Development of model observers applied to 3D breast tomosynthesis microcalcifications and masses

The development of model observers for mimicking human detection strategies has followed from symmetric signals in simple noise to increasingly complex backgrounds. In this study we implement different model observers for the complex task of detecting a signal in a 3D image stack. The backgrounds come from real breast tomosynthesis acquisitions and the signals were simulated and reconstructed within the volume. Two different tasks relevant to the early detection of breast cancer were considered: detecting an 8 mm mass and detecting a cluster of microcalcifications. The model observers were calculated using a channelized Hotelling observer (CHO) with dense difference-of-Gaussian channels, and a modified (Partial prewhitening [PPW]) observer which was adapted to realistic signals which are not circularly symmetric. The sustained temporal sensitivity function was used to filter the images before applying the spatial templates. For a frame rate of five frames per second, the only CHO that we calculated performed worse than the humans in a 4-AFC experiment. The other observers were variations of PPW and outperformed human observers in every single case. This initial frame rate was a rather low speed and the temporal filtering did not affect the results compared to a data set with no human temporal effects taken into account. We subsequently investigated two higher speeds at 5, 15 and 30 frames per second. We observed that for large masses, the two types of model observers investigated outperformed the human observers and would be suitable with the appropriate addition of internal noise. However, for microcalcifications both only the PPW observer consistently outperformed the humans. The study demonstrated the possibility of using a model observer which takes into account the temporal effects of scrolling through an image stack while being able to effectively detect a range of mass sizes and distributions.

A stratified approach for modeling the distribution of a threatened ant species in the Swiss National Park

We present models predicting the potential distribution of a threatened ant species, Formica exsecta Nyl., in the Swiss National Park ( SNP). Data to fit the models have been collected according to a random-stratified design with an equal number of replicates per stratum. The basic aim of such a sampling strategy is to allow the formal testing of biological hypotheses about those factors most likely to account for the distribution of the modeled species. The stratifying factors used in this study were: vegetation, slope angle and slope aspect, the latter two being used as surrogates of solar radiation, considered one of the basic requirements of F. exsecta. Results show that, although the basic stratifying predictors account for more than 50% of the deviance, the incorporation of additional non-spatially explicit predictors into the model, as measured in the field, allows for an increased model performance (up to nearly 75%). However, this was not corroborated by permutation tests. Implementation on a national scale was made for one model only, due to the difficulty of obtaining similar predictors on this scale. The resulting map on the national scale suggests that the species might once have had a broader distribution in Switzerland. Reasons for its particular abundance within the SNP might possibly be related to habitat fragmentation and vegetation transformation outside the SNP boundaries.

A 2D/3D image analysis system to track fluorescently labeled structures in rod-shaped cells: application to measure spindle pole asymmetry during mitosis.

BACKGROUND: The yeast Schizosaccharomyces pombe is frequently used as a model for studying the cell cycle. The cells are rod-shaped and divide by medial fission. The process of cell division, or cytokinesis, is controlled by a network of signaling proteins called the Septation Initiation Network (SIN); SIN proteins associate with the SPBs during nuclear division (mitosis). Some SIN proteins associate with both SPBs early in mitosis, and then display strongly asymmetric signal intensity at the SPBs in late mitosis, just before cytokinesis. This asymmetry is thought to be important for correct regulation of SIN signaling, and coordination of cytokinesis and mitosis. In order to study the dynamics of organelles or large protein complexes such as the spindle pole body (SPB), which have been labeled with a fluorescent protein tag in living cells, a number of the image analysis problems must be solved; the cell outline must be detected automatically, and the position and signal intensity associated with the structures of interest within the cell must be determined. RESULTS: We present a new 2D and 3D image analysis system that permits versatile and robust analysis of motile, fluorescently labeled structures in rod-shaped cells. We have designed an image analysis system that we have implemented as a user-friendly software package allowing the fast and robust image-analysis of large numbers of rod-shaped cells. We have developed new robust algorithms, which we combined with existing methodologies to facilitate fast and accurate analysis. Our software permits the detection and segmentation of rod-shaped cells in either static or dynamic (i.e. time lapse) multi-channel images. It enables tracking of two structures (for example SPBs) in two different image channels. For 2D or 3D static images, the locations of the structures are identified, and then intensity values are extracted together with several quantitative parameters, such as length, width, cell orientation, background fluorescence and the distance between the structures of interest. Furthermore, two kinds of kymographs of the tracked structures can be established, one representing the migration with respect to their relative position, the other representing their individual trajectories inside the cell. This software package, called "RodCellJ", allowed us to analyze a large number of S. pombe cells to understand the rules that govern SIN protein asymmetry. CONCLUSIONS: "RodCell" is freely available to the community as a package of several ImageJ plugins to simultaneously analyze the behavior of a large number of rod-shaped cells in an extensive manner. The integration of different image-processing techniques in a single package, as well as the development of novel algorithms does not only allow to speed up the analysis with respect to the usage of existing tools, but also accounts for higher accuracy. Its utility was demonstrated on both 2D and 3D static and dynamic images to study the septation initiation network of the yeast Schizosaccharomyces pombe. More generally, it can be used in any kind of biological context where fluorescent-protein labeled structures need to be analyzed in rod-shaped cells. AVAILABILITY: RodCellJ is freely available under http://bigwww.epfl.ch/algorithms.html, (after acceptance of the publication).

