Table 3 Skin lesion management prediction results \(\rm{MGMT}_{\rm{pred, multi}}\) obtained using a multi-modal multi-task model.

From: Predicting the clinical management of skin lesions using deep learning

Management labels

Metrics

Sensitivity

Specificity

Precision

AUROC

Overall accuracy

NONE

0.6500

0.9747

0.7429

0.9225

–

CLNC

0.6071

0.8375

0.5965

0.8065

–

EXC

0.8107

0.6776

0.8008

0.8226

–

Average

0.6893

0.8299

0.7134

0.8505

0.7367

3-Fold cross validation

0.6528 ± 0.0477

0.8215 ± 0.0094

0.7123 ± 0.0114

0.8449 ± 0.0135

0.7301 ± 0.0150

  1. All the prediction models have been trained using all the input data modalities (i.e., clinical image, dermoscopic image, and patient metadata). Mean ± standard deviation reported for all the metrics for the 3-fold cross validation.
Back to article page