Thesis (MScEng (Industrial Engineering))--University of Stellenbosch, 2010. / ENGLISH ABSTRACT: Current technology enables researchers to identify a broad spectrum of opportunities in the biomedical industry to develop new and innovative products. Imaging technology, such as Computerised Tomography (CT) scanners or Magnetic Resonance Imaging (MRI) scanners, allow doctors to visualise a patient’s internal organs and bone structure in high quality three-dimensional images. Rapid Prototyping Technology (RPT) can already produce high quality complex parts, such as concept parts for the automobile industry and medical models for preoperative planning. These parts are divided into thin layers and manufactured layer by layer. At the same time the layers are joined together to produce the desired part. Generic artificial intervertebral disc implants already exist. However, these discs are only available in standard geometrical dimensions. The possibility of using imaging technology and RPT to design and manufacture a customized, patient specific implant will be investigated. A simple design (ball and socket) is used to illustrate the design process of a customized disc. It should be noted that this project does not attempt to design a new artificial intervertebral disc implant, but rather describes the design process. The research question is: How accurate can the customised disc implant’s inverse geometry represent the geometry of the vertebrae’s endplates? A preliminary research was done and the results were used to calculate the sample size for this study. A cadaver, provided by Stellenbosch University’s Faculty of Health Sciences’ Anatomy and Histology Department, was CT scanned. The L4- and L5-vertebrae were dissected, cleaned and measured using a photogrammetry measuring machine. Meanwhile, the data gathered from the CT scan is used to design the customised disc implant. The disc is manufactured from Ti6Al4V using a RPT technique called Direct Metal Laser Sintering. After the part is manufactured it is also measured using a photogrammetry measuring machine. The photogrammetry data from the vertebrae and the manufactured customised disc implant are compared, analysed and a hypothesis is formed. It can now be determined, with a certain degree of confidence, how accurate the customised disc implant’s geometry can represent the geometry of the vertebrae’s endplates. The design of a customised disc implant demands many work hours from a qualified engineer or designer, which in turn increases the production costs. This study describes a user-friendly program which will semi-automate the design process. Only limited input from the physician will be required. This program will decrease design time, which will have a direct effect on the production costs. The manufacturing costs are investigated as well. The results from this study indicates that it is possible to design a customized prosthetic, with the help of a custom disc generator, within 27 minutes. The customized disc can then be manufactured with an accuracy of 0.37 mm using rapid prototyping. / AFRIKAANSE OPSOMMING: Huidige tegnologie maak dit vir navorsers moontlik om ʼn breë spektrum geleenthede in die biomediese bedryf te identifiseer en nuwe produkte te ontwikkel. ʼn Pasiënt kan met ʼn Gerekenariseerde Tomografie (GT) -flikkergram of ʼn Magnetiese Resonansiebeelding (MRB) - masjien geskandeer word om sodoende ʼn drie-dimensionele beeld van sy of haar interne organe en beenstrukture te verkry. Deur gebruik te maak van snelle prototiperingstegnologie (SPT) kan daar alreeds enige komplekse geometriese vorm vervaardig word. Hierdie tegnologie word ingespan om parte, ontwerp met die hulp van RGO (Rekenaargesteunde Ontwerp), te vervaardig. Die spesifieke part word in dun lae opgedeel en daarna laag vir laag vervaardig. Terselfdertyd word die lae aan mekaar geheg, totdat die gewenste vorm gegenereer is. Die moontlikheid om ʼn GT-flikkergram én SPT te gebruik, met die doel om ʼn pasklaar-vervaardigde, persoon-spesifieke implantaat te ontwerp en te vervaardig, word in hierdie projek ondersoek. Daar bestaan alreeds generiese kunsmatige intervertebrale skyf-implantate (KISI’s). Hierdie skywe word egter beperk deurdat dit slegs in standaard geometriese dimensies vervaardig word. Met dié projek word die moontlikheid van ʼn pasklaar-vervaardigde intervertebrale skyf-implantate (PVKISI) vir ʼn bepaalde pasiënt, ondersoek. ʼn Eenvoudige meganiese ontwerp (bal-en-pootjie) word gebruik om die ontwerpproses van ʼn pasklaar-skyf in hierdie projek te beskryf. Let daarop dat die projek nie poog om ʼn nuwe kunsmatige intervertebrale skyf te ontwerp nie, maar om die ontwerpproses te beskryf. Die vraag wat ondersoek word, is: Hoe akkuraat kan ʼn PVKISI die inverse geometrie van die pasiënt se intervertebrale kontakoppervlaktes voorstel? ʼn Voorafgaande ondersoek is gedoen en die resultate is gebruik om die steekproef-grootte vir hierdie studie te bepaal. ʼn Kadawer, voorsien deur die Universiteit van Stellenbosch se Fakulteit Gesondheidwetenskappe se Departement Anatomie en Histologie, is met ʼn GT-flikkergram geskandeer. Die L4- en L5-werwels is gedissekteer, skoon gemaak en met ʼn fotogrammetriemeetmasjien gemeet. Intussen is die data, verkry van die GT-flikkergram, gebruik om die PVKISI te ontwerp. Die PVKISI is van Ti6Al4V vervaardig deur Direkte Metaal Laser-Sintering (DMLS). Die part is ook met ʼn fotogrammetrie-meetmasjien gemeet. Die fotogrammetrie-data van die werwels en die PVKISI is vergelyk, geanaliseer en ʼn hipotese is daar gestel. Daar kan dus met statistiese sekerheid bepaal word hoe akkuraat die PVKISI die inverse geometrie van die intervertebrale kontakoppervlaktes kan voorstel. Die ontwerp van ʼn PVKISI vereis baie werksure van ʼn gekwalifiseerde ingenieur of ontwerper, wat veroorsaak dat die vervaardigingskoste van so ʼn implantaat kan verhoog. In dié projek word ʼn gebruikersvriendelike koppelprogram beskryf wat die ontwerpproses semi-outomatiseer. Daar sal slegs beperkte bystand van die betrokke medici vereis word. Dié koppelprogram behoort heelwat te bespaar aan die hoeveelheid werksure bestee aan die ontwerp van die PVKISI, wat direk die koste van ʼn implantaat sal verlaag. Die kostes vir die vervaardiging van die PVKISI met DMLS is ook ondersoek om te bepaal hoe kostes bespaar kan word. Daar is getoon dat ʼn pasklaar-prostese se kontakoppervlaktes met ʼn akkuraatheid van 0.37 mm, deur snelle prototipering, vervaardig kan word. Deur van die koppelprogram, wat in die studie beskryf word, gebruik te maak, sal dit moontlik wees om ʼn pasklaar-protese binne 27 minute te ontwerp.