Temporal SNR characteristics in segmented 3D-EPI at 7T.

Three-dimensional segmented echo planar imaging (3D-EPI) is a promising approach for high-resolution functional magnetic resonance imaging, as it provides an increased signal-to-noise ratio (SNR) at similar temporal resolution to traditional multislice 2D-EPI readouts. Recently, the 3D-EPI technique has become more frequently used and it is important to better understand its implications for fMRI. In this study, the temporal SNR characteristics of 3D-EPI with varying numbers of segments are studied. It is shown that, in humans, the temporal variance increases with the number of segments used to form the EPI acquisition and that for segmented acquisitions, the maximum available temporal SNR is reduced compared to single shot acquisitions. This reduction with increased segmentation is not found in phantom data and thus likely due to physiological processes. When operating in the thermal noise dominated regime, fMRI experiments with a motor task revealed that the 3D variant outperforms the 2D-EPI in terms of temporal SNR and sensitivity to detect activated brain regions. Thus, the theoretical SNR advantage of a segmented 3D-EPI sequence for fMRI only exists in a low SNR situation. However, other advantages of 3D-EPI, such as the application of parallel imaging techniques in two dimensions and the low specific absorption rate requirements, may encourage the use of the 3D-EPI sequence for fMRI in situations with higher SNR.

Creación de una demo 3D mediante gráficos procedurales

Aquest projecte consisteix en el desenvolupament d’una demo 3D utilitzant exclusivament gràfics procedurals per tal d’avaluar la seva viabilitat en aplicacions més complexes com els videojocs. En aquesta aplicació es genera un terreny aleatori explorable amb vegetació i textures creades proceduralment.

Modeling the biophysical impacts of global change in mountain biosphere reserves

Mountains and mountain societies provide a wide range of goods and services to humanity, but they are particularly sensitive to the effects of global environmental change. Thus, the definition of appropriate management regimes that maintain the multiple functions of mountain regions in a time of greatly changing climatic, economic, and societal drivers constitutes a significant challenge. Management decisions must be based on a sound understanding of the future dynamics of these systems. The present article reviews the elements required for an integrated effort to project the impacts of global change on mountain regions, and recommends tools that can be used at 3 scientific levels (essential, improved, and optimum). The proposed strategy is evaluated with respect to UNESCO's network of Mountain Biosphere Reserves (MBRs), with the intention of implementing it in other mountain regions as well. First, methods for generating scenarios of key drivers of global change are reviewed, including land use/land cover and climate change. This is followed by a brief review of the models available for projecting the impacts of these scenarios on (1) cryospheric systems, (2) ecosystem structure and diversity, and (3) ecosystem functions such as carbon and water relations. Finally, the cross-cutting role of remote sensing techniques is evaluated with respect to both monitoring and modeling efforts. We conclude that a broad range of techniques is available for both scenario generation and impact assessments, many of which can be implemented without much capacity building across many or even most MBRs. However, to foster implementation of the proposed strategy, further efforts are required to establish partnerships between scientists and resource managers in mountain areas.

