision metric).

More »

Table 8 Expand

Table 9.

The best ML learner for each combination method and its average classification performance (using Recall metric).

More »

Table 9 Expand

Table 10.

The best ML learner for each combination method and its average classification performance (using F1-Score metric).

More »

Table 10 Expand

Table 11.

The best FC learning optimizer for each combination method and its average classification performance (using Accuracy metric).

More »

Table 11 Expand

Table 12.

The best FC learning optimizer for each combination method and its average classification performance (using Precision metric).

More »

Table 12 Expand

Table 13.

The best FC learning optimizer for each combination method and its average classification performance (using Recall metric).

More »

Table 13 Expand

Table 14.

The best FC learning optimizer for each combination method and its average classification performance (using F1-Score metric).

More »

Table 14 Expand

Table 15.

Summary of the classification performance of ML learners using features from 11 DCNNs.

More »

Table 15 Expand

Table 16.

Summary of the classification performance of FC neural networks using features from 11 DCNNs.

More »

Table 16 Expand

Table 17.

Summary of classification performance of ML methods for each DCNN.

More »

Table 17 Expand

Fig 8.

The Accuracy of DR/FS/FL-ML methods using DenseNet201 features.

The ML learners used can be found in the second row of Table 7.

More »

Fig 8 Expand

Fig 9.

The Precision of DR/FS/FL-ML methods using DenseNet201 features.

The ML learners used can be found in the second row of Table 8.

More »

Fig 9 Expand

Fig 10.

The Recall of DR/FS/FL-ML methods using DenseNet201 features.

The ML learners used can be found in the second row of Table 9.

More »

Fig 10 Expand

Fig 11.

The F1-Score of DR/FS/FL-ML methods using DenseNet201 features.

The ML learners used can be found in the second row of Table 10.

More »

Fig 11 Expand

Fig 12.

Per-class classification performance using Fine Gaussian SVM classifier on full-length DenseNet201 features (Precision).

More »

Fig 12 Expand

Fig 13.

Per-class classification performance using Fine Gaussian SVM classifier on full-length DenseNet201 features (Recall).

More »

Fig 13 Expand

Fig 14.

Per-class classification performance using Fine Gaussian SVM classifier on full-length DenseNet201 features (F1-Score).

More »

Fig 14 Expand