Search

Today’s mechanical tools for bone cutting (osteotomy) lead to mechanical trauma that prolong the healing process. Medical device manufacturers continuously strive to improve their tools to minimize such trauma. One example of such a new tool and procedure is minimally invasive surgery with laser as the cutting element. This setup allows for tissue ablation using laser light instead of mechanical tools, which reduces the post-surgery healing time. During surgery, a reliable feedback system is crucial to avoid collateral damage to the surrounding tissues. Therefore, we propose a tissue classification method that analyzes the acoustic waves produced during laser ablation and show its applicability in an ex-vivo experiment. The ablation process with a microsecond pulsed Erbium-doped Yttrium Aluminium Garnet (Er:YAG) laser produces acoustic waves that we captured with an air-coupled transducer. Consequently, we used these captured waves to classify five porcine tissue types: hard bone, soft bone, muscle, fat, and skin tissue. For automated tissue classification of the measured acoustic waves, we propose three Neural Network (NN) approaches: A Fully-connected Neural Network (FcNN), a one-dimensional Convolutional Neural Network (CNN), and a Recurrent Neural Network (RNN). The time- and the frequency-dependent parts of the measured waves’ pressure variation were used as separate inputs to train and validate the designed NNs. In a final step, we used Grad-CAM to find the frequencies’ activation map and conclude that the low frequencies are the most important ones for this classification task. In our experiments, we achieved an accuracy of 100% for the five tissue types for all the proposed NNs. We tested the different classifiers for their robustness and concluded that using frequency-dependent data together with a FcNN is the most robust approach.

Countless low-surface brightness objects - including spiral galaxies, dwarf galaxies, and noise patterns - have been detected in recent large surveys. Classically, astronomers visually inspect those detections to distinguish between real low-surface brightness galaxies and artefacts. Employing the Dark Energy Survey (DES) and machine learning techniques, Tanoglidis et al. (2020) have shown how this task can be automatically performed by computers. Here, we build upon their pioneering work and further separate the detected low-surface brightness galaxies into spirals, dwarf ellipticals, and dwarf irregular galaxies. For this purpose, we have manually classified 5567 detections from multi-band images from DES and searched for a neural network architecture capable of this task. Employing a hyperparameter search, we find a family of convolutional neural networks achieving similar results as with the manual classification, with an accuracy of 85% for spiral galaxies, 94% for dwarf ellipticals, and 52% for dwarf irregulars. For dwarf irregulars - due to their diversity in morphology - the task is difficult for humans and machines alike. Our simple architecture shows that machine learning can reduce the workload of astronomers studying large data sets by orders of magnitudes, as will be available in the near future with missions such as Euclid., 8 pages, 6 figures, 1 table, accepted for publication in The Open Journal of Astrophysics. The code can be found on GitLab: https://gitlab.com/VoltarCH/deeplearning_des

Segmentation of distinct bones plays a crucial role in diagnosis, planning, navigation, and the assessment of bone metastasis. It supplies semantic knowledge to visualisation tools for the planning of surgical interventions and the education of health professionals. Fully supervised segmentation of 3D data using Deep Learning methods has been extensively studied for many tasks but is usually restricted to distinguishing only a handful of classes. With 125 distinct bones, our case includes many more labels than typical 3D segmentation tasks. For this reason, the direct adaptation of most established methods is not possible. This paper discusses the intricacies of training a 3D segmentation network in a many-label setting and shows necessary modifications in network architecture, loss function, and data augmentation. As a result, we demonstrate the robustness of our method by automatically segmenting over one hundred distinct bones simultaneously in an end-to-end learnt fashion from a CT-scan.