L’ostéoporose est définie par une diminution de la masse osseuse et une altération de la structure osseuse. L’évaluation de l’architecture de l’os trabéculaire vertébral est un des enjeux majeurs de recherche actuels visant à la caractérisation des modifications du tissu osseux liées au vieillissement. Par ailleurs, le traitement de la fracture vertébrale ostéoporotique par vertébroplastie a démontré son efficacité. Le ciment actuel utilisé pour la vertébroplastie est le polyméthylméthacrylate. Ce ciment est efficace pour stabiliser la vertèbre et diminuer la douleur mais il existe une toxicité pour les tissus mous avoisinant, notamment la moelle en cas de fuite, et des inquiétudes quant au risque de fracture des vertèbres adjacentes. De nouveaux ciments biointégrables du type ciments phosphocalciques supplémentés en bisphosphonates pourraient avoir leur place à l’avenir dans le traitement des vertèbres.Notre travail de thèse s’est inscrit dans un projet pluridisciplinaire de recherche sur les altérations de la structure de l’os trabéculaire vertébral, reposant sur une analyse en imagerie médicale et une étude biomécanique, et sur le développement d’un nouveau ciment phosphocalcique supplémenté en bisphosphonates.L’ensemble de nos résultats pourra nous permettre :-D’améliorer la prédiction du risque de fracture vertébrale ostéoporotique-D’optimiser les nouveaux ciments biointégrables / The current definition of osteoporosis is a low bone mass associated with a microarchitecture deterioration. Vertebral trabecular bone microarchitecture assessment is an important research topic aiming at the characterization of the modifications of the bone tissue bound to ageing. On the other hand, the vertebral fracture treatment by vertebroplasty is known to be effective. Currently the cement used in vertebroplasty is the polymethylmetacrylate (PMMA). This cement stabilizes the vertebral fracture and decreases the pain, but there are a toxicity for nearby soft tissues, in particular spinal marrow in case of leakage, and concerns as for the risk of fracture of the neighboring vertebrae. In the future, new biointegrable calcium phosphate cement supplemented in bisphosphonate could be used in the vertrebrae treatment.Our work was based on a multidisciplinary project which aimed firstly at evaluating trabecular vertebral bone microarchitecture using medical imaging and biomechanical testing in this crucial anatomical region, secondly at developing a new calcium phosphate cement supplemented in bisphosphonate. Our results may lead to:- improve the fracture risk prediction in osteoporotic vertebra- improve the new biointegrable cements

