|
Lääketieteellisen fysiikan ja tekniikan yhdistys (LFTY) Finnish Society for Medical Physics and Medical Engineering |
LFT-päivä 13.2.2014, Itä-Suomen yliopisto
Posterikilpailuun osallistuneet
1. Hanna Halme, Aalto yliopistoMeasuring neural mechanisms of error processing with fMRI: model-based and data-driven methods
2. Tero Ilvesmäki, Tampereen teknillinen yliopisto
Neural tract analysis; a software implementation
3. Satu Inkinen, Itä-Suomen yliopisto
Collagen and chondrocyte concentrations control ultrasound scattering in agarose scaffolds
4. Lari Koponen, Aalto-yliopisto
Large thin overlapping coils, a novel approach for multichannel transcranial magnetic stimulation
6. Tuomas Mutanen, Aalto-yliopisto
Brain state dynamics in transcranial magnetic stimulation – a combined TMS-EEG study
7. Niko Mäkelä, Aalto-yliopisto
Probing speech with TMS and EEG
8. Simo Näkki, Itä-Suomen yliopisto
Biocompatibility improvement of mesoporous silicon by PEGylation
9. Mikko Peltokangas, Tampereen teknillinen yliopisto
Analysis methods for arterial pulse wave signals recorded with body sensor network
10. Paruthi Pradhapan, Tampereen teknillinen yliopisto
Heart rate variability based study on sleep recovery in shift working truck drivers
11. Defne Us, Tampereen teknillinen yliopisto
Metal Artifact Reduction in Sinograms of Dental Computed Tomography
12. Mikko Venäläinen, Itä-Suomen yliopisto
Importance of material properties and structure of bone in biomechanical modeling of knee joint function
13. Joni Vuorio, Tampereen teknillinen yliopisto
Binding of Hyaluronic Acid to Its CD44 Receptor
14. Tiziana Zedda, Tampereen teknillinen yliopisto
Construction and testing of a PET demostrator
Postereiden abstraktit
MEASURING NEURAL MECHANISMS OF ERROR PROCESSING WITH FMRI: MODEL-BASED AND DATA-DRIVEN METHODSHanna Halme, Aalto-yliopisto
Background
Even though processing of errors is one of the most fundamental cognitive functions, it is still unclear whether the human brain processes self-generated and observed errors similarly. Furthermore, is should be validated whether observed errors in naturalistic and simulated conditions produce different responses.
Methods
In this study, we examined the neural mechanisms of error processing with functional magnetic resonance imaging (fMRI) during self-committed errors, as well as observed errors made by others in both simulated and naturalistic conditions. In the first experiment the subjects played a simple response selection game, occasionally making errors. In the second experiment they watched a video recording of the aforementioned game played by another player. In the third experiment they watched short video clips depicting other people making errors in everyday situations. The fMRI data were analyzed with the general linear model (GLM) and independent component analysis (ICA) in order to detect error-related activation and functional connectivity in each condition. With the third task we also examined the activity caused by error anticipation with GLM. In addition, correlations between error-related BOLD responses in 23 regions of interest (ROI) were calculated, and the regional responses in different experimental conditions were compared.
Results
Similar activations were detected during self-committed and observed naturalistic errors in striatal subregions (caudate nucleus and globus pallidus), rostral anterior cingulate cortex (rACC) and visual cortical regions. Dorsal ACC, inferior frontal gyrus and insula showed similar activation during self-committed errors and anticipation of observed naturalistic errors. Observed errors in the game could not produce a robust hemodynamic response. Both ICA and ROI-based analyses indicated high functional connectivity between the key regions of the error monitoring circuit during self-committed and observed naturalistic errors.
Conclusions
The results of this study support the theories advocating distinct functions of rostral and dorsal ACC in error monitoring and suggest that the striatum processes self-generated and observed errors quite similarly. In addition, this study provides further evidence for the feasibility of naturalistic stimuli in neuroscientific research, since errors in the naturalistic video clips elicited significant hemodynamic responses, whereas simulated video game errors did not.
