izpis_h1_title_alt

Deformabilna poravnava slik za adaptivno radioterapijo
ID KUŠNIK, BLAŽ (Avtor), ID Špiclin, Žiga (Mentor) Več o mentorju... Povezava se odpre v novem oknu

.pdfPDF - Predstavitvena datoteka, prenos (4,72 MB)
MD5: 76FF18590848E6CB960A31A465A859E1

Izvleček
Cilj adaptivne radioterapije je manjšanje doznih odstopanj od prvotnega obsevalnega načrta, ki so običajno posledica geometrijskih negotovosti pri lokalizaciji bolnika v obsevalni napravi in sprememb anatomije bolnika. Pred začetkom procesa radioterapije bolnika najprej slikamo z uporabo računalniške tomografije (CT), magnetne resonance (MR) in/ali pozitronske emisijske tomografije (PET), nato pa na podlagi teh načrtovalnih slik ustvarimo obrise struktur in obsevalni načrt s predpisano minimalno dozo obsevanja za tarčo (tumor) in maksimalnimi dozami za kritične zdrave organe. Radioterapija poteka v več zaporednih dnevnih frakcijah, ki naj zagotovijo dostavo predpisanih doz. Tekom radioterapije pride do anatomskih sprememb, položaj bolnika pa se med frakcijami razlikuje, kar povišuje geometrijske negotovosti in je zato smiselno pred frakcijo zajeti nove slike in na podlagi teh slik posodobiti načrt obsevanja. Te slike se pogosto zajame s pomočjo računalniške tomografije s stožčastim žarkom (CBCT), ko je bolnik že v obsevalni napravi. Vendar ima CBCT nižjo kakovost v primerjavi načrtovalnimi slikami, kar otežuje postopke avtomatskega obrisovanja struktur in naknadnega prilagajanja načrta obsevanja. Ena izmed obetavnih rešitev je uporaba visoko kakovostnih načrtovalnih slik in pridruženih obrisov tako, da te slike prostorsko poravnamo na CBCT slike in nato preslikamo še obrise. Ker se strukture v telesu pogosto deformirajo in premikajo med seboj neodvisno, je za poravnavo teh slik potrebno uporabiti deformabilno poravnavo slik (DIR; ang. deformable image registration). V nalogi je podan splošen opis delovanja DIR in možnosti uporabe pri adaptivni radioterapiji. Cilj naloge je bil najti, preučiti in oceniti različne algoritme za deformabilno poravnavo slik. Za vrednotenje poravnave sta bili uporabljeni dve podatkovni bazi; prva je vsebovala šestdeset parov CT in CBCT slik radioterapije prostate, druga pa dva para slik iz javne podatkovne baze TG-132. Kvantitativno smo poravnavo vrednotili (i) s površinskim in volumskim Dice koeficientom med obrisi telesa bolnika, prostate in leve stegnenice, in (ii) z razdaljo med desetimi pari pripadajočih oslonilnih točk, ki so bile ročno označene v vsakem paru slik (TRE; ang. target registration error), (iii) s štirimi različnimi merami podobnosti, ki smo jih ovrednotili pred in po poravnavi para slik, in (iv) številom pikslov z vrednostjo determinante Jacobijeve matrike manjšo od nič (točke s fizikalno nesmiselno deformacijo). Rezultati so predstavljeni v treh delih, kjer smo v prvem podrobno vrednotili postopke poravnave pTVreg, Elastix, Plastimatch in Deeds z eksperimentalno določenimi parametri. V drugem delu smo s pomočjo zasnove poskusov z uporabo Taguchi matrik (DOE; ang. design of experiments) vpeljali zanesljiv način iskanja optimalnih parametrov ter testiranja njihovega medsebojnega vpliva in vrednotenje poravnave primerjali z eksperimentalno pridobljenimi parametri iz prvega dela. V tretjem delu pa smo s postopkom DOE poskušali najti parametre za hitrejšo izvedbo poravnave z algoritmom Deeds, ki je v prvih dveh delih kazal najboljše kakovosti poravnav, vendar na račun daljših časov izvedbe. Na koncu smo pokazali, da se je algoritem Deeds izkazal najbolje na področju točnosti in zanesljivosti poravnave, saj je dosegal najboljše rezultate pri vrednotenju z Dice koeficienti in razdaljami med oslonilnimi točkami TRE. Analize determinante Jacobijevih matrik so pokazale, da je algoritem Deeds vseboval najmanj fizikalno nesmiselnih deformacij. Na paru slik virtualnega fantoma podatkovne baze TG-132 so vsi testirani algoritmi dosegli priporočene vrednosti Dice koeficientov, na TG-132 paru CT slik prsnega koša iz podatkovne baze DIR-Lab, pa je priporočene TRE vrednosti dosegel le algoritem Deeds, in sicer s povprečno vrednostjo in standardnim odklonom 1.69 ± 2.37 mm. S pomočjo postopka DOE, pa smo brez opazne izgube na točnosti in zanesljivosti algoritem Deeds z novimi parametri pohitrili iz prvotnega povprečnega časa poravnave enega para slik 433 sekund na le 249 sekund.