Les effets des agents anti-ostéoporotiques sur le tissu osseux sont évalués au niveau de la crête iliaque (CI) mais les réponses aux traitements peuvent varier selon le site osseux. Le but de cette étude était de comprarer les effets de l’acide zolérodronique (ZOL) et du tériparatide (TPTD) au niveau de la crête iliaque et de la vertèbre lombaire L1 (VL1) chez la brebis âgée. Le ZOL a induit une forte diminution du remodelage osseux et une augmentation des microendommagements au niveau des deux sites et une modification des crosslinks du collagène surtout au niveau de l’os cortical de la CI. Trois mois de TPTD ont augmenté le remodelage osseux uniquement au niveau de la VL1. En conclusion, les délais et les amplitudes de réponses au ZOL ou au TPTD diffèrent entre la CI et la VL1 chez la brebis. Ces résultats montrent l’importance de prendre en compte le site osseux pour évaluer les effets des agents anti-ostéoporotiques / The effects of anti-osteoporotic agents on bone tissue are evaluated on iliac crest (IC) but the answers to treatments may vary according to the skeletal site. The purpose of this study was to compare the effect of zoledronic acid (ZOL) and teriparatide (TPTD) on IC and lumbar vertebrae (LV1) in ewes. ZOL has induced a high decrease of bone remodeling, an increase in microdamages in both sites and a modification of collagen crosslinks mainly in cortical bone of IC. Three months of TPTD has increased the bone remodeling only in LV1. In conclusion, the delays and the magnitudes of responses to ZOL or to TPTD differ between IC and LV1 in ewes. These results show that the distinction of bone sites to study the early effects of antiosteoporotic therapies appears meaningful

This thesis deals with stress strain analysis of an aortic abdominal aneurysm (AAA) and the influence of its contact with the spine on the extreme wall stress. The influence was tested on the idealized geometry, as well as on ten patient specific geometries obtained from computer tomography (CT-A) scans. Hyperelastic constitutive models were used for the AAA wall and intraluminal thrombus (ILT) tissue description. The prestress algorithm was used for reconstruction of the unloaded geometry to get more trustworthy results against the geometry from CT which was obtained under the blood pressure. Statistical analysis was used for the results evaluation. The maximal increase of peak wall stress was as high as 81 %.

This master's thesis focuses on detection of pathological vertebrae in spinal CT utilized by machine learning. Theoretical part describes anatomy of the spine and occurrence of pathologies in CT image data, contains an overview of existing methods intended for automated detection of pathological vertebrae. Practical part devotes to design a computer aided detection systems to identify pathological vertebrae and to classify a type of pathology. Designed classification system is based on using neural network, which performs classification step and on principal component analysis (PCA), which is used to reducing the original number of observation features. For completing this task were used real data. Conclusion contains evaluation of obtained results.

The aim of this work was the development of algorithms for detection segmentation and classification of difficult to define bone metastatic cancerous lesions from spinal CT image data. For this purpose, the patient database was created and annotated by medical experts. Successively, three methods were proposed and developed; the first of them is based on the reworking and combination of methods developed during the preceding project phase, the second method is a fast variant based on the fuzzy k-means cluster analysis, the third method uses modern machine learning algorithms, specifically deep learning of convolutional neural networks. Further, an approach that elaborates the results by a subsequent random forest based meta-analysis of detected lesion candidates was proposed. The achieved results were objectively evaluated and compared with results achieved by algorithms published by other authors. The evaluation was done by two objective methodologies, technical voxel-based and clinical object-based ones. The achieved results were subsequently evaluated and discussed.