NEURAL TRACT ANALYSIS; A SOFTWARE IMPLEMENTATION
Tero Ilvesmäki, Tampereen teknillinen yliopisto
Background:
Diffusion tensor imaging (DTI) is an excellent imaging modality to study tissue microstructures such as brain white matter. White matter consists mostly of glial cells and axons. Abnormal changes to neural tracts (axons) can be induced by various causes, e.g. neurological disorders or brain trauma. In order to use DTI data for analytical purposes, the raw diffusion images need to be heavily processed. Fractional anisotropy (FA) derived from DTI is the most used diffusion parameter, and correlates positively with the magnitude and directionality of water diffusion in matter. Other useful diffusion parameters are apparent diffusion coefficient (ADC), axial diffusivity (AD) and radial diffusivity (RD).
Objective:
The objective of my Master’s thesis was to implement an MRI data analysis toolbox, FSL (FMRIB Software Library, University of Oxford, http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/FSL), for hospital environment in Pirkanmaa Hospital District, Medical Imaging Center and apply it in the analysis of spinal cord injury patient data. The essential purpose was to study whether the toolbox could be used as a supportive system for clinical diagnostics.
Materials and methods:
In this research, the tool mainly used of FSL was Tract-Based Spatial Statistics (TBSS). TBSS is used to research groupwise differences in brain connectivity. DTI data provided for the research by Tampere University Hospital consisted of 35 spinal cord injury patients and 40 healthy control subjects, of which 24 patients and 24 controls were used in the analysis. After sufficient familiarization with the software, TBSS was applied to data provided by the hospital.
Results:
Application of the analysis method was successful, and several Unix shell scripts were developed as a byproduct to simplify and partially automate the analysis process. Results of comparison between spinal cord injury patients and controls were obtained and are presentable in both qualitative and quantitative way. Widespread differences in FA values in brain white matter between patient and control groups were observed.
Conclusions:
The method was deemed applicable for research purposes in the hospital environment. Spinal cord injury patient analysis results were promising, but included bias due to the disregard of age effect on DTI parameters. Further analyses with TBSS were planned following the encouraging results of the thesis. It is imperative to account for white matter changes caused by aging in future DTI analyses.
COLLAGEN AND CHONDROCYTE CONCENTRATIONS CONTROL ULTRASOUND SCATTERING IN AGAROSE SCAFFOLDS
Satu Inkinen, Itä-Suomen yliopisto
Introduction Ultrasound arthroscopy is a novel technique proposed for diagnostics of cartilage injuries and osteoarthritis in vivo [1]. Collagen content and organization of collagen network significantly control ultrasound reflection and scattering in cartilage [2]. Although cartilage tissue contains millions of chondrocytes per milliliter, their specific contribution to ultrasound scattering is unknown. This creates uncertainty when interpreting changes in ultrasound scattering within cartilage matrix and warrants for further investigations. Therefore, the aim of this study was to evaluate ultrasound scattering in agarose scaffolds with varying collagen content (type I) and chondrocyte density. Use of agarose scaffolds facilitates controlled, cartilage-like environments for ultrasound measurements.
Materials and methods Two different sets of 4% agarose gel samples were prepared. The first set contained nine series of samples with different chondrocyte contents (1, 2, 4, 8, 16 and 32 million/ml). The second set contained nine series of samples with varying collagen type I concentration (1.50, 3.13, 6.25, 12.50, 25.0, 50.0 and 100.00 mg/ml). In all series, reference samples of agarose gel (4%) with no chondrocytes or collagen were prepared. In agarose gel samples both chondrocytes and collagen were evenly distributed and randomly oriented. All samples were imaged with an intravascular ultrasound device (Clear View Ultra, Boston Scientific Corporation, San Jose, CA, USA). Two catheters with frequencies of 9 MHz (diameter = 2.7 mm, -6 dB bandwidth 7.1-11.0 MHz) and 40 MHz (diameter = 1 mm, -6 dB bandwidth, 30.1-45.3 MHz) were used. Ultrasound image data was recorded and stored for off-line analysis with MATLAB (Math Works Inc., Natick, MA, USA). For each sample, ultrasound backscattering was determined as an average grayscale value within a rectangular window (1 x 2 mm2 for 40 MHz, 2.8 x 4 mm2 for 9 MHz) positioned immediately below the sample surface. An average image, including 30 frames inside the rectangular window from same spatial location, was calculated for every sample. Subsequently, a subtraction image between the average image of the reference sample (pure agarose) and the average image of scaffold containing chondrocytes or collagen was calculated. Based on the subtraction image an intensity histogram of 100 bins was formed. To evaluate the changes in distribution of ultrasound backscattering amplitude within the scaffolds the skewness of the histogram was determined. Kruskal-Wallis test was used to evaluate the statistical significance of difference in ultrasound scattering in scaffolds with different compositions (p < 0.05).
