ebook img

POLITECNICO DI MILANO Generation of a patient specific predictor for osteoporotic risk of fracture PDF

132 Pages·2013·10.64 MB·English
by  
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview POLITECNICO DI MILANO Generation of a patient specific predictor for osteoporotic risk of fracture

POLITECNICO DI MILANO Facoltá di ingegneria dei sistemi Corso di Laurea Specialistica in Ingegneria Biomedica Generation of a patient specific predictor for osteoporotic risk of fracture of the femoral neck Relatore: Tomaso Villa Maria Tobia Correlatore: Marco Viceconti Relazione della prova finale di: Anna Caimi 770324 Gloria Casaroli 765186 Anno Accademico 2011/2012 Ringraziamenti Desideriamo porgere il primo sentito ringraziamento al Prof. Marco Viceconti per l’estrema disponibilitá e cortesia dimostrateci in questi mesi, per le grandi opportunitá che ci ha offerto, per la realtá che ci ha fatto conoscere e nella quale ci ha introdotto; un forte ringraziamento va al Prof. Tomaso Villa, che ci ha offerto la possibilitá di compiere un’esperienza di tesi all’estero che è stata la migliore della nostra carriera, sostenuto durante tutto il suo corso e ci ha sempre mostrato grande disponibilitá. Un ringraziamento speciale va a Giovanna Farinella, che ci ha seguito passo a passo nello svolgimento del nostro lavoro, fornendoci sempre grande aiuto e competenza, oltre che aver stretto con noi un forte legame di amicizia. Un grazie particolare va a tutti coloro che hanno condiviso con noi l’esperienza a Sheffield rendendola davvero speciale, e in particolare alle nostre colleghe Sandra e Francesca che sono state due perfette compagne d’avventura. Grazie anche a tutti i nostri compagni di corso per gli anni di studio, per i progetti di gruppo, per le risate e per tutti i momenti belli passati qui al Politecnico e durante la vita normale di tutti i giorni. Grazie anche a tutti gli amici di sempre ed alle persone noi care, per le serate, i week-end, per averci parlato di tutto fuorché dello studio, per averci capito, distratto e consolato con la loro simpatia e il loro affetto; in particolare grazie a Massimo e a Riccardo, che ci sono stati accanto piú di ogni altro e ci hanno sostenuto nonostante la grande distanza. Un ringraziamento particolare va infine alle nostre famiglie, per averci incoraggiato nelle scelte, sostenuto nei 5 anni, aver condiviso con noi i momenti di soddisfazione e consolato in quelli piú difficili e per averci regalato la possibilitá di compiere questa esperienza all’estero. Grazie, senza di voi, per noi, oggi sarebbe un giorno qualunque. Anna e Gloria iii Contents Contents v List of Figures vii List of Tables xi Abstract xiii Sommario xix 1 Introduction 1 1.1 Osteoporosis: social and economic burden . . . . . . . . . . . . . . . . . 1 1.2 The Clinical State of the Art . . . . . . . . . . . . . . . . . . . . . . . . 3 1.3 The VPHOP project . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1.4 The aim of our work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 2 Anatomy and Mechanics 13 2.1 Body’s reference system and planes . . . . . . . . . . . . . . . . . . . . . 13 2.2 Relative position . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 2.3 Musculoskeletal anatomy . . . . . . . . . . . . . . . . . . . . . . . . . . 15 2.3.1 Anatomy and physiology of bone . . . . . . . . . . . . . . . . . . 15 2.3.2 The pelvis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 2.3.3 The femur . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 2.3.4 The hip joint and angle . . . . . . . . . . . . . . . . . . . . . . . 22 2.4 Bone tissue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 2.4.1 Composition of bone . . . . . . . . . . . . . . . . . . . . . . . . . 25 2.4.2 Bone Cells. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 v vi CONTENTS 2.4.3 Bone mechanobiology . . . . . . . . . . . . . . . . . . . . . . . . 29 2.4.4 Effects of underloading and overloading . . . . . . . . . . . . . . 31 2.5 Hip Biomechanics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 2.5.1 Hip muscles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 2.5.2 Hip kinematic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 2.6 Definition and basic information . . . . . . . . . . . . . . . . . . . . . . 35 2.6.1 The importance of Gait analysis for this study . . . . . . . . . . 37 2.6.2 Effect of sub-optimal neuromotor control . . . . . . . . . . . . . 41 2.6.3 Hypothesis of our model . . . . . . . . . . . . . . . . . . . . . . . 43 3 Material and methods 45 3.1 Patients’ cohort and CT scanning . . . . . . . . . . . . . . . . . . . . . . 