Christian Kargel

IN VIVO TIME HARMONIC ELASTOGRAPHY OF THE HUMAN HEART

Time harmonic elastography is introduced as a modality for assessing myocardial elasticity changes during the cardiac cycle. It is based on external stimulation and real-time analysis of 30-Hz harmonic shear waves in axial direction of a parasternal line of sight through the lateral heart wall. In 20 healthy volunteers, the externally induced waves showed smaller amplitudes during systole (76.0 ± 30.8 μm) and higher amplitudes during diastole (126.7 ± 52.1 μm). This periodic wave amplitude alteration preceded ventricular contraction and dilation by about 100 ms. The amplitude ratio of 1.75 ± 0.49 indicates a relative change in myocardial shear elasticity on the order of 14 ± 11. These results well agree with observations made by cardiac magnetic resonance elastography for a similar displacement component and region of the heart. The proposed method provides reproducible elastodynamic information on the heart in real-time and may help in diagnosing myocardial relaxation abnormalities in the future.

Demand controlled ventilation strategies for high indoor air quality and low heating energy demand

In this study several demand controlled ventilation (DCV) strategies are compared with regard to indoor air quality, thermal comfort, and heating energy demand. In order to appropriately model the office room equipped with a PID controlled floor heating system and 4 commonly used bottom-hinged tilted windows located in our low-energy research house (SmartHOME), the thermal time constants and air exchange rates including the infiltration rate and natural ventilation rate were measured. Simulations were then performed for 2 typical winter days in Munich in the accordingly modeled office room with 2 adults working from 8 a.m. to 6 p.m. The DCV strategies ensure that the CO2 concentration of the indoor air stays within the preset range. The Pettenkofer limit of 1000 ppm is a hygienically approved limit value for long-term occupied spaces. The results show that the combination of appropriate DCV and floor heating control strategies can greatly reduce the heating energy consumption and at the same time guarantees high indoor air quality and low room temperature variations. It can decrease the total heating energy demand of the office room during working hours by approximately 27% and save 4.4 kWh per typical winter day compared with the usually recommended “brief and intensive airing” method. Moreover, based on our measurements we performed CFD (Computational Fluid Dynamics) simulations in order to determine the temporally resolved 3D distributions of the indoor CO2 concentration. These results help determine the best location in the room for a single sensor inside the room to representatively measure the CO2 concentration.

Multiple hypothesis tracking of two persons using a network of LIDAR sensors with stationary and directional beams

A sensor network based on the LIDAR (LIght Detection And Ranging) principle is investigated in order to track persons inside a surveillance area and be able to identify security-relevant behavior. In order to minimize the overall sensor network complexity, power consumption and costs, we recently investigated the network topology based on a quality measure in terms of the number of nodes, measurement distance, width of the LIDAR beams and localization as well as classification performance. As a result, stationary beams with rather small opening angles of up to a few degrees are a good compromise. Since certain regions of the surveillance area are not directly assessed by the LIDAR beams, the accurate tracking of a target throughout the entire area of surveillance is challenging. We demonstrated that tracking based on a Kalman filter approach can nevertheless deliver satisfactory results for a single person inside the surveillance area. In this paper we focus on the task of reliably tracking two persons inside the surveillance area at the same time. The tracking of multiple moving targets is carried out by applying a Multiple Hypothesis Tracking filter approach. The localization and tracking performances are derived from simulations and experiments carried out with commercial laser scanners. In the near future, the rather expensive laser scanners will be replaced by appropriate LIDAR range finders.

A new approach to improve the reconstruction quality in ultrasound elastography

The development of new medical imaging methods with high sensitivity and specificity for early cancer detection is of paramount interest in medical diagnostics. Since carcinomas are known to often show an increased stiffness compared to the surrounding healthy tissue, the so-called elastography is well-suited for their detection in soft tissue like the female breast. In elastography, the spatial distribution of elastic properties like the E-modulus is displayed as a two-dimensional map. In order to deduce the E-modulus distribution from the tissues response (i.e. the estimated internal tissue displacements) to an external compression force, the inverse problem governed by the equilibrium equations of linear elastostatics must be solved. In this paper we propose a new approach to reduce the influence of the spatially non-stationary displacement estimation variance on the E-modulus reconstruction, and thus improve the reconstruction quality.

Decentralized multi-target-tracking using a LIDAR sensor network

A LIDAR (LIght Detection And Ranging) sensor network to track walking persons inside a surveillance area is investigated. A small number of sensor nodes with spatially stationary and partially overlapping narrow LIDAR beams is chosen. As a consequence of this network topology, the area of surveillance is not fully covered with LIDAR beams and thus the accurate tracking of persons walking inside the area of surveillance is challenging, especially in a multi-target situation. To tackle this problem, multiple target tracking based on a sophisticated decentralized track-to-track fusion architecture is developed and evaluated in this paper: dynamic multi-hypothesis tracking (MHT) by independent local trackers is carried out in all sensor nodes; then local track favorites are sent to a fusion center where global track candidates are derived and fed back to the local trackers in order to improve the local tracking. With this architecture a track association success rate of (98.8 ± 0.3)% and a mean square position error of Δp = 6.7 cm were derived from 1000 pairs of random and intersecting trajectories of two walking persons (mean velocity 1.5 m/s). Furthermore, the tracking performance as a function of the target velocity was quantified. The performance of the proposed algorithm was also experimentally evaluated using general-purpose laser scanners.

