Examinator: Fredrik Kahl, Inst för elektroteknik
This thesis investigated whether a deep learning model could learn features of Chaos, from the Chaos & Clues evaluation protocol, in a given dermatoscopic image data set. A successful result could be of use in a future decision-support system for when dermatologists examine skin lesions for traces of melanoma (a type of skin cancer).
The chosen deep learning model (Inception V3) was trained to recognise four classes related to Chaos. Nearly 5000 anonymous patient data entries was used, provided by the partnering company Gnosco. The data was partitioned into one or two classes depending on the symmetry properties found in the corresponding image annotation. More than twenty different model configurations were run to obtain the results in this thesis.
The results indicate that the chosen model was not capable of learning features of Chaos from the dermoscopic image data-set. Training the model to recognise features of Chaos resulted in an overfit system with low validation accuracy (close to 30%).
The prediction target was changed to contrast the negative results from the Chaos classification. The chosen model was therefore configured to learn two classes, ’melanoma’ and ’nevus’. This prediction target yielded a more positive result as the validation accuracy was close to 85%. However, the corresponding confusion matrix showed that these results are not trustworthy.
It is inconclusive whether the negative results from the Chaos classification stem from the chosen approach or if the data set was insufficient for the task-difficulty. We propose adjustments to the data set for future work which could disclose if the outlined approach is viable or not.