Results Both collagen and chondrocytes affected ultrasound scattering. Ultrasound backscattering quantified in a region of interest starting right below sample surface was significantly different (p<0.05) with collagen and chondrocyte concentrations.
Discussion Both collagen and chondrocytes were found to significantly affect ultrasound scattering in agarose. With present experimental material, the relative contribution of chondrocytes was greater than that of collagen. This was surprising and may be explained by the applied concentrations. In human cartilage, collagen content is ~200 mg/ml [3] and chondrocyte density is ~14 million/ml [4]. In this study, collagen contents of 1-50 mg/ml and chondrocyte densities of 1-32 million/ml were applied. Thus, the collagen concentrations used in this study were less than one third of that normally present in human cartilage. Considering this, the results suggest that collagen may be more important ultrasound scatterer than the chondrocytes. However, further research with higher collagen concentrations is needed to verify this. Both 9 MHz and 40 MHz ultrasound were found to be sensitive to variation in collagen and chondrocyte concentrations in agarose. The present measurements covered only a limited range of ultrasound frequencies. In future, effects of chondrocytes and collagen on ultrasound scattering should be evaluated with wider range of ultrasound frequencies. In the present study, chondrocytes and collagen were investigated separately but also samples with both components should be investigated. Furthermore, the effects of orientation of collagen fibers on ultrasound scattering in cartilage warrants for investigation.
Significance It was shown that quantitative ultrasound imaging is sensitive to variations in collagen and chondrocyte concentrations in agarose. This knowledge can improve the interpretation of quantitative ultrasound imaging and modeling of ultrasound propagation in cartilage. References 1. Virén et al. 2010 Ultrasound in Med & Biol. 2. Töyräs et al. 2002 Biorheology. 3. Basser et al. 1998. Arch. Biochem. Biophys. 4. Stockwell et al. 1978. J. Clin. Suppl. (Roy. Coll. Path)
TRANSCRANIAL MAGNETIC STIMULATION
Lari Koponen, Aalto-yliopisto
Transcranial magnetic stimulation (TMS) allows for studying the functionality of the brain. Present TMS devices have one, or two separate, stimulation coils. More stimulation coils would allow new types of stimulation sequences, and thus they could be used to reveal more about brain functionality. However, due to the dimensions of the existing TMS coils, having multiple separate coils is a very limited approach. Rather, the coils should be combined into a single multichannel (mTMS) device.
The purpose of my Master’s Thesis was to make mTMS more feasible. In order to realize this purpose, a new coil design paradigm was introduced which employs large thin overlapping coils. This paradigm required a new coil design method and a new coil-former design method, which were developed and tested in the Thesis.
The new coil design paradigm requires significantly lower number of channels for similar degrees of freedom for the TMS-induced E-field than the previously proposed designs utilizing a lattice of small coils (a minimum of 4 channels are required for making small corrections to the coil position and orientation, instead of 16 (a four-by-four lattice)). At the same time, the proposed paradigm requires significantly smaller coil currents for a similar intensity stimulus. Thus, the Thesis solved two major problems that appeared with existing mTMS designs, taking use one step closer to a working mTMS device.
Brain state dynamics in transcranial magnetic stimulation – a combined TMS-EEG study
Tuomas Mutanen, Aalto-yliopisto
Transcranial magnetic stimulation (TMS) and electroencephalography (EEG) have been successfully combined to study the connectivity and reactivity of the brain. However, it is not yet well understood how TMS modulates the ongoing brain activity largely because the present methods used to analyze TMS–EEG signals usually describe the average response to TMS rather than the immediate effects. The purpose of this Thesis was to improve our understanding on the dynamics of TMS by analyzing EEG signals; How does TMS affect the state of the brain, and on the other hand, how does the state of the brain change the effects of TMS? Deeper understanding on this subject is vital when seeking for more elaborate and effective stimulation sequences and methods.
In this Thesis, we introduced two quantitative tools called mean state shift (MSS) and state variance (SV) which we used to measure changes in the brain’s electric activity due to a single TMS pulse. MSS uses EEG data to quantify the immediate TMS-elicited shift in the brain state, whereas SV shows whether TMS modulates the rate at which the brain state is changing. We used the introduced measures to analyze EEG data, which had been measured from subjects that received single-pulse TMS on the primary motor cortex. Additionally, we performed TMS–EEG measurements where the subjects received paired TMS pulses. The purpose of these measurements was to test whether MSS and SV would change as a function of stimulation efficiency.