Jezik:Slovenski jezik
Ključne besede:adaptivna radioterapija, deformabilna poravnava slik, Diceov koeficient, zasnova poskusov, Taguchi matrika, Deeds, Plastimatch, Elastix, pTVreg
Vrsta gradiva:Magistrsko delo/naloga
Organizacija:FE - Fakulteta za elektrotehniko
Leto izida:2021
PID:20.500.12556/RUL-126480 Povezava se odpre v novem oknu
Datum objave v RUL:23.04.2021
Število ogledov:906
Število prenosov:99
Metapodatki:XML RDF-CHPDL DC-XML DC-RDF
:
Kopiraj citat
Objavi na:Bookmark and Share

Sekundarni jezik

Jezik:Angleški jezik
Naslov:Deformable image registration for adaptive radiotherapy
Izvleček:
The aim of adaptive radiotherapy is to reduce dose deviations from the initial irradiation plan that may result from geometric uncertainties. Prior to the start of radiotherapy, the so-called planning images of the patient are first acquired using computed tomography (CT), magnetic resonance (MR) and/or positron emission tomography (PET), then segmentations of structures of interest, i.e. targets (tumors) and organs-at-risk, and treatment plans are created based on these images. Based on the plan, the radiotherapy is carried out across several daily fractions. Because geometric uncertainties are introduced during treatment in the form of anatomical changes and changes in the patient's positioning, it is necessary to acquire new images prior to each fraction and adapt the treatment plan based on these images. Pre-fraction images are often acquired using cone-beam computed tomography (CBCT) because such imaging is performed in-room, while the patient is immobilized. The low-dose CBCT imaging exhibits a lower image quality, which complicates the procedures of reliable automatic segmentation of the structures and adjustments of the treatment plan. An alternative approach is to use high-quality planning images and register them onto the CBCT images, and then map the segmentations into the CBCT image space. Because structures in the body often deform and move independently of each other, a type of non-rigid registration called deformable image registration (DIR) is required. This thesis gives a general description of the workings of DIR and lists some of its uses in adaptive radiotherapy. The aim was to find, study and quantitatively and comparatively evaluate different algorithms for deformable image registration. To this end we employ datasets of sixty pairs of CT and CBCT images of prostate cancer patients, acquired during an actual radiotherapy, and two image pairs from the public TG-132 database in order to assess the success of DIR algorithms. Quantitative evaluation was based on the following metrics: (i) surface and volumetric Dice coefficients between segmentations of the patient skin, prostate and left femur, (ii) distance between ten corresponding landmarks of each image pair using target registration errors (TRE), (iii) four similarity measures computed before and after application of DIR to each image pair, and (iv) the count of voxels with negative determinant of the Jacobian matrix (i.e. voxels with physically implausible deformation). The results are presented in three parts. In the first part we validated the pTVreg, Elastix, Plastimatch and Deeds registration algorithms with experimentally determined parameters. In the second part we introduced a robust way of finding optimal registration parameters with the design of experiments method using Taguchi matrices (DOE), and then compared the results to previous ones. In the third part, we further optimized the parameters to speed-up DIR with the Deeds algorithm, which to this stage showed the best registration results. In the end, we showed that the Deeds algorithm proved the best in terms of accuracy and robustness, as it achieved the best results during evaluations with the Dice coefficients and TRE. The analysis of the determinants of the Jacobian matrices showed that the Deeds algorithm exhibited little or no folding. On the virtual phantom image pair from the TG-132 dataset all tested algorithms achieved the recommended values of Dice coefficients, and on the TG-132 thorax CT image pair from the DIR-Lab dataset, the recommended values of TRE were only achieved by the Deeds algorithm, with an average value and standard deviation of 1.69 ± 2.37 mm. With the use of DOE methodology, we also significantly reduced the registration time of one image pair with the Deeds algorithm from 433 seconds to 249 seconds without any noticeable loss in registration accuracy and robustness.

Ključne besede:adaptive radiotherapy, deformable image registration, Dice coefficient, design of experiments, Taguchi matrix, Deeds, Plastimatch, Elastix, pTVreg

Podobna dela

Podobna dela v RUL:
Podobna dela v drugih slovenskih zbirkah:

Nazaj