Skip to main content
Fig. 3 | Chinese Neurosurgical Journal

Fig. 3

From: A deep-learning system to help make the surgical planning of coil embolization for unruptured intracranial aneurysms

Fig. 3

Establishment of a DL model for the diameter of first coils. A The flowchart of establishing NFM model for the diameter of first coils. Patients in the derivation set were grouped as the training set (80%, n = 90) and testing set (20%, n = 23). Based on the training set, a model for the diameter of first coils was trained. The accuracy of this model was validated based on the testing set. B The performance of model for recommending the diameter of first coils within the training set. The absolute match ratio was 67.8% (61/90), and the general match ratio was 91.1% (82/90). C The scatter dot plots present the correlation of diameter of coil actually used and coil predicted by model within the training set. D The performance of model for recommending the diameter of first coils within the training set. The absolute match ratio was 78.2% (18/23), and the general match ratio was 91.3% (21/23). E The scatter dot plots present the correlation of diameter of coil actually used and coil predicted by model within the testing set. NFM, neural factorization machines

Back to article page