Publications

Isothermal microcalorimetry (IMC) is a promising tool for diagnosing periprosthetic joint infection (PJI), based on real-time measurement of growth-related heat production of pathogens, and faster than conventional microbial cultures. However, the feasibility of identifying specific pathogens in clinical samples using IMC has yet to be proven. This study implements machine learning and transfer learning convolutional neural network (CNN) models to detect and identify pathogens causing PJI, using IMC data alone. IMC data were obtained from synovial fluid samples, including 174 aseptic samples and 239 PJI samples containing five different bacterial strains. XGBoost, multi-layer perceptron, support vector machine, random forest, and three transfer learning CNN models were implemented to detect PJI and identify five bacterial strains in PJI samples. The binary XGBoost classifier yielded a 100% accuracy in PJI detection, whereas the multiclass XGBoost classifier and the combined transfer learning CNN classifier reached an overall accuracy of 90.3% and 91.5%, respectively, in PJI identification, with biological significance of extracted features in the XGBoost model facilitating its interpretability and usage in clinical practice. The strain with the lowest recall (83.3%) was PA, whereas SE was the strain with the lowest precision (78.9%). The results demonstrate the feasibility of automatic detection and identification of pathogens causing PJI using their IMC growth patterns and machine learning models. This adds a critical missing feature to IMC, contributing to accelerating the diagnosis of PJI and the selection of antibiotic therapy.

JTD Keywords: Bacterial strain classification, Convolutional neural network, Growth, Infection, Isothermal microcalorimetry, Model, Periprosthetic joint infection, Xgboos

Providing force feedback as relevant information in current Robot-Assisted Minimally Invasive Surgery systems constitutes a technological challenge due to the constraints imposed by the surgical environment. In this context, force estimation techniques represent a potential solution, enabling to sense the interaction forces between the surgical instruments and soft-tissues. Specifically, if visual feedback is available for observing soft-tissues’ deformation, this feedback can be used to estimate the forces applied to these tissues. To this end, a force estimation model, based on Convolutional Neural Networks and Long-Short Term Memory networks, is proposed in this work. This model is designed to process both, the spatiotemporal information present in video sequences and the temporal structure of tool data (the surgical tool-tip trajectory and its grasping status). A series of analyses are carried out to reveal the advantages of the proposal and the challenges that remain for real applications. This research work focuses on two surgical task scenarios, referred to as pushing and pulling tissue. For these two scenarios, different input data modalities and their effect on the force estimation quality are investigated. These input data modalities are tool data, video sequences and a combination of both. The results suggest that the force estimation quality is better when both, the tool data and video sequences, are processed by the neural network model. Moreover, this study reveals the need for a loss function, designed to promote the modeling of smooth and sharp details found in force signals. Finally, the results show that the modeling of forces due to pulling tasks is more challenging than for the simplest pushing actions.

JTD Keywords: Convolutional neural networks, Force estimation, LSTM networks, Robotic surgery