Kinetics of crystal growth and equilibrium domains in eclogite of the Sesia Zone, Western Alps

Kinetics of crystal growth and equilibrium domains in eclogite of the Sesia Zone, Western Alps Darbellay Bastien Institut de Minéralogie et Géochimie Résumé grand public Comme toute matière, la roche est sensible à son environnement et cherche à s'adapter pour acquérir un état stable (état d'équilibre). Les changements des conditions physiques (température et pression) vont ainsi impliquer des modifications dans la roche. Le métamorphisme est l'étude de ces changements. Les minéraux qui constituent la roche peuvent modifier, leur structure, leur chimie ou être remplacer par d'autres minéraux plus stables. Il est ainsi crucial de déterminer les processus responsables et limitant de la croissance minérale. Trois processus permettent la croissance ; (1) la dissolution des éléments du réactant, (2) le transport de ces éléments vers le site de croissance, (3) l'incorporation de ces éléments dans la nouvelle structure. Cette thèse se focalise sur les structures des minéraux de haute pression (forme, zonation chimique, structure interne) pour essayer de déterminer les facteurs importants à l'origine de leur état final. Les zones d'étude se situent dans la zone de Sésia. La première partie traite de la problématique liée à l'incorporation d'un élément dans une structure minérale. A l'image de la croissance humaine, les irrégularités minéralogiques permettent de mettre en lumière un dysfonctionnement de la croissance due à un excès ou à une carence d'un élément. Bien dosé, cet élément est cependant essentiel à la croissance. Les zoisites (épidotes) des métabasites de la région de Cima di Bonze montrent une zonation chimique en sablier. Dans cette zonation la teneur en fer excède la capacité maximum que peut contenir la structure orthorhombique de la zoisite. Des défauts de structure permettent l'accommodation de cet excès. La zoisite peut ainsi adapter sa structure pour permettre l'incorporation d'une relativement grande quantité de fer. Les études précédentes montraient, pour des conditions similaires, la formation de deux épidotes distinctes. La deuxième partie se penche sur la compétition entre le minéral qui fait sa croissance et les minéraux (réactants) qui l'entourent. Les métapélites de la région du Monte Mucrone contiennent des grenats atollaires. Des études détaillées de la texture et de la zonation chimique du grenat ainsi qu'une modélisation thermodynamique ont permis de mieux cerner les facteurs importants responsables de la forme atollaire. Cette structure est obtenue par un changement du comportement de la croissance du grenat le long d'un chemin P-T hercynien. Dans un premier stade, le grenat croît rapidement et consume peu le quartz de la matrice. La croissance se fait ainsi le long des jointures des grains de quartz. Dans un second temps, les changements de conditions PT donnent une croissance lente du grenat et une forte consommation du quartz. Le grenat peut ainsi développer sa forme dodécaédrale classique. La troisième partie s'intéresse aux distances de transport par diffusion d'un élément (ici l'argon) durant la haute pression. Pour ce faire, un profile d'âges 40Ar/39Ar sur biotite a été mesuré depuis un veine de haute pression riche en argon jusque dans son encaissant (granitoïd du Monte Mucrone). Le profile montre une répartition des âges suivant une courbe de diffusion. Le transport se fait sur une longueur de deux centimètre avec l'aide d'un fluide. Il est réduit à une échelle millimétrique quand la phase fluide disparaît. Cette étude montre ainsi les difficultés de transport des éléments durant la haute pression ne permettant pas un rééquilibrage de la roche à grande échelle. Kinetics of crystal growth and equilibrium domains in eclogite of the Sesia Zone, Western Alps Darbellay Bastien Institut de Minéralogie et Géochimie Résumé de thèse Les processus de croissance (diffusion des éléments et les réactions d'interface) et les conditions dans lesquelles les minéraux grandissent (température, pression, fluide, composition chimique de la roche), déterminent la texture ainsi que la zonation des minéraux. Cette thèse se focalise, par le biais de textures peu communes, sur trois différents processus impliqués dans la croissance minérale à haute pression (Zone de Sésia, Alpes de l'Ouest, Italie). L'incorporation d'un élément dans une structure minérale ne peut se faire que dans des sites en accord avec la taille et la charge ionique de l'élément. De plus, la balance de charge doit être maintenue dans le minéral. La régularité de la structure cristalline fixe ainsi une limite maximum de concentration d'un élément donné. Les zoisites provenant des métabasites de la région de Cima di Bonze montrent des zonations en sablier caractérisées par une concentration anormale en fer. La zonation se marque par une différente teinte de biréfringence et par un plus grand angle d'extinction que le reste de la zoisite. Une inter-croissance de clinozoisite à l'intérieur de la structure orthorhombique de la zoisite peut ainsi être suspectée. Les analyses XRD (diffraction des rayons x) ainsi que les analyses Raman ne confirment pas cette suspicion. Seules les analyses TEM (microscope à électrons transmis) montrent des défauts de structure pouvant être interprétés comme des modules de clinozoisite. Ils ne peuvent cependant pas être considérés comme une phase thermodynamique. Un nouveau trou d'immiscibilité entre deux zoisite (X ep= 0.1 and Xep = 0.15) a ainsi pu être établi. Dans les métapélites la région