In vitro and in silico assessment of flow modulation after deploying the Contour Neurovascular System in intracranial aneurysm models

815 Ko eJ, e al. J Neu oIn e en Su g 2024;16:815–823. doi:10.1136/jnis-2023-020403 O iginal esea ch In i o and in silico assessmen o low modula ion a e deploying he Con ou Neu o ascula Sys em in in ac anial aneu ysmmodels Jana Ko e ,1,2 F anziska Gaidzik ,1,2 Naomi La sen ,3 E ik Schü z,2 Timo Damm,4 F i z Woda g ,3 Jan- Be nd Hö ene ,4 Ola Jansen,3 Gábo Janiga ,1,2 Philipp Be g ,2,5 Ma iya S P a di se a 4 Basic science To ci e: Ko eJ, GaidzikF, La senN, e al. J Neu oIn e en Su g 2024;16:815–823. ►Addi ional supplemen al ma e ial is published online only. To iew, please isi he jou nal online (h p:// dx. doi. o g/ 10. 1136/ jnis- 2023- 020403). 1Labo a o y o Fluid Dynamics and Technical Flows, Uni e si y o Magdebu g, Magdebu g, Ge many 2Resea ch campus STIMULATE, Uni e si y o Magdebu g, Magdebu g, Ge many 3Depa men o Radiology and Neu o adiology, Uni e si y Medical Cen e Schleswig- Hols ein (UKSH), Kiel, Ge many 4Sec ion Biomedical Imaging, Molecula Imaging No h Compe ence Cen e (MOIN CC), Depa men o Radiology and Neu o adiology, Uni e si y Medical Cen e Schleswig- Hols ein (UKSH), Kiel Uni e si y, Kiel, Ge many 5Depa men o Heal hca e Telema ics and Medical Enginee ing, Uni e si y o Magdebu g, Magdebu g, Ge many Co espondence o Jana Ko e; ko e. j@ ou look. com JK and FG con ibu ed equally. PB and MSP con ibu ed equally. Recei ed 6 Ap il 2023 Accep ed 15 July 2023 Published Online Fi s 18Oc obe 2023 © Au ho (s) (o hei employe (s)) 2024. Re- use pe mi ed unde CC BY- NC. No comme cial e- use. See igh s and pe missions. Published by BMJ. ABSTRACT Backg ound The no el Con ou Neu o ascula Sys em (Con ou ) has been epo ed o be e icien and sa e o he ea men o in ac anial, wide- necked bi u ca ion aneu ysms. Flow in he aneu ysm and pos e io ce eb al a e ies (PCAs) a e Con ou deploymen has no been analyzed in de ail ye . Howe e , his in o ma ion is c ucial o p edic ing aneu ysm ea men ou comes. Me hods Time- esol ed h ee- dimensional eloci y maps in 14 combina ions o pa ien - based basila ip aneu ysm models wi h and wi hou Con ou de ices (sizes be ween 5 and 14 mm) we e analyzed using ou - dimensionsal (4D) low MRI and nume ical/image- based low simula ions. A complex i ual p ocessing pipeline was de eloped o mimic he expe imen al shape and posi ion o he Con ou oge he wi h he simula ions. Resul s On a e age, he Con ou signi ican ly educed in a- aneu ysmal low eloci y by 67% (mean w/ = 0.03m/s; mean w/o = 0.12m/s; p- alue=0.002), and he ime- a e aged wall shea s ess by mo e han 87% (mean w/ = 0.17Pa; mean w/o = 1.35Pa; p- alue=0.002), as obse ed by nume ical simula ions. Fu he mo e, a signi ican educ ion in low (P<0.01) was con i med by he neck in low a e, kine ic ene gy, and in low concen a ion index a e Con ou deploymen . No ably, de ice size has a s onge e ec on educing low han de ice posi ioning. Howe e , posi ioning a ec ed low in he PCAs, while being obus in e ec i ely educing low. Conclusions This s udy showed he high e icacy o he Con ou de ice in educing low wi hin aneu ysms ega dless o he exac posi ion. Howe e , we obse ed an e ec on he low in PCAs, which needs o be in es iga ed u he . INTRODUCTION In ac anial aneu ysms (IAs) ca y he isk o up u e and need o be iden i ied and ea ed i his isk is high.1–3 Neu o ascula hemodynamics play a i al ole in IA o ma ion and up u e.4 Indeed, coil emboliza ion has become an es ablished ech- nique o endo ascula ea men o IA.5 Howe e , emboliza ion o wide- necked bi u ca ion aneu- ysms (WNBAs) wi h coils is mo e challenging as i equi es assis ing de ices such as s en s o balloons and esul s in ela i ely low occlusion (40%) and high