We show that the introduced measures are able to quantify transient effects of TMS on the electric brain state. Furthermore, both MSS and SV are sensitive to the stimulation parameters and seem to reflect the amount of activation induced on the cortex.
Large MSS implies that TMS shifts the brain to a state that differs from the spontaneous ones. Furthermore, the increase in SV indicates that the TMS-modulated activity is changing more vigorously. Hence, it seems that TMS shifts the brain into a state that is unnatural. This is followed by the increased variation in the electric state until the brain has settled itself into a more natural condition. The findings obtained from the paired pulse measurements support this interpretation.
Probing speech with TMS and EEG
Niko Mäkelä, Aalto yliopisto
a) Objectives
Transcranial magnetic stimulation (TMS) has been used to study cortical speech processing. Combined with simultaneous electroencephalography (EEG), TMS provides a unique, causal mapping tool that could give new information about speech functions. TMS-evoked large muscle artifacts may mask the TMS-evoked short-latency brain activity measured by EEG, especially when applied on lateral areas. Classical speech-related cortical areas are located laterally, which limits the usability of TMS–EEG on studying the cortical speech network. In the Thesis, the applicability of TMS–EEG on studying the lateral speech areas was evaluated. The main hypothesis was that using a functionally located speech area instead of an anatomically located speech area would enhance the signal-to-artifact ratio, especially when TMS is applied during a language task.
b) Methods
First, a novel TMS speech mapping paradigm was used to functionally locate speech-related areas from the left hemisphere of a healthy subject. The subject was naming pictures while repetitive TMS was applied on the cortex. The TMS-induced errors were categorized by type and location. Second, the cortical site with most TMS-induced naming errors was selected as the stimulation target for the TMS– EEG experiment. The TMS–EEG paradigm was repeated for several stimulation intensities with and without a covert picture naming task. A new analysis measure, effective masking length (EML), was introduced to estimate the duration of the disturbing artifact in relation to the underlying brain activation. The TMS–EEG data was analyzed in signal, sensor, and source spaces. In addition, independent component analysis (ICA) was used to separate artifactual components from brain signals. Different approaches to apply ICA on multi-trial TMS–EEG data were used and compared with each other.
c) Results
The results showed that both the amplitude and the duration of the muscle artifacts decreased fast when the stimulation intensity was lowered. The artifacts masked the brain signals for the first 20 milliseconds for the lowest intensity and for 40 ms for the conventional intensity. ICA was able to remove some artifacts, e.g., eye-blinks, but not the large muscle artifacts. ICA on trial-averaged data led to different separation results than on concatenated or mean-subtracted data.
d) Conclusions
TMS speech mapping allows functional localization of speech-related areas that can be used as targets for subsequent TMS–EEG. Using lower stimulation intensities may partially help to overcome the artifact problems of TMS–EEG on artifactual areas. However, the 20 first milliseconds still remain a challenging task in TMS–EEG of lateral areas. In addition, the results put the use of conventional trialaveraged data in ICA in serious doubt.
Biocompatibility improvement of mesoporous silicon by PEGylation
Simo Näkki, Itä-Suomen yliopisto
Introduction
According to the study done by the International Agency for Research on Cancer (IARC), there were 14.1 million new cancer cases and 8.2 million cancer deaths in the year 2012, being the leading cause of death in the world. Nanotechnology and especially porous silicon nanoparticles (PSiNPs) have shown their potential to be the next big leap in the cancer medicine. PSi is especially suitable for medical purposes due its biodegradability, biocompatibility and nontoxicity. However, the plain PSiNPs face the problems of heavy aggregation and protein adsorption in plasma. This causes the reticuloendothelial system (RES) to remove the PSiNPs quickly from blood stream, leading in insufficient circulation times in the body.
Aims
The aim of this study was to improve the biocompatibility of PSiNPs with the surface functionalization of polyethylene glycols (PEGs, PEGylation), which enhances the PSiNP colloidal stability in ionic solutions and decreases the protein adsorption in plasma. Another objective was to functionalize the PEGylated PSiNPs with radiolabelable compounds to enable imaging. As a result, the modified PSiNPs could be a potential material for intravenous administrations such as cancer treatments and imaging.