du Monte Mucrone, des grenats fortement zonés montrent une évolution texturale singulière. Ils présentent une forme initiale de `champignon' qui se développe pour former une structure atollaire finale. L'étude conjuguée de la structure 3D et des zonations, ainsi que l'établissement d'un model thermodynamique, indiquent que ces structures proviennent de deux épisodes de croissances : (1) La croissance du grenat durant un chemin prograde hercynien (de 525 °C et 6.2 kbar à 640 °C et 9 kbar) permet la formation des textures atollaires. Elles sont le résultat d'une croissance poecilitique initiale suivie d'une croissance idiomorphique du grenat. (2) La structure est rendue plus complexe par la cristallisation d'un grenat homogène tout autour ainsi qu'à l'intérieur du grenat hercynien durant la haute pression alpine (550 °C and 20 kbar). L'arrivée de l'eau durant la haute pression facilite le transport d'éléments et permet une cristallisation rapide du grenat. La diffusion peut être un facteur limitant de la croissance minéralogique. Elle a aussi une grande importance pour la géochronologie. Une veine de haute pression à l'intérieur du granitoïde du Monte Mucrone a été étudiée dans le but de déterminer la distance de diffusion de l'argon. Le profile d'âges 40Ar/39Ar sur biotites, établi de la veine vers le métagranitoïde, suit une courbe de diffusion. Les âges sont élevés proche de la veine (800 Ma) puis décroissent jusqu'à des âges homogènes (170-150 Ma) à deux centimètres de la veine. La présence de fluide, marqué par de hautes concentrations en chlore, permet une diffusion centimétrique. Cependant, la distance est réduite à une échelle millimétrique quand le fluide est absent. Les très faibles distances de diffusion préservent les âges pré-alpins et impliquent un événement géologique pour les âges de 170-150 Ma. Kinetics of crystal growth and equilibrium domains in eclogite of the Sesia Zone, Western Alps Darbellay Bastien Institut de Minéralogie et Géochimie Thesis abstract Rock textures and zonings are the consequence of growth processes (element diffusion and interface reaction) steered by the environment in which they grew (pressure, differential stress, temperature, fluid and rock composition). The thesis presented here focuses on three different topics, each of it dealing with aspects of mineral growth processes during subduction, in a high-pressure environment. All studies were conducted in the Sesia Zone of the Western European Alps, Italy. The first study addresses the crystallography and geochemistry of element incorporation in zoisite, one of the major hydrous minerals found in subduction zone rocks. Elements can be incorporated into a mineral structure only on crystallographic sites that offer enough space for the ion and the overall charge balance has to be maintained. Element concentrations are hence limited. Incorporation of some elements produces complex zoning, including hourglass like patterns, which are the focus of the first contribution. Zoisites from Cima di Bonze (Sesia Zone) show spectacular hourglass zoning defined by Fe-content variations. The hourglass zones have a distinct birefringence and a different extinction angle than the regular part of the zoisite. We show by detailed XRD (X-ray diffraction) and confocal Raman analyses that the high Fe-zones are nevertheless zoisite, and not clinozoisite as one might expect. High resolution TEM (transmission electron microscopy) analyses show planar defects on (100) that can be interpreted as small-scale clinozoisite modules. However, these clinozoisites cannot be interpreted as a distinctive thermodynamic phase and the entire mineral has to be considered as zoisite. The miscibility gap between two zoisites (Xep = 0.1 and Xep = 0.15) can be then definite at 550 ± 50°C and 14 to 20 Kbar. Strongly zoned garnets in quartz rich metapelite from the Monte Mucrone area (Sesia Zone) show evolution form 3D mushroom to atoll structure. The second contribution presents textural investigations, garnet zoning and thermodynamic modeling that demonstrate that atoll garnets are the result of two distinctive growth events. (1) Garnet atoll structure is already formed during a prograde Hercynian path from 525 °C and 6.2 kbar to 640 °C and 9 kbar. It results in an initial poikilitic growth followed by a final idiomorphic growth event. (2) Alpine HP garnet are homogenous (550 °C and 20 kbar) and grew around and also inside the Hercynian garnet. Lack of prograde Alpine garnet and fast growth of the HP garnet is explained by the absence of water during much of the prograde path. Water saturation was only observed towards the end, close towards the peak metamorphic conditions. Diffusion could be a limiting factor for crystal growth. It has also a great importance in geochronology. HP vein inside the metagranitoide of the Monte Mucrone (~300 Ma) was investigated to determine argon diffusion scales during high-pressure metamorphism. 40Ar/39Ar biotite ages profile from the vein toward the metagranodiorite show a diffusion curve: old ages (800 Ma) located close to the vein decrease until homogenous 170-150 Ma ages are obtained, two centimeter away from the vein. Centimeter-scale diffusion occurs with help of a fluid phase marked by high chlorine concentrations. Argon diffusion is reduced to a millimeter scale when free fluid is absent. Very short diffusion distance permits to preserve pre-Alpine ages. The 170-150 Ma ages are considered to be geologic meaningful, probably resulting from the extensional tectonics linked to opening of the Tethian ocean.

