Convolutional neural networks (CNNs) have been widely used to detect and classify various objects and structures in computer vision and medical imaging. Access to large sets of annotated data is commonly a prerequisite for achieving good performance. In medical imaging, acquiring adequate amounts of labelled data can often be time consuming and costly. There- fore, reducing the need for data and in particular associated annotations, is of high importance for medical imaging applications. In this work we investigated whether we can lower the need of annotated data for a super- vised learning classification problem. We chose to tackle the problem of lung nodule detection in thoracic computed tomography (CT) imaging, as this widely investigated applica- tion allowed us to benefit from publicly available data and benchmark our methods. We designed a 3D CNN architecture to perform patch-wise clas- sification of candidate nodules for false positive reduction. Its training, testing and fine-tuning procedure is optimized, we evaluated its perfor- mance, and we compared it with other state-of-the-art approaches in the field. Next, we explored how data augmentation can contribute towards more accurate and less data-demanding models. We investigated the relative benefit of increasing the amount of original data, with respect to computa- tionally augmenting the amount of training samples. Our result indicated that in general, better performance is achieved when increasing the amount of unique data samples, or augmenting the data more extensively, as ex- pected. Surprisingly however, we observed that after reaching a certain amount of training samples, data augmentation led to significantly better performance compared to adding unique samples. Amongst investigated augmentation methods, rotations were found to be most beneficial for im- proving model performance. Following, we investigated the benefit of combining deep learning with handcrafted features. We explored three fusion strategies with in- creasing complexity and assessed their performance for varying amounts of training data. Our findings indicated that combining handcrafted fea- tures with a 3D CNN approach significantly improved lung nodule detec- tion performance in comparison to an independently trained CNN model, regardless of the fusion strategy. Comparatively larger increases in perfor- mance were obtained when less training data was available. The fusion strategy in which features are combined with a CNN using a single end- to-end training scheme performed best overall, allowing to reduce training data by 33\% to 43\%, while maintaining performance. Among the investi- gated handcrafted features, those that describe the relative position of the candidate with respect to the lung wall and mediastinum, were found to be of most benefit. Finally, we considered the case in which abundant data is available, but annotations are scarce, and investigated several methods to improve label-efficiency and their combined effect. We proposed a framework that utilizes both annotated and unannotated data, can be pretrained via self- supervision, and allows to combine handcrafted features with learned rep- resentations. Interestingly, the improvements in performance derived from the proposed learning schemes were found to accumulate, leading to in- creased label-efficiency when these strategies are combined. We observed a potential to decrease the amount of annotated data up to 68\% when com- pared to traditional supervised training, while maintaining performance. Our findings indicate that the investigated methods allow consider- able reduction of data and/or annotations while maintaining model per- formance for lung nodule detection from CT imaging. Future work should investigate whether these results generalize to other domains, such that more applications that face challenges due to a shortage of annotated data may benefit from the potential of deep learning.