complica ion (21%) a es.6 In asaccula low dis up o s such as he Wo en EndoB idge (WEB; Mic o en ion/Te umo, Aliso Viejo, CA) we e designed as single implan s o simpli y endo ascula WNBA ea men . Recen ly, a new in asaccula de ice, called Con ou Neu o ascula Sys em (Con ou , Ce us Endo ascula , F emon , CA), was de eloped. Ini ial s udies showed he e icacy and sa e y o ea ing un up u ed IAs wi h his de ice, including WNBAs.7–11 Mo eo e , Con ou has been used o ea acu ely up u ed aneu ysms bo h as a s and- alone de ice and in combina ion wi h coils.12–15 A sys ema ic e iew based on WHAT IS ALREADY KNOWN ON THIS TOPIC ⇒Flow modula ion is an e ec i e ea men o wide- necked aneu ysms. The Con ou Neu o ascula Sys em (Con ou ) is a no el, hal - sphe e- shaped, low- modula ing de ice ha is placed in he aneu ysm neck and comes in di e en sizes. WHAT THIS STUDY ADDS ⇒In a- aneu ysmal low modula ion and he e ec o di e en de ice sizes and posi ions we e quan i ied o he i s ime. ⇒Fou - dimensional low MRI and compu a ional luid dynamics we e used o ob ain ime- esol ed, h ee- dimensional maps o he low inside pa ien - de i ed aneu ysm models wi h and wi hou he Con ou de ice. ⇒The Con ou de ice induced conside able low educ ion o all sizes and deploymen s. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY ⇒The low educ ion caused by he Con ou de ice was cons an and abo e 60%, ega dless o he exac posi ioning o he de ice wi hin a 5° angle, possibly a e ing he need o epea ed de ice eposi ioning. Howe e , he Con ou posi ioning in luences he low di ision in o he pos e io ce eb al a e ies (PCAs), which equi es u he in es iga ion. Mo eo e , he e ec o s onge Con ou disposi ioning (highe ha 5°) on aneu ysm and PCA low is unknown. Zei sch i enab eilung. P o ec ed by copy igh . on Sep embe 9, 2024 a Uni e si a sbiblio hekh p://jnis.bmj.com/J Neu oIn e en Su g: i s published as 10.1136/jnis-2023-020403 on 18 Oc obe 2023. Downloaded om 816 Ko eJ, e al. J Neu oIn e en Su g 2024;16:815–823. doi:10.1136/jnis-2023-020403 Basic science six s udies,7–9 11 14 15 including 131 IAs ea ed wi h ei he Con ou o Con ou and coils, showed a pooled adequa e occlusion a e o 84%.16 Expe ience wi h he Con ou de ice is limi ed. Mos o he a o emen ioned s udies had a e ospec i e design, including angiog aphic (X- ay- based) and clinical neu ological ollow- up. Only one s udy discussed MRI ollow- ups, epo ing s ong me al a i ac s o igina ing om he implan .8 The p ima y pu pose o he Con ou is o dis up and di e he low om an IA. Howe e , o he bes o he au ho s’ knowledge, he di ec low changes caused by he Con ou ha e no been e alua ed ye . In a p e ious expe imen al s udy, he washou ime o an angiog aphic con as agen obse ed wi h digi al sub ac ion angiog aphy (DSA) was used as a su oga e ma ke o he e ec i eness o he de ice.17 Howe e , he anal- ysis was limi ed o a wo- dimensional (2D) e alua ion o he low and migh be ope a o dependen . The e o e, his s udy aimed o quan i y he in a- aneu ysmal changes in low, induced by placing he Con ou in IA models, which we e based on pa ien da a, using ime- dependen , high- dimensional compu a ional luid dynamic (CFD) simula ions. Fu he mo e, ou - dimensional (4D) low MRI expe imen s we e conduc ed o assess he easibili y o e alua ing IA hemodynamics wi h in i o modali y in he p esence o he Con ou de ice. MATERIALS AND METHODS In his s udy, a complex p ocessing pipeline was de eloped, s a ing wi h ou pa ien - based basila ip aneu ysm models. The p ocessing pipeline, comp ising he expe imen al me hod and he in- silico low assessmen , is shown in igu e 1. De ails will be explained wi hin he ollowing subsec ions. Pa ien -based aneu ysm models The ou pa ien - based basila ip aneu ysm models A1–4 ( igu e 1A), we e designed