45 3.2 Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 3.2.1 Segmentation and morphing . . . . . . . . . . . . . . . . . . . . . 48 3.2.2 BoneMat software . . . . . . . . . . . . . . . . . . . . . . . . . . 53 3.2.3 Structure of msf file . . . . . . . . . . . . . . . . . . . . . . . . . 58 3.3 The mechanical load scenarios . . . . . . . . . . . . . . . . . . . . . . . . 64 3.3.1 Strength loading scenario . . . . . . . . . . . . . . . . . . . . . . 64 3.3.2 Femur-fall Charité database . . . . . . . . . . . . . . . . . . . . . 68 4 Results 71 5 Conclusion 87 6 Future development 91 Appendix 93 Descriptive statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 The Mann-Whitney test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 ROC curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Logistic regression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Bibliography 107 List of Figures 1.1 Image of the DXA of a femur. . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.2 FRAX interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.3 The graph shows fracture rates per 100 person-years by phenotypic and T- scoreaccordingtotheNationalOsteoporosisRiskAssessmentStudy. Across all phenotypic groups, low BMD is a consistent risk factor for fracture. . . . 6 1.4 Image representation of the VPHOP project. . . . . . . . . . . . . . . . . . 7 1.5 Biomed Town interface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 1.6 Physiomspace interface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 1.7 OpenClinica interface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1.8 VOP Hypermodel. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.1 Body planes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 2.2 Structure of long bone: external structure, epiphysis and diaphysis and a section of bone with a view of internal structure and component. . . . . . . 16 2.3 Distribution of compact and cancellous bone in the upper part of the femur. 17 2.4 Structureofthecancellousandcorticalbone: arepresentationoftrabeculae and of osteons. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 2.5 Maximum strength with minimum weight. . . . . . . . . . . . . . . . . . . . 20 2.6 Structure of the pelvis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 2.7 Arepresentationofforcesthatactonthepelvisfromfemurandfrompelvis to femur. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 2.8 Anterior and posterior view of the femur. . . . . . . . . . . . . . . . . . . . 22 2.9 The coxofemoral joint. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 2.10 CCD-angle in three configuration: coxa norma, coxa vara and coxa valga. . 25 vii viii List of Figures 2.11 Lightmicrographofosteoclasts(arrows). Typicalmultinucleatedosteoclast nestled in its Howship’s lacuna. Bone (B), calcified cartilage (CC). Decal- cified, methylene blue, and azure II stained section; original magnitude X 800. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 2.12 Light micrograph of osteoblasts. Spicule of calcified core line with os- teoblasts (Ob) and thin osteoid (arrows). Osteoprogenitor cells (Opc) are located between osteoblasts and blood vessel. Original magnification X600. 28 2.13 Diagrammatic representation of working hypothesis of bone resorption. A typical resting bone surface is lined by a thin demineralized layer (OO), a lamina lamitans (LL), and a flat bone-lining cells (BLC). . . . . . . . . . . 30 2.14 Hip joint lateral view. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 2.15 Representation of the main hip muscles. . . . . . . . . . . . . . . . . . . . . 34 2.16 Principal movements’ angles . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 2.17 Representationofanatomicalsegmentsofhumanbodyandglobalandlocal coordinate systems.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 2.18 Coordinate system for measured hip contact forces. The hip contact force vector –F and its components –F , (cid:0)F , (cid:0)F acts from the pelvis to the x y z implant head and is measured in the femur coordinate system x;y;z. . . . . 38 2.19 Joint centres, reference points and coordinate system for gait analysis. . . . 39 2.20 ContactforceFoftypicalpatientNPAduringnineactivities. Contactforce F and its components –F , (cid:0)F , (cid:0)F . F and –F are nearly identical. . . . 40 x y z z 2.21 Contact force vector F of typical patient NPA during nine activities. The z-scales go up to 300% BW. Upper diagrams: Force vector F and direction Ay of F in the frontal plane. Lower diagrams: Force vector F and direction Az of F in the transverse plane. . . . . . . . . . . . . . . . . . . . . . . . . . 