Intracellular nanomanipulation by a photonic-force microscope with real-time acquisition of a 3D stiffness matrix.

A traditional photonic-force microscope (PFM) results in huge sets of data, which requires tedious numerical analysis. In this paper, we propose instead an analog signal processor to attain real-time capabilities while retaining the richness of the traditional PFM data. Our system is devoted to intracellular measurements and is fully interactive through the use of a haptic joystick. Using our specialized analog hardware along with a dedicated algorithm, we can extract the full 3D stiffness matrix of the optical trap in real time, including the off-diagonal cross-terms. Our system is also capable of simultaneously recording data for subsequent offline analysis. This allows us to check that a good correlation exists between the classical analysis of stiffness and our real-time measurements. We monitor the PFM beads using an optical microscope. The force-feedback mechanism of the haptic joystick helps us in interactively guiding the bead inside living cells and collecting information from its (possibly anisotropic) environment. The instantaneous stiffness measurements are also displayed in real time on a graphical user interface. The whole system has been built and is operational; here we present early results that confirm the consistency of the real-time measurements with offline computations.

Genetic and genomic analysis modeling of germline c-MYC overexpression and cancer susceptibility

Background: Germline genetic variation is associated with the differential expression of many human genes. The phenotypic effects of this type of variation may be important when considering susceptibility to common genetic diseases. Three regions at 8q24 have recently been identified to independently confer risk of prostate cancer. Variation at 8q24 has also recently been associated with risk of breast and colorectal cancer. However, none of the risk variants map at or relatively close to known genes, with c-MYC mapping a few hundred kilobases distally. Results: This study identifies cis-regulators of germline c-MYC expression in immortalized lymphocytes of HapMap individuals. Quantitative analysis of c-MYC expression in normal prostate tissues suggests an association between overexpression and variants in Region 1 of prostate cancer risk. Somatic c-MYC overexpression correlates with prostate cancer progression and more aggressive tumor forms, which was also a pathological variable associated with Region 1. Expression profiling analysis and modeling of transcriptional regulatory networks predicts a functional association between MYC and the prostate tumor suppressor KLF6. Analysis of MYC/Myc-driven cell transformation and tumorigenesis substantiates a model in which MYC overexpression promotes transformation by down-regulating KLF6. In this model, a feedback loop through E-cadherin down-regulation causes further transactivation of c-MYC.Conclusion: This study proposes that variation at putative 8q24 cis-regulator(s) of transcription can significantly alter germline c-MYC expression levels and, thus, contribute to prostate cancer susceptibility by down-regulating the prostate tumor suppressor KLF6 gene.

