University of Twente Student Theses
Improving focus on uterine fibroids : visualizing the treatment effect after ultrasound ablation by using diffusion weighted imaging.
Slotman, D.J. (2020) Improving focus on uterine fibroids : visualizing the treatment effect after ultrasound ablation by using diffusion weighted imaging.
|
PDF
16MB |
| Abstract: | Uterine fibroids (UFs) are common benign neoplasms of the uterus, entailing a high socioeconomic burden. MR-HIFU is a recent alternative treatment option for symptomatic UFs and has an appealing non-invasive character. Shorter procedure times and improved treatment outcomes could increase the availability of MR-HIFU for patients suffering from UFs. As for now, no established method is available to assess the treatment progress intraprocedurally, leading occasionally to unneeded prolongations of ablations and limited treatment outcomes. Currently, the treatment effect is examined postprocedurally by acquiring a contrast-enhanced (CE)-T1w scan, visualizing the non-perfused volume (NPV). A major drawback of a contrast agent is the inability to use it between sonications, due to risk of gadolinium (Gd) entrapment and distortion of the thermometry. However, there has been some suggestions about applying diffusion weighted imaging (DWI) to visualize tissue perfusion between MR-HIFU sonications. Until now, previous work about DWI has only been analyzed with a mono-exponential model in the context of Gd-free NPV visualization. This thesis has examined the way in which intravoxel incoherent motion analysis (IVIM) and deep learing (DL) can aid in predicting perfusion based on DWI data after MR-HIFU treatment. A dataset of 56 patients with a paired DWI and CE-T1w scan was used to conduct this analysis. With the first strategy, a bi-exponential IVIM model was applied. Quantitative analysis disclosed a significant difference between parameters within the perfused and non-perfused volumes, implicating contrast differences between the ablated and viable tissue. With the second approach, a DL-based method was used to create synthetic CE-T1w scans with a conditional generative adverserial network. With quantitative and qualitative evaluation it was established that the DL-method could produce synthetic CE-T1w scans that can be used for adequate treatment assessment. These findings show that these methods for DWI-data analysis are feasible for MR-HIFU treatment evaluation of UF. Further work should focus on prospective assessment of these methods and their clinical value. The findings from this work are a next step towards more efficient healthcare by improving the efficiency of MR-HIFU treatments of UFs. |
| Item Type: | Essay (Master) |
| Faculty: | TNW: Science and Technology |
| Subject: | 33 physics, 42 biology, 44 medicine, 54 computer science |
| Programme: | Technical Medicine MSc (60033) |
| Link to this item: | https://purl.utwente.nl/essays/83635 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page