40 2.22 Comparison of the predicted pattern of the hip load (solid black line) with thevariabilityofthehiploadmagnitude(greyband)measuredon4subjects through an hip prosthesis instrumented with a telemetric force sensor. . . . 42 3.1 Phantom and patient set-up during CT-scan. . . . . . . . . . . . . . . . . . 47 3.2 Pointsofmorphing:head,greatertrochanter,underbelowgreatertrochanter, and lesser trochanter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 3.3 Points of morphing:the four points on the basis of the femur: anterior, posterior, medial and lateral. . . . . . . . . . . . . . . . . . . . . . . . . . . 54 List of Figures ix 3.4 Steps of morphing algorithm: (a) original template mesh, (b) original STL mesh, (c) result from morphing the template mesh on the STL using RBF method,(d)resultsafterprojection(c)on(b),(e)resultfromtheLaplacian smoothing, (f) final result represented in a high quality mesh of the STL geometry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 3.5 Representation of the regression line between experimental and predicted strain [STM+07]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 3.6 BoneMatofthefemur,withtherangeofYoung’smodulusbetween18000MPa and the maximum value obtained from the software (19832 MPa). . . . . . 59 3.7 final structure of VME data tree (final structure of msf file).. . . . . . . . . 60 3.8 TF_IF_CH reference system. . . . . . . . . . . . . . . . . . . . . . . . . . . 61 3.9 CHA reference system; blue line is z-axis, red line is x-axis and green line is y-axis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 3.10 Disposition of the Ansys key-points on CHA reference system. . . . . . . . 62 3.11 Frontal plane and transversal plane. . . . . . . . . . . . . . . . . . . . . . . 63 3.12 Femur’s local coordinate system and keypoints. . . . . . . . . . . . . . . . . 66 3.13 Representetion of the nominal direction of load. . . . . . . . . . . . . . . . . 67 3.14 Representetion of all direction of load. . . . . . . . . . . . . . . . . . . . . . 67 4.1 Visualization of the distribution of principal deformation on the femur (an- terior view). The highest deformation is located on the top of the neck. . . 72 4.2 Visualization of the maximum strain point in posterior view of the femur. . 73 4.3 Box plot of BMD. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 4.4 Box plot of FRAX. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 4.5 Box plot of Strength. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 4.6 Box plot of WorkFlow 2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 4.7 ROC curve of BMD. AUC is of 0:73. . . . . . . . . . . . . . . . . . . . . . . 79 4.8 ROC curve of FRAX. AUC is of 0:64. . . . . . . . . . . . . . . . . . . . . . 79 4.9 ROC curve of strength. AUC is of 0:72. . . . . . . . . . . . . . . . . . . . . 80 4.10 ROC curve of WF2. AUC is of 0:75. . . . . . . . . . . . . . . . . . . . . . . 80 4.11 ROC curve of strength, WF2, N_BMD, TH_BMD and FRAX. AUC is of 0:84. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 4.12 ROC curve of strength and FRAX. AUC is of 0:74. . . . . . . . . . . . . . . 83 4.13 ROC curve of WF2 and FRAX. AUC is of 0:76.. . . . . . . . . . . . . . . . 83 x List of Figures 4.14 ROC curve of strength, WF2, and FRAX. AUC is of 0:80. . . . . . . . . . . 84 4.15 ROC curve of strength and WF2. AUC is of 0:80. . . . . . . . . . . . . . . 84 4.16 ROC curve of strength, WF2, N_BMD and TH_BMD. AUC is of 0:83. . . 85 1 Representation of a box plot. The dots indicate the outliers.. . . . . . . . . 96 2 Representation of Mann-Whitney method. In (a) the different treatment cause different effects, while in (b) they don’t produce any differences. . . . 97 3 Gaussian distribution of two population completely separated. . . . . . . . 101 4 Gaussian distribution of two population with overlapping area. . . . . . . . 101 5 Contingency table.TP represents the true positive results, TN represents the true negative results, FP represents the false positive results and FN represents the false negative results. . . . . . . . . . . . . . . . . . . . . . . 102 6 Representation of the ROC curve as a function of specificity (Sp) and sen- sitivity (Se). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 7 Representation of the AUC, the area under ROC curve. . . . . . . . . . . . 104 8 Logistic function with (cid:12) =(cid:0)1 and (cid:12) =2. . . . . . . . . . . . . . . . . . . 105 0 1

Description:
extracted the geometry of their femurs; then we meshed the specific geometry applying the morphing patient-specific mechanical model through the pre-processing phase and then to eval- uate the In the processing phase we imported the femur in ANSYS Mechanical APDL and sub- sequently:
See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.