De TU Delft leidt natuurkundestudenten op tot natuurkundig ingenieurs voor de 21e eeuw. Een kernvaardigheid die hen onderscheidt van veel andere academici is het ontwerpen van fysische experimenten en apparaten, die als doel hebben een fysisch verschijnsel te kunnen waarnemen, beïnvloeden, of kwantificeren. Natuurkundig ingenieurs vervullen zo een eigen rol binnen teams van wetenschappers. Daar waar ontwerpen traditioneel een vaardigheid is die natuurkundig ingenieurs impliciet verwerven, waardoor de huidige generatie natuurkundedocenten deze vaardigheid slechts impliciet kan overbrengen, wil de TU Delft ontwerpen tot een expliciet onderdeel van het curriculum maken. Hierbij werken we met onderwijs gebaseerd op de Maker Education filosofie. Maaktechnologie is tegenwoordig zoveel bereikbaarder geworden, dat ontwerpprocessen zélf veranderen: maken, bijstellen, en opnieuw maken, vervangt het oude ‘eerst bedenken dan maken’-model. De Maker Education aanpak combineert de mogelijkheden van ons huidige tijdperk – 3D-printing, lasersnijden, internet-of-things – met de behoefte aan meer aansluiting tussen onderwijs en de multidisciplinaire vragen van de samenleving. In ons voorstel zullen studenten Technische Natuurkunde niet alleen leren zelf opstellingen en systemen te ontwerpen, maar deze ook daadwerkelijk te maken, rekening houdend met praktische, technische, duurzaamheids- en financiële randvoorwaarden, en intellectueel eigendom. Zo leren studenten Technische Natuurkunde gebruik te maken van de mogelijkheden die nieuwe maaktechnologieën bieden en krijgen zij ontwerponderwijs dat aansluit bij de hedendaagse hightech-beroepspraktijk. Het voorstel omvat een disseminatieplan waarin de opzet van de leergang, het ontwikkelde cursusmateriaal, en een wetenschappelijke evaluatie van Maker Education in ontwerponderwijs zullen worden gedeeld met VO, HBO, WO, onderwijskundigen en het brede publiek

Heart failure is a major public health problem leading to impaired quality of life, shortness of breath, and high mortality rates. In more than half of the cases, in women even more than in men, the important underlying cause is stiffening of the left ventricular part of the heart muscle, leading to poor filling of the heart. Currently there is no good test to diagnose increased stiffening of the heart and there is no good method of monitoring any therapy. The investigators propose a novel method of diagnosis that is based on safe and noninvasive echographic imaging of the heart. This method aims to measure minute natural vibrations of the heart wall. The research focuses to understanding and interpreting these vibrations to deduce the stiffness of heart wall. Experimental validation in both a laboratory environment and in three hospital centers will show its applicability.

Background: Cancer is first-leading cause of adult deaths. Radiation therapy based on x-rays cure approximately 50% of all cancer patients and is a fundamental pillar in cancer treatment. But, collateral damage of healthy surrounding tissues is unavoidable. Proton beams differ from x-rays by the fact that their penetration depth is sharply determined and release their energy at the Bragg peak. The problem: In proton beam therapy the dosimetry is determined by simulations of the proton deposition in the tissue. However, organ movement and errors in the assumed material properties lead to inaccuracy of the deposition. The Solution: We propose a non-invasive, in-situ, real-time localization system for proton therapy monitoring using ultrasound contrast agents and highly sensitive optical-acoustical receivers. Our concept consists of two innovative steps. The first step is the interaction of the proton beam with a medical ultrasound contrast agent consisting of coated microbubbles. The energy deposition from individual protons in the Bragg peak creates a broadband excitation in the vicinity of the bubble forcing them to vibrate at their resonance frequency (1-10 MHz). This creates a low amplitude pressure wave that can be used for localization and dose measurement of the proton beam. The second step entails the development of an array of ultra-sensitive acousto-optical ultrasound sensors for detecting the acoustic pressure waves generated by the microbubbles, which is one order of magnitude below the detection limit of current state-of-the-art ultrasonic sensors. Acousto-optical sensors consist of a silicon chip with an extremely thin membrane that will already be deflected by very small acoustic pressure amplitudes. This deflection will be detected by a micro-optical circuit that is integrated on the membrane. Using the microbubbles and these highly sensitive receivers allows for a real-time monitoring of the proton deposition with a spatial resolution better than 1 mm.

The task of translating from one language into another using the computer, called Machine Translation (MT), has been one of the central challenges for the natural language processing community for decades now. Recently, neural models for machine translation (MT) have received much attention. This interest is partially fueled by the successes of neural and other representation learning methods in some domains (e.g., image and speech processing, reinforcement learning) but it is also motivated by recognized limitations of traditional MT systems (e.g., these systems do not directly model paraphrasing or semantic similarity). The aim of this (sub-)project is to exploit fast parallel GPU computation to train large neural networks that learn meaningful representations of input sentences informed by hierarchical structure of a sentences so that better translation quality can be achieved.