and h ee- dimensionally (3D) p in ed in- house17 18 ( o de ails see online supplemen al ile S1 and able S1). The diame e s (heigh , neck, dome) o he IA sac we e 3.5×2.7×3.2 mm (A1), 6.9×2.8×3.3 mm (A2), 8.4×6.7×8.4 mm (A3), and 16.4×9.2×10.2 mm (A4), espec- i ely ( igu e 1A). IA models we e designed as WNBAs wi h a compa able dome- o- neck a io (1.2±0.1). All models sha ed he same pa en essel and pos e io ce eb al a e ies. The models eady o 3D p in ing a e eely a ailable a Zenodo.19 The i ual model used o CFD simula ions is sligh ly di e en om he one used o 4D low MRI measu emen s. Namely, supe io ce eb al a e ies (SCA) we e ini ially modelled o expe imen s, bu due o hei small size hey we e pa ially occluded du ing he 3D p in ing p ocess. Thus, o p ese e he compa abili y be ween 4D low MRI and CFD he SCAs we e emo ed om he i ual models used o CFD analysis. The exclusion o he b anches does no a ec he change o he hemodynamics in he IAs by placemen o he Con ou sys em analyzed in his s udy (see online supplemen al able S2, online supplemen al igu e S1). Expe imen al me hods Flow se up and in i o de ice deploymen IA models we e in eg a ed in o a closed cycle low se up and supplied wi h saline solu ion a a mean low a e o 150 mL/min o mimic low in he basila a e y obse ed in i o (Isma ec MCP S anda d, Cole Pa me , IL).20 Time- dependen low and p essu e wa e o ms we e measu ed a he inle and ou le s (see online supplemen al igu es S2/S3) and only p essu e a he ip o he IA sac using low and p essu e senso s (ME8PXL- M12, T ansonic Sys em Inc, NY; PRESS- N- 000; PendoTech, NJ), espec i ely ( igu e 2B). Ten Con ou s (C1–10) o h ee sizes—5 mm (C1–5), 11 mm (C6–9), and 14 mm (C10)—we e deployed in o IAs (A1–4) unde luo oscopy (Allu a Xpe FD, Philips, The Ne he lands) by an expe ienced neu o adiologis (>10 yea s o expe ience, FW) (deploymen : A1: C1–2; A2: C3–5; A3: C6–9; A4: C10; o de ails see online supplemen al ile S2). Figu e 1 P ocessing pipeline used o mimic he expe imen s by nume ical CFD simula ions. On he expe imen al side, pa ien - based IA models (A)and he co esponding low se up (B)we e used o acqui e µCT and 4D low MRI da a (C).Flow and p essu e senso s p o ided he bounda y condi ions o CFD. Fo placing he Con ou in he i ual aneu ysm model, i s , he CAD- model o he Con ou was de o med acco ding o µCT images (D).Nex , he posi ion o he Con ou inside he model was de e mined om µCT da a and he de o med Con ou acco dingly placed (E).Finally, CFD simula ions we e ca ied ou and compa ed wi h 4D low MRI indings (F). CAD, Compu e Aided Design; CFD, compu a ional luid dynamic; 4D, ou - dimensional; IA, in ac anial aneu ysm; μCT, mic o- CT; w/o, wi hou . Zei sch i enab eilung. P o ec ed by copy igh . on Sep embe 9, 2024 a Uni e si a sbiblio hekh p://jnis.bmj.com/J Neu oIn e en Su g: i s published as 10.1136/jnis-2023-020403 on 18 Oc obe 2023. Downloaded om 817 Ko eJ, e al. J Neu oIn e en Su g 2024;16:815–823. doi:10.1136/jnis-2023-020403 Basic science Figu e 2 Quali a i e and quan i a i e compa ison o 4D low MRI and CFD eloci y ields. Fo each aneu ysm size, one ep esen a i e case is shown. (A)Veloci y s eamlines be o e Con ou deploymen o A1–A4. (B)His og am plo s be o e Con ou deploymen o he 4D low MRI and CFD eloci y alues a e in e pola ing hem on he same g id. Median alues a e displayed wi h a dashed e ical line. (C)The eloci y magni ude in he co onal plane is displayed on he le ( op ow: wi hou Con ou , bo om ow: wi h Con ou ) o bo h modali ies. Resul s wi hou Con ou show high co espondence be ween 4D low MRI and CFD. Me al a i ac s caused signal oids in he model wi h Con ou and pa ly no eloci ies could be ob ained in 4D low MRI o hese cases (black a ows), despi e o IA sac o A4 C10 (g een dashed ci cle). The