| LDR | | 00000nam u2200205 4500 |
| 001 | | 000000435032 |
| 005 | | 20200227113735 |
| 008 | | 200131s2018 ||||||||||||||||| ||eng d |
| 020 | |
▼a 9781687934840 |
| 035 | |
▼a (MiAaPQ)AAI27536296 |
| 035 | |
▼a (MiAaPQ)umichrackham002173 |
| 040 | |
▼a MiAaPQ
▼c MiAaPQ
▼d 247004 |
| 082 | 0 |
▼a 574 |
| 100 | 1 |
▼a Bray, Mathieu. |
| 245 | 10 |
▼a Advances in Methods, Algorithms and Software for Optimization and Simulation of Kidney Paired Donation Programs. |
| 260 | |
▼a [S.l.]:
▼b University of Michigan.,
▼c 2018. |
| 260 | 1 |
▼a Ann Arbor:
▼b ProQuest Dissertations & Theses,
▼c 2018. |
| 300 | |
▼a 105 p. |
| 500 | |
▼a Source: Dissertations Abstracts International, Volume: 81-05, Section: B. |
| 500 | |
▼a Advisor: Song, Peter Xuekun. |
| 502 | 1 |
▼a Thesis (Ph.D.)--University of Michigan, 2018. |
| 506 | |
▼a This item must not be sold to any third party vendors. |
| 506 | |
▼a This item must not be added to any third party search indexes. |
| 520 | |
▼a Kidney Paired Donation (KPD) is a public health initiative wherein kidney transplant candidates with willing but incompatible donors are pooled together with the goal of finding transplant opportunities through the exchange of donors. Transplants are completed either via exchange cycles among incompatible donor-candidate pairs, or by transplant chains initiated by altruistic donors not associated with any particular candidate. Selection of cycles and chains in a KPD program has been modeled as a constrained optimization problem over a directed network, where each vertex represents either a donor-candidate pair or an altruistic donor, and each edge represents a potential transplant between a donor and candidate, weighted by utility (Roth et al. 2007). The goal is to select the disjoint set of cycles and chains that maximize the total utility among the selected transplants. Our aim in this dissertation is to develop and expand existing methods to optimize KPD programs, to generalize concepts and address current realities in KPD management, and to validate these methods through simulation.In the first project, we consider the issues of uncertainty in transplant viability after selection and recourse to fallback options in cases of non-viability. We extend the standard KPD model to select more general subgraphs of the KPD network, where fallback options consisting of successful sub-cycles and sub-chains can be realized. Methods for determining the appropriate expected utility under uncertainty to assign to such subgraphs are established, through exact calculation as well as by estimation using Monte Carlo sampling. Simulations of KPD programs are performed, demonstrating a substantial advantage in selecting such subgraphs, in terms of realized utility, compared to the standard KPD model.In the second project, we generalize the previous formulation to account for candidates joining a KPD program with more than one incompatible donor, introducing additional potential transplant opportunities and fallback options in cases of non-viability. Such KPD models have been sparsely considered in the literature. In this setting, edge properties depend on the specific donor involved, and there exists the possibility of more than one directed edge between two vertices, with implications on the expected utility calculation. We also present a state-transition model for donor and candidate availability within KPD. Through simulation, we demonstrate the benefits of our generalized formulation, not only for individual candidates but for the KPD program as a whole. In the third project, we investigate temporal aspects of KPD programs, with candidates and donors joining and departing over time, along with matches being confirmed or rejected. We consider modeling the KPD as a 3-dimensional tensor, and decompose the tensor to uncover latent factors within the KPD that allow pairs to be clustered based on their propensity to match with other pairs. This cluster assignment can be used to inform allocation systems, allowing for greater balance between utility and equity. We briefly overview alternative techniques to model the dynamics of KPD programs. We introduce our software application, which renders an interactive virtual KPD network and incorporates methods developed in this dissertation for research and clinical use. Finally, we outline current realities in KPD management that have yet to be thoroughly explored in the literature as directions for future research. |
| 590 | |
▼a School code: 0127. |
| 650 | 4 |
▼a Public health. |
| 650 | 4 |
▼a Biostatistics. |
| 690 | |
▼a 0308 |
| 690 | |
▼a 0573 |
| 710 | 20 |
▼a University of Michigan.
▼b Biostatistics. |
| 773 | 0 |
▼t Dissertations Abstracts International
▼g 81-05B. |
| 773 | |
▼t Dissertation Abstract International |
| 790 | |
▼a 0127 |
| 791 | |
▼a Ph.D. |
| 792 | |
▼a 2018 |
| 793 | |
▼a English |
| 856 | 40 |
▼u http://www.riss.kr/pdu/ddodLink.do?id=T15494246
▼n KERIS
▼z 이 자료의 원문은 한국교육학술정보원에서 제공합니다. |
| 980 | |
▼a 202002
▼f 2020 |
| 990 | |
▼a ***1008102 |
| 991 | |
▼a E-BOOK |