Methods
First, PSiNPs with two different surface chemistries were manufactured. These nanoparticles were coated with different sized polyethylene glycols (PEGs) to potentially improve the behaviour of the PSiNPs in ionic solutions. The materials were characterized with thermogravimetric analysis, gas sorption, electrophoretic light scattering and Fourier transform infrared spectroscopy. The colloidal stability of PEG coated PSiNPs was measured with dynamic light scattering in phosphate buffered saline (PBS, pH 7.0, incubated at 37 ?C). Finally, the PEGylated PSiNPs were modified with radiolabelable compounds that enable radiolabeling and imaging of the particles.
Results
PEGylation was found to increase the colloidal stability in ionic solutions even though it decreased ? –potential of the PSiNPs. With the best method, the colloidal stability increased up to 9 days, whereas in the initial situation the particles aggregated instantly. The addition of radiolabelable compounds was found decrease the colloidal stability but with the suitable PEGylation sufficient stability was obtained.
Conclusions
The results indicate that it is possible to modify PSiNPs with radiolabelable compounds and coat the particles with PEG. The PEGylation was found to increase colloidal stability of PSiNPs significantly and thus improve their potential for cancer therapy. With these modifications it is possible to manufacture PSiNPs for intravenous applications such as cancer therapy and imaging.
Analysis methods for arterial pulse wave signals recorded with body sensor network
Mikko Peltokangas, Tampereen teknillinen yliopisto
Background and objectives: Arterial pulse waves (PW) provide valuable information on the subject’s vascular health. The PW analysis is based on the fact that the peripheral arterial PW is a combination of heart-beat-induced forward wave caused by a rapid blood ejection from the left ventricle to the aorta and the reflections of this pressure wave. The pulse wave velocity (PWV) in the arteries depends on the elasticity of the arterial wall: the stiffer the arteries are, the faster the PW propagates. Thus, the shape of the observed PW contour at peripheral measurement point depends on the arrival times of the forward wave (percussion wave) and the reflected waves (tidal and dicrotic waves as well as higher order reflections). Because the PW recorded from a person who has stiffened arteries differs significantly from the PW recorded from a person whose arteries are elastic, there are parameters that can be obtained from the PW contour to describe the arterial condition. The main objective of the Master’s thesis is to find methods for analyzing the PW signals recorded with a synchronous wireless body sensor network. However, there are also important tasks before the PW analysis: to assemble the measurement devices, find the suitable measurement points from the human body and study the effects of static bias force on the PW amplitude recorded with a force sensor.
Methods: The presented wireless body sensor enables eight sensor nodes in maximum. Two different types of PW signals are recorded: pressure-related mechanical PWs from wrists, the bends of the arms, and ankles by using sensors made of electromechanical film (EMFi) as well as blood-volume related PWs from the index finger, the second toe and the forehead by using photoplethysmographic (PPG) transducers. In addition to the PW signals, the sensor network enables also the ECG and respiration measurements. At the analysis point of view, the main focus of the thesis is in the pulse wave decomposition (PWD) analysis. The goal of the PWD is to decompose an individual PW into a set of highly non-linear basis functions and these resulting components represent the forward PW and its reflections. In this study, three different sets of totally five Gaussian and logarithmic-normal shaped basis functions are tested and their performances are compared in the terms of residual error between the decomposition result and the measured PW, the ratio of successful PWD results and beat-to-beat variability. The parameters of the basis functions are found by using Levenberg-Marquardt algorithm and the challenge is to find a set of parameters that forms both physiologically and mathematically relevant fit. The PWDs are computed for over 20000 individual PWs that are registered from eight different test subjects. Besides the PWD analysis, other parameters presented in literature such as AGI (aging index), pAIx (peripheral augmentation index), SI (stiffness index) and RI (reflection index) are calculated with automatic methods for the recorded PW contours.
