Applied AI/ML for automatic customisation of medical implants
Saved in:
| Title: | Applied AI/ML for automatic customisation of medical implants |
|---|---|
| Authors: | Burge, Thomas Alexander |
| Committee Members: | Myant, Connor; Jeffers, Jonathan |
| Summary: | Most knee replacement surgeries are performed using 'off-the-shelf' implants, supplied with a set number of standardised sizes. X-rays are taken during pre-operative assessment and used by clinicians to estimate the best options for patients. Manual templating and implant size selection have, however, been shown to be inaccurate, and frequently the generically shaped products do not adequately fit patients' unique anatomies. Furthermore, off-the-shelf implants are typically made from solid metal and do not exhibit mechanical properties like the native bone. Consequently, the combination of these factors often leads to poor outcomes for patients. Various solutions have been outlined in the literature for customising the size, shape, and stiffness of implants for the specific needs of individuals. Such designs can be fabricated via additive manufacturing which enables bespoke and intricate geometries to be produced in biocompatible materials. Despite this, all customisation solutions identified required some level of manual input to segment image files, identify anatomical features, and/or drive design software. These tasks are time consuming, expensive, and require trained resource. Almost all currently available solutions also require CT imaging, which adds further expense, incurs high levels of potentially harmful radiation, and is not as commonly accessible as X-ray imaging. This thesis explores how various levels of knee replacement customisation can be completed automatically by applying artificial intelligence, machine learning and statistical methods. The principal output is a software application, believed to be the first true 'mass-customisation' solution. The software is compatible with both 2D X-ray and 3D CT data and enables fully automatic and accurate implant size prediction, shape customisation and stiffness matching. It is therefore seen to address the key limitations associated with current implant customisation solutions and will hopefully enable the benefits of customisation to be more widely accessible. |
| URL: | https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.889986 |
| Database: | OpenDissertations |
| FullText | Text: Availability: 0 |
|---|---|
| Header | DbId: ddu DbLabel: OpenDissertations An: ddu.oai.ethos.bl.uk.889986 AccessLevel: 6 PubType: Dissertation/ Thesis PubTypeId: dissertation PreciseRelevancyScore: 0 |
| IllustrationInfo | |
| Items | – Name: Title Label: Title Group: Ti Data: Applied AI/ML for automatic customisation of medical implants – Name: Author Label: Authors Group: Au Data: <searchLink fieldCode="AR" term="%22Burge%2C+Thomas+Alexander%22">Burge, Thomas Alexander</searchLink> – Name: Author Label: Committee Members Group: Au Data: <searchLink fieldCode="CO" term="%22Myant%2C+Connor%22">Myant, Connor</searchLink>; <searchLink fieldCode="CO" term="%22Jeffers%2C+Jonathan%22">Jeffers, Jonathan</searchLink> – Name: Abstract Label: Summary Group: Ab Data: Most knee replacement surgeries are performed using 'off-the-shelf' implants, supplied with a set number of standardised sizes. X-rays are taken during pre-operative assessment and used by clinicians to estimate the best options for patients. Manual templating and implant size selection have, however, been shown to be inaccurate, and frequently the generically shaped products do not adequately fit patients' unique anatomies. Furthermore, off-the-shelf implants are typically made from solid metal and do not exhibit mechanical properties like the native bone. Consequently, the combination of these factors often leads to poor outcomes for patients. Various solutions have been outlined in the literature for customising the size, shape, and stiffness of implants for the specific needs of individuals. Such designs can be fabricated via additive manufacturing which enables bespoke and intricate geometries to be produced in biocompatible materials. Despite this, all customisation solutions identified required some level of manual input to segment image files, identify anatomical features, and/or drive design software. These tasks are time consuming, expensive, and require trained resource. Almost all currently available solutions also require CT imaging, which adds further expense, incurs high levels of potentially harmful radiation, and is not as commonly accessible as X-ray imaging. This thesis explores how various levels of knee replacement customisation can be completed automatically by applying artificial intelligence, machine learning and statistical methods. The principal output is a software application, believed to be the first true 'mass-customisation' solution. The software is compatible with both 2D X-ray and 3D CT data and enables fully automatic and accurate implant size prediction, shape customisation and stiffness matching. It is therefore seen to address the key limitations associated with current implant customisation solutions and will hopefully enable the benefits of customisation to be more widely accessible. – Name: URL Label: URL Group: URL Data: <link linkTarget="URL" linkTerm="https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.889986" linkWindow="_blank">https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.889986</link> |
| PLink | https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=ddu&AN=ddu.oai.ethos.bl.uk.889986 |
| RecordInfo | BibRecord: BibEntity: Languages: – Code: eng Text: English Titles: – TitleFull: Applied AI/ML for automatic customisation of medical implants Type: main BibRelationships: HasContributorRelationships: – PersonEntity: Name: NameFull: Burge, Thomas Alexander IsPartOfRelationships: – BibEntity: Dates: – D: 01 M: 01 Type: published Y: 2023 |
| ResultId | 1 |