Detection and tracking of humans with a sparse network of LIDAR sensors

The applicability of a sparse sensor network with only two sensor nodes and a small number of directional LIDAR sensors to detect and track humans in an area of surveillance is investigated. The detection and tracking performances are evaluated for various positions of the two nodes as a function of the number of sensors per node and the sensor beamwidths. A quality factor incorporating the area coverage ratio and the position error is introduced to find the best network configuration with a minimal number of sensors yielding a position accuracy sufficient for the task at hand. Extensive simulations and measurements with two laserscanners to emulate the LIDAR sensors were carried out for straight trajectories uniformly distributed over the area of surveillance. In order to improve the tracking performance, we used a Kalman filter based approach. As in our application a spatial mean RMS position error of approx. 0.6 m is sufficient, each of the two sensor nodes must be equipped with 4 LIDAR sensors with a -3dB-beamwidth of 12°.

Optimized Time-Gated Fluorescence Spectroscopy for the Classification and Recycling of Fluorescently Labeled Plastics.

For the production of high-quality parts from recycled plastics, a very high purity of the plastic waste to be recycled is mandatory. The incorporation of fluorescent tracers (“markers”) into plastics during the manufacturing process helps overcome typical problems of non-tracer based optical classification methods. Despite the unique emission spectra of fluorescent markers, the classification becomes difficult when the host plastics exhibit (strong) autofluorescence that spectrally overlaps the marker fluorescence. Increasing the marker concentration is not an option from an economic perspective and might also adversely affect the properties of the plastics. A measurement approach that suppresses the autofluorescence in the acquired signal is time-gated fluorescence spectroscopy (TGFS). Unfortunately, TGFS is associated with a lower signal-to-noise (S/N) ratio, which results in larger classification errors. In order to optimize the S/N ratio we investigate and validate the best TGFS parameters—derived from a model for the fluorescence signal—for plastics labeled with four specifically designed fluorescent markers. In this study we also demonstrate the implementation of TGFS on a measurement and classification prototype system and determine its performance. Mean values for a sensitivity of TPR ¯  = 99.93% and precision PPV ¯  = 99.80% were achieved, proving that a highly reliable classification of plastics can be achieved in practice.

Extended spectral unmixing for the classification of fluorescently labeled plastic waste

Plastics are indispensable materials in modern life and their worldwide production has been steadily increasing in the last 20 years. Unfortunately, only about one quarter of the produced plastics is recycled into new products. Given the enormous (global) challenges associated with plastic waste, the recycling rate must be increased. We present the results of a hyperspectral imaging prototype system developed for the real-time identification (classification) and sorting of a mixture of small plastic flakes delivered on a conveyor belt. The spectra emitted from fluorescent tracers incorporated into the different plastics at the ppm level according to a binary coding scheme serve as unique optical fingerprints for classification purposes. An extended form of hyperspectral unmixing provides highly appropriate features for the maximum-a-posteriori (MAP) classifier used, and additionally suppresses the adverse effects from inherent (auto-)fluorescence emitted from the hosting plastics. Results derived from 160,000 plastic flakes (4 bit binary coding with 16 different plastic classes) analyzed with the prototype system prove the excellent performance of this approach. Very high values for the overall sensitivity TPR̅=99.867% and precision PPV̅=99.8725% were achieved, which are essential in practice for automated sorting with high reliability at the required purity of the sorted material.

Reduction of the lateral localization error of targets moving through a LIDAR beam

To track walking persons inside a surveillance area we use LIDAR (LIght Detection And Ranging) sensors with a small number N of spatially stationary LIDAR beams in order to keep the sensor costs to a minimum. To achieve high target detectability and tracking performance, the coverage of the surveillance area by the N LIDAR beams must be large, which is why the beamwidth is to be set to a practically feasible maximum. As a result, the lateral localization error inside these wide LIDAR beams is high while the area of surveillance can still not be entirely covered by LIDAR beams. Thus, the accurate tracking of persons walking inside the area of surveillance is challenging. In the classical tracking approach, the axial position of a target inside a LIDAR beam is obtained from time-of- ight measurements. However, the lateral deviation of the target position from the optical beam axis remains unknown. In this paper, a novel approach to reduce the lateral localization error is proposed and investigated. From consecutively measured (axial) distances to the target while it moves through a LIDAR beam the target velocity vector is estimated and used as observation for a Kalman-based tracking algorithm. The localization and tracking performances of the novel approach are determined and compared with those of the classical approach.