Results: Both visual and numerical comparison of the results from three tested PWD methods reveal differences in the performances of the implemented methods. The mean absolute error between the results of the decomposition and the measured PW contour varies in the range of 0.81%–1.09% and proportion of the successful PWD results varies from 88.6% to 89.8% depending on the set of basis functions. These differences are small, but the beat-tobeat variability of the decomposition results shows more clear differences: the set of five 4-parametric logarithmicnormal functions produces the smallest the beat-to-beat variability for the peak locations of the components. The other tested methods produced often roughly 50% higher variation for the peak locations of the components, both in amplitude and time scale. The analysis of more traditional indices, such as AGI, pAIx, RI and SI shows wide beat-to-beat variations in each parameter value. An interesting observation is that not only values of the traditional indices, but also the results of PWD are often dependent on the respiration, indicating that PW decomposition may reflect the state of vascular system which is affected also by the respiration.
Conclucions: The implemented automatic analysis methods enable using of a large number of PWs in order to obtain a more comprehensive view on the vascular state which is a benefit compared with the analysis of single PWs that can be affected by various physiological factors. However, to find real applications for the PWD and find new PW- and PWD-related parameters, measurements with subjects suffering from vascular diseases are planned in future.
Heart rate variability based study on sleep recovery in shift working truck drivers
Paruthi Pradhapan, Tampereen teknillinen yliopisto
Prolonged work hours, shortened and irregular sleep patterns often leads to inadequate recovery in shift workers resulting in increased sleepiness or fatigue during the day. Heart rate and heart rate variability (HRV) have been often used in occupational health studies to examine sleep quality and recovery. The aim of the current study was to determine the factors affecting the recovery process in shift working long-haul truck drivers and to assess the impact different shifts have on the drivers’ sleep health.
Of the recruited volunteers, data collected from 38 volunteers (Age: 38.46 ± 10.89 years) satisfied the inclusion criteria for this study. Driver demographics and background questionnaires were obtained prior to measurements. R-R intervals and actigraphy data were collected for three intensive measurement days (non-night shift, night shift and leisure day) and subjective measures of sleep quality, recorded on the sleep-diary, were used for the analyses. Several time- and frequency-domain HRV indices were calculated in 10-minute segments and averaged on an hourly basis and for the entire duration of sleep. All tests for statistical significance were conducted on a within-subject basis.
Comparison of HRV indices over the entire sleep duration recorded on different intensive measurement days revealed no significant differences except for LF/HF ratio (Leisure day vs. Night shift, p <0.05). Sleep duration and efficiency were significantly lower on duty days. Regression analyses indicated VLF power was strong predictor of recovery and 31% of the outcome was influenced by explanatory factors. SDNN (r = 0.555, adjusted r2 = 0.248, F(9, 92) = 5.166, p <0.001), RMSSD (r = 0.414, adjusted r2 = 0.131, F(9.92) = 4.229, p <0.05) and HF power (r = 0.460, adjusted r2 = 0.165, F(9.92) = 4.526, p <0.001) were significantly associated with age and sleep duration. Short-term variability indices, RMSSD and HF power, were moderately influenced by diurnal variations.
The results suggest that despite the fact that shift type does not have any direct consequences on sleep recovery, the odd work hours and irregular sleep schedules pose an indirect effect. The truncated sleep length, especially seen after night shift work, have been significantly associated with the impaired recovery and is contributed to by other short-term (diurnal variations) and long-term (ageing) factors. These results provide a basis for planning shift schedules such that direct or indirect manifestations of shift type related influence on recovery are mitigated.
Acknowledgement: Thesis was part of project An educational intervention to promote safe and economic truck driving. Mia Pylkkönen, Maria Sihvola, and Mikael Sallinen from Finnish Institute of Occupational Health have participated in the data collection and processing.
Metal Artifact Reduction in Sinograms of Dental Computed Tomography
Defne Us, Tampereen teknillinen yliopisto
Prolonged work hours, shortened and irregular sleep patterns often leads to inadequate recovery in shift workers resulting in increased sleepiness or fatigue during the day. Heart rate and heart rate variability (HRV) have been often used in occupational health studies to examine sleep quality and recovery. The aim of the current study was to determine the factors affecting the recovery process in shift working long-haul truck drivers and to assess the impact different shifts have on the drivers’ sleep health.
Of the recruited volunteers, data collected from 38 volunteers (Age: 38.46 ± 10.89 years) satisfied the inclusion criteria for this study. Driver demographics and background questionnaires were obtained prior to measurements. R-R intervals and actigraphy data were collected for three intensive measurement days (non-night shift, night shift and leisure day) and subjective measures of sleep quality, recorded on the sleep-diary, were used for the analyses. Several time- and frequency-domain HRV indices were calculated in 10-minute segments and averaged on an hourly basis and for the entire duration of sleep. All tests for statistical significance were conducted on a within-subject basis.
