Dear Colleagues,

The deep learning breakthrough dramatically affected all areas of computer vision and image analysis with subsequent repercussions in biomedical image analysis. The state of the art in standard related tasks, such as classification, retrieval and segmentation of biomedical images has been conquered by methods based on convolutional neural networks (CNNs), recurrent neural networks (RNNs) or reinforcement learning. Various CNN-based configurations, such as CifarNet, AlexNet, GoggLeNet, ResNet50, VCG16, and DenseNet, have demonstrated classification accuracy that challenges the one obtained by domain experts. Still, several issues complicate the successful adoption of such methods in clinical practice: 1) deep learning approaches usually rely on the availability of large labelled datasets, which are often hard to obtain, 2) despite the high overall accuracies obtained, deep architectures often appear as ‘black boxes’ and their results cannot be intuitively interpreted by domain experts, 3) the deployment of deep architectures often relies on GPUs and expensive hardware.

On the other hand, ‘classical’ image analysis methods, based on handmade features (SIFT, SURF, LBP, BRIEF etc.), bag-of-visual words, standard classifiers (random forests, SVMs,…), standard segmentation approaches (level-sets, graph-cuts,…), and rule-based decisions, have been refined for several years and, despite their limitations, provide alternatives with respect to the aforementioned issues. This Special Issue is dedicated to biomedical applications of such classical methods, in a complementary fashion to deep neural networks (DNNs). Submitted works could be original contributions, including but not limited to:

• Hybrid schemes combining handcrafted features and DNN-learned features,
• Computer-aided Diagnosis systems combining rule-based and DNN-based components,
• Intelligent labelling techniques employing classical methods,
• Pretasks for DNN training based on classical image analysis methods,
• Standalone ‘classical’ methods, provided that quantitative or qualitative comparisons with DNNs indicate some advantage in terms of accuracy, efficiency, dataset requirements etc.

Reviews on related challenging topics or comparative studies evaluating classical and DNN-based methods are also welcome.

Prof. Dr. Michalis Savelonas
Prof. Dr. Dimitris Iakovidis
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Sensors is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2600 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

• Machine learning
• pattern recognition
• biomedical image analysis
• biomedical imaging
• image classification
• image retrieval
• image segmentation
• decision support systems
• CAD systems
• unsupervised learning
• handcrafted features