Skin cancer is one of the most common types of cancer and, with its increasing incidence, accurate early diagnosis is crucial to improve prognosis of patients. In the process of visual inspection, dermatologists follow specific dermoscopic algorithms and identify important features to provide a diagnosis. This process can be automated as such characteristics can be extracted by computer vision techniques. Although deep neural networks can extract useful features from digital images for skin lesion classification, performance can be improved by providing additional information. The extracted pseudo-features can be used as input (multimodal) or output (multi-tasking) to train a robust deep learning model. This work investigates the multimodal and multi-tasking techniques for more efficient training, given the single optimization of several related tasks in the latter, and generation of better diagnosis predictions. Additionally, the role of lesion segmentation is also studied. Results show that multi-tasking improves learning of beneficial features which lead to better predictions, and pseudo-features inspired by the ABCD rule provide readily available helpful information about the skin lesion.

Teledermatology has developed rapidly in recent years and is nowadays an essential tool for early diagnosis. In this work, we aim to improve existing Teledermatology processes for skin lesion diagnosis by developing a deep learning approach for risk prioritization with a dataset of retrospective data from referral requests of the Portuguese National Health System. Given the high complexity of this task, we propose a new prioritization pipeline guided and inspired by domain knowledge. We explored automatic lesion segmentation and tested different learning schemes, namely hierarchical classification and curriculum learning approaches, optionally including additional patient metadata. The final priority level prediction can then be obtained by combining predicted diagnosis and a baseline priority level accounting for explicit expert knowledge. In both the differential diagnosis and prioritization branches, lesion segmentation with 30% tolerance for contextual information was shown to improve classification when compared with a flat baseline model trained on original images; furthermore, the addition of patient information was not beneficial for most experiments. Curriculum learning delivered better results than a flat or hierarchical approach. The combination of diagnosis information and a knowledge map, created in collaboration with dermatologists, together with the priority achieved interesting results (best macro F1 of 43.93% for a validated test set), paving the way for new data-centric and knowledge-driven approaches.

No abstract

The ongoing reading process of digital meters is time-consuming and prone to errors, as operators capture images and manually update the system with the new readings. This work proposes to automate this operation through a deep learning-powered solution for universal controllers and flow meters that can be seamlessly incorporated into operators’ existing workflow. Firstly, the digital display area of the equipment is extracted with a screen detection module, and a perspective correction step is performed. Subsequently, the text regions are identified with a fine-tuned EAST text detector, and the important readings are selected through template matching based on the expected graphical structure. Finally, a fine-tuned convolutional recurrent neural network model recognizes the text and registers it. Evaluation experiments confirm the robustness and potential for workload reduction of the proposed system, which correctly extracts 55.47% and 63.70% of the values for reading in universal controllers, and 73.08% of the values from flow meters. Furthermore, this pipeline performs in real time in a low-end mobile device, with an average execution time in preview of under 250 ms for screen detection and on an acquired photo of 1500 ms for the entire pipeline.