Full Text Available
Note: Clicking the button above will open the full text document at the original institutional repository in a new window.
The application of a deep convolution neural network for the automated delineation of the target and organs at risk in High Dose Rate Cervical Brachytherapy
Thesis (MSc)--Stellenbosch University, 2022.
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | en_ZA |
| Published: |
Stellenbosch : Stellenbosch University
2022
|
| Subjects: | |
| Tags: |
No Tags, Be the first to tag this record!
|
| _version_ | 1867614016506953728 |
|---|---|
| access_status_str | Open Access |
| author | Duprez, Didier Raphael Roger |
| author2 | Trauernicht, Christoph Jan |
| author_browse | Duprez, Didier Raphael Roger Trauernicht, Christoph Jan |
| author_facet | Trauernicht, Christoph Jan Duprez, Didier Raphael Roger |
| author_sort | Duprez, Didier Raphael Roger |
| collection | Thesis |
| dc_rights_str_mv | Stellenbosch University |
| description | Thesis (MSc)--Stellenbosch University, 2022. |
| format | Thesis |
| id | oai:scholar.sun.ac.za:10019.1/125999 |
| institution | Stellenbosch University (South Africa) |
| language | en_ZA |
| last_indexed | 2026-06-10T12:45:20.375Z |
| license_str | Other — see source repository |
| provenance_str_mv | Harvested via OAI-PMH from SUNScholar — Stellenbosch University Repository |
| publishDate | 2022 |
| publishDateRange | 2022 |
| publishDateSort | 2022 |
| publisher | Stellenbosch : Stellenbosch University |
| publisherStr | Stellenbosch : Stellenbosch University |
| record_format | dspace |
| source_str | SUNScholar — Stellenbosch University Repository |
| spelling | oai:scholar.sun.ac.za:10019.1/125999 The application of a deep convolution neural network for the automated delineation of the target and organs at risk in High Dose Rate Cervical Brachytherapy Duprez, Didier Raphael Roger Trauernicht, Christoph Jan Stellenbosch University. Faculty of Medicine and Health Sciences. Dept. of Medical Imaging and Clinical Oncology. Medical Physics. Cervix uteri -- Cancer -- Radiotherapy Cervix uteri -- Cancer -- Treatment Radioisotope brachytherapy UCTD Thesis (MSc)--Stellenbosch University, 2022. ENGLISH SUMMARY: Low/middle income countries suffer from large deficits in experienced Radiation Oncologists, Medical Physicists and Radiation Therapists. Due to these deficits, the bottlenecks experienced in the High-dose rate (HDR) cervical brachytherapy treatment planning workflow are amplified. Image-guided HDR cervical brachytherapy is a complex, labour intensive, manual, time-consuming and expertise driven process. Automation in radiotherapy treatment planning, especially in brachytherapy, has the potential to substantially reduce the overall planning time however most of these algorithms require high level of expertise to develop. The aim of this study is to implement the out of the box self-configuring deep neural network package, known as No New U-Net (nnU-Net), for the task of automatically delineating the organs at risk (OARs) and high-risk clinical target volume (HR CTV) for HDR cervical brachytherapy. The computed tomography (CT) scans of 100 previously treated patients were used to train and test three different nnU-Net configurations (2D, 3DFR and 3DCasc). The performance of the models was evaluated by calculating the Sørensen-Dice similarity coefficient, Hausdorff distance (HD), 95th percentile Hausdorff distance, mean surface distance (MSD) and precision score for 20 test patients. The dosimetric accuracy between the manual and predicted contours was assessed by looking at the various dose volume histogram (DVH) parameters and volume differences. Three different radiation oncologists (ROs) scored the predicted bladder, rectum and HR CTV contours generated by the best performing model. The manual contouring, prediction and editing times were recorded. The mean DSC, HD, HD95, MSD and precision scores for our best performing model (3DFR) were 0.92/7.5 mm/3.0 mm/ 0.8 mm/0.91 for the bladder, 0.84/13.8 mm/5.2 mm/1.4 mm/0.84 for the rectum and 0.81/8.5 mm/6.0 mm/2.2 mm/0.80 for the HR CTV. Mean dose differences (D2cc/90%) and volume differences were 0.08 Gy/1.3 cm3 for the bladder, 0.02 Gy/0.7 cm3 for the rectum and 0.33 Gy/1.5 cm3 for the HR CTV. On average, 65 % of the generated contours were clinically acceptable, 33 % requiring minor edits, 2 % required major edits and no contours were rejected. Average manual contouring time was 14.0 minutes, while the average prediction and editing times were 1.6 and 2.1 minutes respectively. Our best performing model (3DFR) provided fast accurate auto generated OARs and HR CTV contours with a large clinical acceptance rate. Future work should focus on including larger datasets to eliminate inconsistencies, as well as focus on automating the generation of treatment plans. AFRIKAANSE OPSOMMING: Lae/middel-inkomste lande ly aan groot tekorte in ervare