highly- esol ed CFD da a can p o ide a de ailed iew o he low in he aneu ysm wi h Con ou . P essu e alues a he aneu ysm ip (senso da a and CFD) a e displayed on he igh . CFD, compu a ional luid dynamic; 4D, ou - dimensional. Zei sch i enab eilung. P o ec ed by copy igh . on Sep embe 9, 2024 a Uni e si a sbiblio hekh p://jnis.bmj.com/J Neu oIn e en Su g: i s published as 10.1136/jnis-2023-020403 on 18 Oc obe 2023. Downloaded om 818 Ko eJ, e al. J Neu oIn e en Su g 2024;16:815–823. doi:10.1136/jnis-2023-020403 Basic science 4D low MRI 4D low MRI da a we e acqui ed by using a 3T whole- body MR sys em equipped wi h a 32- channel head coil (Ingenia CX, R5 V6.1, Philips Heal hca e, Bes , The Ne he lands). Veloci- ies we e quan i ied wi h a ime- esol ed phase- con as MR sequence wi h 3D co e age (4D low MRI, igu e 1C). Tempo al and spa ial esolu ion was 63 ms and (0.75 mm)3, espec i ely. The eloci y- encoding pa ame e was se o 75 cm/s o all IA models, abou 10% highe han he maximum eloci y obse ed a he inle essel wi hou Con ou . Linea o se phase co ec- ion, eloci y aliasing, and essel masking we e pe o med using GT low (V3.1.12, Gy o ools, Swi ze land) ( o de ails see online supplemen al ile S3). In silico low assessmen To accu a ely mimic he exac shape and posi ion o he Con ou by using CFD, he Con ou was digi alized and i ually deployed ( igu e 1) as desc ibed in de ail below. Con ou digi aliza ion To de elop a digi al Con ou eady- o- be- placed in he IA model ( igu e 1D), i s , mic o- CT (μCT) was acqui ed om all Con ou s inside he IA models (a i aCT 80; Scanco Medical AG, B ü isellen, Swi ze land; 45 keV, 80 mm ield o iew, econ- s uc ed o 26 µm iso opic oxel size). He e, a scala mask based on signal in ensi y was c ea ed using a h eshold- based, seeded, egion- g owing algo i hm.21 Nex , he mesh was gene a ed based on he scala mask using a ma ching cubes algo i hm.22 Second, a 3D compu e - aided Con ou design (CAD- Con ou ) was c ea ed based on 2D ep esen a ions o he uncons ained de ice p o ided by he manu ac u e (Fusion 360 2.0, Au odesk Inc, USA). Speci ically, he CAD- Con ou consis ed o 72 ci cles equally spaced om each o he connec ed a he base o he de ice, adjacen o he adiopaque ma ke . In his way, an uncon- s ained model o he Con ou was ob ained, which changed a e deploymen (see igu e 1D). Thus, and hi d, o accu a ely ob ain he shape o he μCT- Con ou , he CAD- Con ou was non- igidly ans o med o he μCT da a, by alignmen o he adiopaque ma ke s. Then, he CAD- Con ou was adap ed o he shape o he μCT- Con ou using a la ice modi ie (Blende , Blende Founda ion, 3.1., Ams e dam, The Ne he lands). The modi ie smoo hly de o med he CAD- Con ou acco ding o he shape o he μCT- Con ou . Hence, a cons ained con igu a ion o he CAD- Con ou was achie ed, which was used o he highly esol ed CFD simula ions. The di ec use o Con ou segmen ed om µCT images was no easible due o he limi ed spa ial esolu ion o he μCT which esul ed in segmen a ion a e ac s such as subs an ially inc eased wi e hickness o he Con ou and he p esence o ully occluded Con ou segmen s (see online supplemen al ile S4, online supplemen al able S3, online supplemen al igu e S4). Vi ual de ice deploymen The i ual IA models and he cons ained CAD- Con ou s we e loca ed in di e en coo dina e sys ems. To ensu e ha he CAD- Con ou s we e co ec ly posi ioned wi hin he IA sac, i s , he 3D- p in ed wall o he aneu ysm model was segmen ed om he same μCT images as he μCT- Con ou ( igu e 1E). The μCT- wall was aligned wi h he CAD- wall using an i e a i e, closes - poin algo i hm (MeshLab 2022.02, ISTI - CNR, Pisa, I aly). Second, he esul ing ans o ma ion ma ix was applied o he CAD- Con ou . As CAD- wall and CAD- aneu ysm lie in an iden ical coo dina e sys em, his single