Comparison of HRV indices over the entire sleep duration recorded on different intensive measurement days revealed no significant differences except for LF/HF ratio (Leisure day vs. Night shift, p <0.05). Sleep duration and efficiency were significantly lower on duty days. Regression analyses indicated VLF power was strong predictor of recovery and 31% of the outcome was influenced by explanatory factors. SDNN (r = 0.555, adjusted r2 = 0.248, F(9, 92) = 5.166, p <0.001), RMSSD (r = 0.414, adjusted r2 = 0.131, F(9.92) = 4.229, p <0.05) and HF power (r = 0.460, adjusted r2 = 0.165, F(9.92) = 4.526, p <0.001) were significantly associated with age and sleep duration. Short-term variability indices, RMSSD and HF power, were moderately influenced by diurnal variations.
The results suggest that despite the fact that shift type does not have any direct consequences on sleep recovery, the odd work hours and irregular sleep schedules pose an indirect effect. The truncated sleep length, especially seen after night shift work, have been significantly associated with the impaired recovery and is contributed to by other short-term (diurnal variations) and long-term (ageing) factors. These results provide a basis for planning shift schedules such that direct or indirect manifestations of shift type related influence on recovery are mitigated.
Importance of material properties and structure of bone in biomechanical modeling of knee joint function
Mikko Venäläinen, Itä-Suomen yliopisto
Background: Mechanical function of knee joint in various loading and pathological conditions such as osteoarthritis (OA) can be evaluated with the aid of finite element (FE) modeling. The modeling outcomes are highly dependent on the material properties implemented for individual tissues and therefore their applicability should be carefully considered before drawing any conclusions about the functional integrity of the joint. Although there has been significant progress in developing constitutive material models for individual tissues such as articular cartilage and meniscus, the articulating bones in the knee joint, femur and tibia, are typically modeled as completely rigid bodies or simple solid structures with linear elastic material behavior. The aim of this study was to evaluate how different structural and material properties of bone affect the estimated mechanical response of articular cartilage in human knee joint as compared to results obtained using rigid bones.
Methods: Based on a cadaver knee joint, a two-dimensional FE model of a knee joint including bone, cartilage and meniscus geometries was constructed. By combining a set of mechanical and structural properties for cortical, trabecular and subchondral bone, a total of six variations of the knee joint model were created. The simplest model included rigid bones, while the most complex included specific mechanical properties for different bone structures and anatomically accurate structure for trabecular bone. Models with different porosities of trabecular bone were also constructed. In all models, cartilages and menisci were modeled as biphasic fibril-reinforced tissues and their properties were kept unaltered throughout analysis. The mechanical behaviour of all model variations was simulated under axial impact loading of 700 N.
Results: As compared to results obtained with rigid bone, stresses, strains and pore pressures observed in tibial cartilage decreased depending on the implemented properties of trabecular bone. The most prominent changes in these parameters were observed using the softest elastic modulus for trabecular bone (measured for macroscopic samples) which showed decreases of up to 62 %, 31 % and 32 % in mean values of maximum principal stresses, strains and pore pressures, respectively. By including the trabecular architecture and using material properties measured at microscopic level, these differences were only partially reproduced. Further increase in trabecular bone porosity led to improved congruence between results. However, substantial differences on different sides of the joint (i.e. lateral and medial) were still present.
Conclusion: The present results demonstrate the importance of long bones and specifically the essential feature of trabecular bone to substantially dampen forces in knee joints under impact loading. This study also suggests that the shock absorption capacity of bone becomes more non-uniform along with the increased bone porosity which indicates that in order to construct accurate patient-specific models and analyze failure points from estimated stress values, structural information of bones is also required to be implemented in the models..
Binding of Hyaluronic Acid to Its CD44 Receptor
Joni Vuorio, Tampereen teknillinen yliopisto
CD44 is a transmembrane glycoprotein that binds hyaluronic acid (HA), its main carbohydrate ligand, in a reversible fashion [1]. In addition to enabling normal cell migration, such as the rolling of white blood cells, CD44-HA interaction is exploited by malignant cancer cells metastasizing through the blood stream [2]. A normal cell hence requires effective regulatory mechanisms for controlling the binding affinity between CD44 and HA. Earlier studies addressing this topic [3, 4] have however been unable to unequivocally identify these regulatory mechanisms especially in an atomic level.
