Deep neural networks have achieved great success in many domains. However, successful deployment of such systems is determined by proper manual selection of the neural architecture. This is a tedious and time-consuming process that requires expert knowledge. Different tasks need very different architectures to obtain satisfactory results. The group of methods called the neural architecture search (NAS) helps to find effective architecture in an automated manner. In this paper, we present the use of an architecture search framework to solve the medical task of malignant melanoma detection. Unlike many other methods tested on benchmark datasets, we tested it on practical problem, which differs greatly in terms of difficulty in distinguishing between classes, resolution of images, data balance within the classes, and the number of data available. In order to find a suitable network structure, the hill-climbing search strategy was employed along with network morphism operations to explore the search space. The network morphism operations allow for incremental increases in the network size with the use of the previously trained network. This kind of knowledge reusing allows significantly reducing the computational cost. The proposed approach produces structures that achieve similar results to those provided by manually designed structures, at the same time making use of almost 20 times fewer parameters. What is more, the search process lasts on average only 18h on single GPU.

Breast cancer is the most frequent female cancer, with a considerable disease burden and high mortality. Early diagnosis with screening mammography might be facilitated by automated systems supported by deep learning artificial intelligence. We propose a model based on a weakly supervised Clustering-constrained Attention Multiple Instance Learning (CLAM) classifier able to train under data scarcity effectively. We used a private dataset with 1174 non-cancer and 794 cancer images labelled at the image level with pathological ground truth confirmation. We used feature extractors (ResNet-18, ResNet-34, ResNet-50 and EfficientNet-B0) pre-trained on ImageNet. The best results were achieved with multimodal-view classification using both CC and MLO images simultaneously, resized by half, with a patch size of 224 px and an overlap of 0.25. It resulted in AUC-ROC = 0.896 ± 0.017, F1-score 81.8 ± 3.2, accuracy 81.6 ± 3.2, precision 82.4 ± 3.3, and recall 81.6 ± 3.2. Evaluation with the Chinese Mammography Database, with 5-fold cross-validation, patient-wise breakdowns, and transfer learning, resulted in AUC-ROC 0.848 ± 0.015, F1-score 78.6 ± 2.0, accuracy 78.4 ± 1.9, precision 78.8 ± 2.0, and recall 78.4 ± 1.9. The CLAM algorithm’s attentional maps indicate the features most relevant to the algorithm in the images. Our approach was more effective than in many other studies, allowing for some explainability and identifying erroneous predictions based on the wrong premises.

To successfully train a deep neural network, a large amount of human-labeled data is required. Unfortunately, in many areas, collecting and labeling data is a difficult and tedious task. Several ways have been developed to mitigate the problem associated with the shortage of data, the most common of which is transfer learning. However, in many cases, the use of transfer learning as the only remedy is insufficient. In this study, we improve deep neural models training and increase the classification accuracy under a scarcity of data by the use of the self-supervised learning technique. Self-supervised learning allows an unlabeled dataset to be used for pretraining the network, as opposed to transfer learning that requires labeled datasets. The pretrained network can be then fine-tuned using the annotated data. Moreover, we investigated the effect of combining the self-supervised learning approach with transfer learning. It is shown that this strategy outperforms network training from scratch or with transfer learning. The tests were conducted on a very important and sensitive application (skin lesion classification), but the presented approach can be applied to a broader family of applications, especially in the medical domain where the scarcity of data is a real problem.

Machine learning-based systems are gaining interest in the field of medicine, mostly in medical imaging and diagnosis. In this paper, we address the problem of automatic cerebral microbleeds (CMB) detection in magnetic resonance images. It is challenging due to difficulty in distinguishing a true CMB from its mimics, however, if successfully solved, it would streamline the radiologists work. To deal with this complex three-dimensional problem, we propose a machine learning approach based on a 2D Faster RCNN network. We aimed to achieve a reliable system, i.e., with balanced sensitivity and precision. Therefore, we have researched and analysed, among others, impact of the way the training data are provided to the system, their pre-processing, the choice of model and its structure, and also the ways of regularisation. Furthermore, we also carefully analysed the network predictions and proposed an algorithm for its post-processing. The proposed approach enabled for obtaining high precision (89.74%), sensitivity (92.62%), and F1 score (90.84%). The paper presents the main challenges connected with automatic cerebral microbleeds detection, its deep analysis and developed system. The conducted research may significantly contribute to automatic medical diagnosis.