Fig. 1 | Scientific Reports

Fig. 1

From: Convolutional neural network-based classification of craniosynostosis and suture lines from multi-view cranial X-rays

Fig. 1

Illustration of the proposed preprocessing steps. (A) The first row contains AP-view X-rays whereas the second contains lateral-view X-rays. AP-view X-ray preprocessing: (a) input, (b) marker-removed, (c) cropped (scale standardisation), and (d) histogram-equalised (intensity standardisation ) images. Lateral-view X-ray preprocessing: (e) input, (f) marker-removed (the green line shows the detected connecting landmarks for visualisation), (g) cropped (scale and orientation standardisation), and (h) histogram-equalised (intensity standardisation) images. (B) Examples of (i) X-ray markers and (j) their removal process. (C) Head-pose standardisation and cropping.

Back to article page