We use all-atom explicit-solvent molecular dynamics (MD) simulations to study the adsorption of a HA oligomer to a human wild-type CD44 HA binding domain (HABD). In practice, we explore the role of three potential regulation mechanisms: size of the ligand, glycosylation of the protein, and conformation of the protein. First, free energy profiles for the adsorption of a HA octamer reveal the strength of an individual CD44-HA interaction to be over 25 kJ mol??1, thereby suggesting ligand binding to be irreversible. Second, by glycosylating the HABD at residues Asn25, Asn100, and Asn110 independently, we show that only the first glycosylation site blocks most of the native binding interactions. In this case, glycosylating the Asn25 residue with a charge-neutral core pentasaccharide reduces the strength of the adsorption by 40 %. More strikingly, our simulation data reveals that a conformation change in CD44 previously reported to improve the binding affinity with HA is, in fact, a molecular mechanism repelling the bound HA oligomer, and thereby dynamically regulating the biological activity of CD44.
The findings of this study unlock how the binding of HA to its CD44 receptor is regulated. This information may facilitate the design and targeting of novel drugs and therapies against severe conditions, such as cancer. Lastly, the insight from this study is of potential value when considering carbohydrate-protein interactions of other cell surface receptors.
[1] H. Ponta, L. Sherman, and P. Herrlich, Nature Reviews Molecular Cell Biology 4, 33–45 (2003). [2] M. Zöller, Nature Reviews Cancer 11, 254–267 (2011). [3] F. Jamison II, T. Foster, J. Barker, R. Hills Jr, and O. Guvench, Journal of Molecular Biology 406, 631–647 (2011). [4] A. Favreau, C. Faller, and O. Guvench, Biophysical Journal 105, 1217–1226 (2013)..
Construction and testing of a PET demostrator
Tiziana Zedda, Tampereen teknillinen yliopisto
a) Aim of the study
In this poster we present the Avantomography demonstrator, which is being implemented and tested at Tampere University of Technology (Tampere, Finland). We also describe the first tests performed with it and the obtained results. The final aim of this master thesis was the energy calibration of the scintillating crystals and the electronic chain for data acquisition.
b) Methods used
This new small Positron Emission Tomography (PET) demonstrator follows the recent innovations presented by the AX-PET group [1], at CERN. This is the first functioning version of a prototype of a light and compact PET scanner. The novel geometry, used to build the Avantomography demonstrator, is based on scintillating crystals and wavelength shifting (WLS) plastic strips, allowing high resolution and high sensitivity at the same time. The device consists in two small and compact modules, with two different adjustable parts inside. Each detector module is built up from long scintillator bars placed in the trans-axial plane and orthogonal WLS strip arrays. Preliminary tests with a standard positron emitter source has been performed in order to test the acquisition chain and to calibrate the demonstrator. First test has been performed measuring the intrinsic radioactivity of the scintillating crystals. For a complete calibration of one crystal, a test with a linear positron emitter source has been performed at the Tampere University Hospital (Tampere, Finland).
c) Results
From these measurements the spectra of different energy peaks are acquired and plotted. Using a dedicate MATLAB code, different Gaussian fits are calculated to find the position of each peak. With these values a 3-parameters fitting curve has been evaluated in order to obtain the nonlinear curve for the energy calibration. Furthermore a first evaluation of the energy resolution has been calculated starting from the acquired data.
d) Conclusions
The Avantomography demonstrator is a functioning prototype of the new concept of new small PET cameras. This device implemented during this thesis work consists in two identical small, light and compact modules, with high resolution and high sensitivity at the same time, unlike the standard total body PET cameras. The design, the implementation and the characterization of the demonstrator derive from the AX-PET collaboration [1] ideas. The scintillating crystals have been characterized in terms of energy calibration and energy resolution. With its compact and light geometry, high resolution and high sensitivity this detector has a promising layout as a preclinical PET scanner.
We can be reasonably satisfied for the results obtained with this first version of the demonstrator, but we consider necessary some further structural improvement and some additional tests. A customized electronics for the acquisition is needed. It may be also possible to enhance the design, for example increasing the number of scintillating bars and the number of WLS strips. Moreover a higher number of LYSO bars will have the effect of increasing the efficiency on revealing photons, and therefore it would be not just a better design but a better performance of the device as well.