stralingsonkoloe, mediese fisici en radioterapeute. Hierdie tekorte lei tot verhoogde werksladings in die beplanning van hoe-dosis-tempo (HDT) servikale bragiterapie behandelings. Beeldgeleide HDT servikale bragiterapie is 'n komplekse-, arbeidsintensiewe-, tydrowende- en kundigheidgedrewe proses. Outomatisering in radioterapie-behandelingsbeplanning, veral in bragiterapie, het die potensiaal om die algehele beplanningstyd aansienlik te verkort, maar die meeste van hierdie algoritmes vereis gevorderde kundigheid om te ontwikkel. Die doel van hierdie studie is die implimentering van die self-konfigurerende diep neurale netwerk pakket, No New U-Net (nnU-Net), om die krities belangrike organe (OARs) en hoe-risiko kliniese teiken-volumes (HR-CTV) outomaties vir HDT servikale bragiterapie te delinieer. Drie verskillende nnU-Net-konfigurasies (2D, 3DFR en 3DCasc) is retrospektief op die rekenaartomografiese (CT) beelde van 100 behandelde pasiente opgelei en getoets. Die doeltreffendheid van die modelle is evalueer deur die Sørensen-dobbelsteen-ooreenkomskoeffisient (DSC), Hausdorff-afstand (HD), 95ste persentiel Hausdorff-afstand (HD95), gemiddelde oppervlakafstand (MSD) en presisietelling vir 20 toets-pasiente te bereken. Die verskillende dosis-volume-histogram (DVH) parameters en volume-verskille is gebruik om die dosimetriese akkuraatheid tussen die hand-gedelinieerde en voorspelde kontoere te beoordeel. Die beste presterende model se voorspelde blaas-, rektum- en HR-CTV-kontoere is daarna deur drie verskillende stralingsonkoloe beoordeel. Die totale tydsduur van hand-gedelinieerde kontoere, asook die voorspelde- en gewysigde kontoere is aangeteken. Die gemiddelde DSC, HD, HD95, MSD en presisietellings vir die beste model (3DFR) is 0.92/7.5 mm/3.0 mm/0.8 mm/0.91 vir die blaas, 0.84/13.8 mm/5.2 mm/1.4 mm/0.84 vir die rektum en 0.81/8.5 mm/6.0 mm/2.2 mm/0.80 vir die HR-CTV. Die gemiddelde dosis- (D2cc/90%) en volume-verskille is 0.08 Gy/1.3 cm3 vir die blaas, 0.02 Gy/0.7 cm3 vir die rektum en 0.33 Gy/1.5 cm3 vir die HR-CTV. Oor die algemeen is 65% van die genereerde kontoere klinies aanvaarbaar, terwyl slegs 33% klein veranderinge benodig het. Verder, het slegs 2% hiervan groot veranderinge benodig, terwyl geen kontoere afgekeur is nie. Die gemiddelde hand-gedelinieerde kontoere het 14.0 minute geneem, terwyl die gemiddelde voorspelde kontoere 1.6 minute geneem het. Dit het slegs 2.1 minute geneem om die voorspelde kontoere te wysig totdat dit klinies aanvaarbaar is. Die beste model (3DFR) het vinnige, akkurate outomaties-gegenereerde kontoere met 'n groot kliniese aanvaardingskoers verskaf. Groter datastelle sal in die toekoms ingesluit word om teenstrydighede uit te skakel en outomatiese behandelingsbeplanning sal ook deel vorm van toekomstige navorsing. Masters 2022-11-16T08:27:59Z 2023-01-16T12:44:59Z 2022-11-16T08:27:59Z 2023-01-16T12:44:59Z 2022-12 Thesis http://hdl.handle.net/10019.1/125999 en_ZA Stellenbosch University xi, 55 pages : illustrations, includes annexures application/pdf Stellenbosch : Stellenbosch University |
| spellingShingle | Cervix uteri -- Cancer -- Radiotherapy Cervix uteri -- Cancer -- Treatment Radioisotope brachytherapy UCTD Duprez, Didier Raphael Roger The application of a deep convolution neural network for the automated delineation of the target and organs at risk in High Dose Rate Cervical Brachytherapy |
| title | The application of a deep convolution neural network for the automated delineation of the target and organs at risk in High Dose Rate Cervical Brachytherapy |
| title_full | The application of a deep convolution neural network for the automated delineation of the target and organs at risk in High Dose Rate Cervical Brachytherapy |
| title_fullStr | The application of a deep convolution neural network for the automated delineation of the target and organs at risk in High Dose Rate Cervical Brachytherapy |
| title_full_unstemmed | The application of a deep convolution neural network for the automated delineation of the target and organs at risk in High Dose Rate Cervical Brachytherapy |
| title_short | The application of a deep convolution neural network for the automated delineation of the target and organs at risk in High Dose Rate Cervical Brachytherapy |
| title_sort | application of a deep convolution neural network for the automated delineation of the target and organs at risk in high dose rate cervical brachytherapy |
| topic | Cervix uteri -- Cancer -- Radiotherapy Cervix uteri -- Cancer -- Treatment Radioisotope brachytherapy UCTD |
| url | http://hdl.handle.net/10019.1/125999 |
| work_keys_str_mv | AT duprezdidierraphaelroger theapplicationofadeepconvolutionneuralnetworkfortheautomateddelineationofthetargetandorgansatriskinhighdoseratecervicalbrachytherapy AT duprezdidierraphaelroger applicationofadeepconvolutionneuralnetworkfortheautomateddelineationofthetargetandorgansatriskinhighdoseratecervicalbrachytherapy |