Image quality in Cherenkov positron emission tomography

Razdevšek, Gašper (author), Dolenec, Rok (mentor)

Positron emission tomography (PET) is an important medical imaging modality, and although the technology is well established, there are still opportunities as well as a demand for better PET systems. The detection of annihilation photons in PET is based on scintillation light detection, but an interesting alternative is detection based on Cherenkov photons. Dense Cherenkov radiators provide an opportunity for high gamma detection efficiency - due to their high stopping power and photofraction - and excellent coincidence time resolution (CTR). However, because only a few tens of Cherenkov photons follow a gamma interaction in the radiator, the detection efficiency and the energy resolution of a pure Cherenkov detector are an issue. This work explores the performance of PET scanners based on Cherenkov detectors through Monte Carlo simulations and aims to determine whether such scanners are clinically feasible. First, single PbF$_2$ crystal based detectors with different surface treatments and photo-detectors covering one or multiple crystal faces were studied. Then, the potential performance of a full-size Cherenkov PET scanner was investigated using the NEMA NU 2-2018 standard and compared with a reference scanner - Siemens Biograph Vision PET scanner. The simulations were performed on a super-computing network using GATE software, and CASToR software was used for (TOF-OSEM) image reconstruction. Cherenkov scanner with single-sided readout performed similarly, while multi-sided readout detector designs performed better than the reference scanner, thanks to their improved coincidence detection efficiency and CTR. This work demonstrates that even though pure Cherenkov scanners have basically no energy resolution, the scatter fraction of around 50\% is not prohibitively large, and images comparable to the state-of-the-art clinical PET scanner can be achieved. Cherenkov detectors are expected to perform even better in low-scatter environments - brain, breast, or preclinical imaging studies - and their potential for low cost could make them very interesting for total-body scanners.

• Language: English
• Keywords: TOF PET, Cherenkov radiation, multi-sided crystal readout, PbF$_2$, Geant4/GATE Monte Carlo simulation, CASToR, NEMA NU 2-2018
• Work type: Doctoral dissertation
• Typology: 2.08 - Doctoral Dissertation
• Organization: FMF - Faculty of Mathematics and Physics
• Year: 2023
• PID: 20.500.12556/RUL-145613
• COBISS.SI-ID: 150407171
• Publication date in RUL: 26.04.2023

Secondary language

• Language: Slovenian
• Title: Kakovost slike pri Čerenkovi pozitronski emisijski tomografiji

Abstract

Pozitronska emisijska tomografija (PET) je pomemben način medicinskega slikanja. Čeprav je metoda široko uveljavljena, še vedno obstajajo priložnosti za njeno izboljšavo, prav tako pa obstaja potreba po boljših sistemih PET. Zaznavanje anihilacijskih fotonov v PET temelji na detekciji scintilacijske svetlobe, zanimiva alternativa pa je detekcija na podlagi fotonov Čerenkova. Čerenkovi sevalci z veliko gostoto ponujajo priložnost za visoko učinkovitost detekcije žarkov gama - zaradi visokega atenuacijskega koeficienta in visokega deleža fotoefekta - in odlično časovno ločljivost koincidenc (CTR). Pri interakciji žarka gama v sevalcu nastane le nekaj deset fotonov Čerenkova, kar vpliva na učinkovitost detekcije in energijsko ločljivost ter predstavljata potencialni problem te metode. To delo raziskuje delovanje skenerjev PET na osnovi Čerenkovih detektorjev z uporabo simulacij z metodo Monte Carlo in se osredotoča na raziskavo, ali so taki skenerji klinično sprejemljivi oz. uporabni. Najprej so bile raziskane različne konfiguracije detektorjev na osnovi svinčevega fluorida (PbF$_2$) z različnimi površinami kristalna in fotodetektorji, ki prekrivajo eno ali več ploskev kristala. Nato je bila preučena potencialna zmogljivost Čerenkovega PET skenerja z uporabo NEMA NU 2-2018 standarda. Zmogljivost je bila primerjana tudi z referenčnim skenerjem - Siemens Biograph Vision PET skenerjem. Simulacije so bile izvedene na superračunalniški mreži z uporabo GATE programske opreme, program CASToR pa je bil uporabljen za rekonstrukcijo slike. Čerenkov skener z enostranskim branjem detektorjev je deloval podobno kot referenčni skener, medtem ko so skenerji z večstranskim branjem delovali bolje, zahvaljujoč izboljšani učinkovitosti detekcije koincidenc in CTR. Delo dokazuje, da lahko s čistimi Čerenkovimi skenerji, kljub odsotnosti energijske ločljivosti, dobimo primerljive, v določenih detektorskih konfiguracijah pa tudi boljše slike, v primerjavi s sodobnimi skenerji. Znatno boljše slike, pa lahko čisti Čerenkovi skenerji dosežejo v okoljih z nizkim sipanjem, pri pregledih možganov, prsi ali predkliničnih raziskavah. Zaradi potencialne cenovne dostopnosti lahko postanejo Čerenkovi detektorji tudi zelo zanimivi za skenerje celotnega telesa.

• Keywords: TOF PET, svetloba Čerenkova, večstransko branje kristala, PbF$_2$, Geant4/GATE Monte Carlo simulacije, CASToR, NEMA NU 2-2018