Patient-specific computational hemodynamics of intracranial aneurysms from 3D rotational angiography and CT angiography: an in vivo reproducibility study

Patient-specific simulations of the hemodynamics in intracranial aneurysms can be constructed by using image-based vascular models and CFD techniques. This work evaluates the impact of the choice of imaging technique on these simulations

Temporal diffeomorphic free-form deformation: application to motion and strain estimation from 3D echocardiography

This paper presents a new registration algorithm, called Temporal Di eomorphic Free Form Deformation (TDFFD), and its application to motion and strain quanti cation from a sequence of 3D ultrasound (US) images. The originality of our approach resides in enforcing time consistency by representing the 4D velocity eld as the sum of continuous spatiotemporal B-Spline kernels. The spatiotemporal displacement eld is then recovered through forward Eulerian integration of the non-stationary velocity eld. The strain tensor iscomputed locally using the spatial derivatives of the reconstructed displacement eld. The energy functional considered in this paper weighs two terms: the image similarity and a regularization term. The image similarity metric is the sum of squared di erences between the intensities of each frame and a reference one. Any frame in the sequence can be chosen as reference. The regularization term is based on theincompressibility of myocardial tissue. TDFFD was compared to pairwise 3D FFD and 3D+t FFD, bothon displacement and velocity elds, on a set of synthetic 3D US images with di erent noise levels. TDFFDshowed increased robustness to noise compared to these two state-of-the-art algorithms. TDFFD also proved to be more resistant to a reduced temporal resolution when decimating this synthetic sequence. Finally, this synthetic dataset was used to determine optimal settings of the TDFFD algorithm. Subsequently, TDFFDwas applied to a database of cardiac 3D US images of the left ventricle acquired from 9 healthy volunteers and 13 patients treated by Cardiac Resynchronization Therapy (CRT). On healthy cases, uniform strain patterns were observed over all myocardial segments, as physiologically expected. On all CRT patients, theimprovement in synchrony of regional longitudinal strain correlated with CRT clinical outcome as quanti ed by the reduction of end-systolic left ventricular volume at follow-up (6 and 12 months), showing the potential of the proposed algorithm for the assessment of CRT.

Dynamic estimation of three-dimensional cerebrovascular deformation from rotational angiography

Purpose: The objective of this study is to investigate the feasibility of detecting and quantifying 3D cerebrovascular wall motion from a single 3D rotational x-ray angiography (3DRA) acquisition within a clinically acceptable time and computing from the estimated motion field for the further biomechanical modeling of the cerebrovascular wall. Methods: The whole motion cycle of the cerebral vasculature is modeled using a 4D B-spline transformation, which is estimated from a 4D to 2D + t image registration framework. The registration is performed by optimizing a single similarity metric between the entire 2D + t measured projection sequence and the corresponding forward projections of the deformed volume at their exact time instants. The joint use of two acceleration strategies, together with their implementation on graphics processing units, is also proposed so as to reach computation times close to clinical requirements. For further characterizing vessel wall properties, an approximation of the wall thickness changes is obtained through a strain calculation. Results: Evaluation on in silico and in vitro pulsating phantom aneurysms demonstrated an accurate estimation of wall motion curves. In general, the error was below 10% of the maximum pulsation, even in the situation when substantial inhomogeneous intensity pattern was present. Experiments on in vivo data provided realistic aneurysm and vessel wall motion estimates, whereas in regions where motion was neither visible nor anatomically possible, no motion was detected. The use of the acceleration strategies enabled completing the estimation process for one entire cycle in 5-10 min without degrading the overall performance. The strain map extracted from our motion estimation provided a realistic deformation measure of the vessel wall. Conclusions: The authors' technique has demonstrated that it can provide accurate and robust 4D estimates of cerebrovascular wall motion within a clinically acceptable time, although it has to be applied to a larger patient population prior to possible wide application to routine endovascular procedures. In particular, for the first time, this feasibility study has shown that in vivo cerebrovascular motion can be obtained intraprocedurally from a 3DRA acquisition. Results have also shown the potential of performing strain analysis using this imaging modality, thus making possible for the future modeling of biomechanical properties of the vascular wall.

Personalized modeling for drug concentration prediction using Support Vector Machine

Building a personalized model to describe the drug concentration inside the human body for each patient is highly important to the clinical practice and demanding to the modeling tools. Instead of using traditional explicit methods, in this paper we propose a machine learning approach to describe the relation between the drug concentration and patients' features. Machine learning has been largely applied to analyze data in various domains, but it is still new to personalized medicine, especially dose individualization. We focus mainly on the prediction of the drug concentrations as well as the analysis of different features' influence. Models are built based on Support Vector Machine and the prediction results are compared with the traditional analytical models.