The World Health Organization estimates that there were around 10 million deaths due to cancer in 2020, and lung cancer was the most common type of cancer, with over 2.2 million new cases and 1.8 million deaths. While there have been advances in the diagnosis and prediction of lung cancer, there is still a need for new, intelligent methods or diagnostic tools to help medical professionals detect the disease. Since it is currently unable to detect at an early stage, speedy detection and identification are crucial because they can increase a patient’s chances of survival. This article focuses on developing a new tool for diagnosing lung tumors and providing thermal touch feedback using virtual reality visualization and thermal technology. This tool is intended to help identify and locate tumors and measure the size and temperature of the tumor surface. The tool uses data from CT scans to create a virtual reality visualization of the lung tissue and includes a thermal display incorporated into a haptic device. The tool is also tested by touching virtual tumors in a virtual reality application. On the other hand, thermal feedback could be used as a sensory substitute or adjunct for visual or tactile feedback. The experimental results are evaluated with the performance comparison of different algorithms and demonstrate that the proposed thermal model is effective. The results also show that the tool can estimate the characteristics of tumors accurately and that it has the potential to be used in a virtual reality application to “touch” virtual tumors. In other words, the results support the use of the tool for diagnosing lung tumors and providing thermal touch feedback using virtual reality visualization, force, and thermal technology.

Plane wave imaging persists as a focal point of research due to its high frame rate and low complexity. However, in spite of these advantages, its performance can be compromised by several factors such as noise, speckle, and artifacts that affect the image quality and resolution. In this paper, we propose an attention-based complex convolutional residual U-Net to reconstruct improved in-phase/quadrature complex data from a single insonification acquisition that matches diverging wave imaging. Our approach introduces an attention mechanism to the complex domain in conjunction with complex convolution to incorporate phase information and improve the image quality matching images obtained using coherent compounding imaging. To validate the effectiveness of this method, we trained our network on a simulated phased array dataset and evaluated it using in vitro and in vivo data. The experimental results show that our approach improved the ultrasound image quality by focusing the network’s attention on critical aspects of the complex data to identify and separate different regions of interest from background noise.

In this paper, we present a low-cost virtual reality (VR) collaborative system equipped with a haptic feedback sensation system. This system is composed of a Kinect sensor for bodies and gestures detection, a microcontroller and vibrators to simulate outside interactions, and smartphone powered cardboard, all of this are put into a network implemented with Unity 3D game engine. CCS CONCEPTS $\bullet$ Interaction paradigms $\rightarrow$ Virtual reality; Collaborative interaction; $\bullet$ Hardware $\rightarrow$ Sensors and actuators; Wireless devices; KEYWORDS collaborative virtual reality, haptic feedback system.

Développement d’une imagerie ultrasonore interventionnelle précise: Cette thèse explore les méthodes avancées de reconstruction d’images ultrasonores à l’aide de techniques d’apprentissage profond, en mettant l’accent sur l’amélioration de la qualité et de la résolution des images pour l’imagerie interventionnelle. L’étude commence par un examen des principes fondamentaux de l’imagerie ultrasonore, y compris la propagation des ondes, la configuration du matériel, les modalités d’imagerie et les techniques de visualisation, suivi d’un examen des méthodes actuelles de formation de faisceaux ainsi que de l’état de l’art des techniques d’apprentissage profond pour la reconstruction d’images ultrasonores et de leurs limites. Deux nouveaux cadres d’apprentissage profond sont proposés : le premier utilise des réseaux à valeurs complexes pour reconstruire des images échographiques de haute qualité à partir d’une seule acquisition d’ondes planes, et le second introduit une méthode de super-résolution pour la reconstruction d’images en utilisant moins d’ondes divergentes. Les résultats démontrent des améliorations significatives en termes de qualité d’image, de résolution et d’efficacité de calcul, ouvrant la voie à des applications d’imagerie ultrasonore plus précises et en temps réel.