ans o ma ion ensu ed ha he Con ou was co ec ly placed wi hin he CAD- aneu ysm. CFD simula ions Nume ical CFD simula ions we e ca ied ou using a ini e olume sol e (S a CCM+2021.3 16.6, Siemens, E langen, Ge many). Bounda y condi ions ob ained in he expe imen s ( igu e 1C) we e applied a he ex uded inle and ou le s (measu ed mass low and p essu e wa e o ms) o he IA models. Fu he mo e, igid essel walls and he mimicking luid p ope ies we e used (wa e : densi y = 998 kg/m³, dynamic iscosi y = 0.001 Pa·s). Spa ial disc e iza ion o he IA and CAD- Con ou models was pe o med wi h a base cell size o 0.1 mm a he aneu ysm sac and pa en essel, while 0.02 mm was chosen a he Con ou s u s and 0.3 mm a he essel ex usions. The o al cell coun wi hin he IA models wi h Con ou anged om 8.8 million (A1) o 11.6 million cells (A4). The models wi hou Con ou ea u ed a o al cell coun o 2.4 o 4 million cells. In o al, 14 ime- dependen CFD simula ions we e ca ied ou ( ou wi hou : A1–4; 10 wi h Con ou : A1 C1–2, A2 C3–5, A3 C6–9, A4 C10). Tempo al esolu ion was 1 ms o e h ee ca diac cycles, whe eas only he las cycle was analyzed. Cycle leng h o 1.26 s was de e mined om he expe imen . Da a analysis The expe imen al 4D low MRI esul s we e compa ed wi h he calcula ed CFD eloci y by in e pola ing he eloci y ields om 4D low MRI and CFD inside he un ea ed IA sac on o a g id wi h he base size o 0.3 mm. Nex , he changes in he in a- aneu ysmal low a e placing he Con ou we e e alua ed. Namely, oscilla o y shea index (OSI), oscilla o y eloci y index (OVI), neck in low a e (NIR), ime- a e aged wall shea s ess (TAWSS), eloci y (V), kine ic ene gy (KE), in low concen a ion index (ICI), and aneu ysm u no e ime (TOT), which is he aneu ysm sac olume di ided by he NIR, we e e alua ed (see online supplemen al ile S5). Fo each pa ame e (P) wi h (w/) and wi hou (w/o) Con ou , he ea men e ec (TE) was calcu- la ed o C1–10, espec i ely. TE p= Mean P w/ −Mean P w/o Mean Pw/o (1) The ou le low o le and igh pos e io ce eb al a e ies (PCAs) was no malized by he o al ou low. S a is ical analysis was pe o med in MATLAB (MATLAB R2022a, The Ma h- Wo ks, Na ick, MA), using he pai ed Wilcoxon es . Wi h he Bon e oni co ec ion he P alue was se o 0.007. The chosen in low plane o calcula ing NIR is shown in igu e 1F. RESULTS Flow compa ison be ween 4D low MRI and CFD Wi hou Con ou The eloci y ields calcula ed in he IA geome ies wi hou Con ou we e simila o hose measu ed wi h 4D low MRI ( igu e 2A, s eamline and eloci y magni ude images). Fo A4, he low je en e ing he aneu ysm sac was obse ed in bo h modali ies. In CFD, he je appea s o be b oade and sligh ly shi ed o he igh side o he IA’s wall. In A3, he simula ed low je was clea ly isible and he 4D low MRI did no dis inc ly show he low a ached o he wall. Howe e , he o e all low ield emained simila be ween 4D low MRI and CFD. Fo he smalle models (A1 and A2), he highly esol ed CFD simula- ions showed ine eloci y s uc u es, which we e no isible on he 4D low MRI eloci y maps. Gene ally, 4D low MRI esul s su e ed om limi ed spa ial esolu ion, especially o Zei sch i enab eilung. P o ec ed by copy igh . on Sep embe 9, 2024 a Uni e si a sbiblio hekh p://jnis.bmj.com/J Neu oIn e en Su g: i s published as 10.1136/jnis-2023-020403 on 18 Oc obe 2023. Downloaded om 819 Ko eJ, e al. J Neu oIn e en Su g 2024;16:815–823. doi:10.1136/jnis-2023-020403 Basic science small aneu ysms (A1 and A2), and a eloci y noise ha is compa- able o he eloci ies obse ed a he cen e o he aneu ysm and along he walls. Quan i a i ely, as shown in he his og am plo s in igu e 2B, wo smalle sized IAs (A1 and A2) had a highe ela i e equency o low eloci y alues measu ed wi h 4D low MRI (median 0.056 m/s o A1 and 0.023 m/s o A2) when compa ed wi h CFD (median 0.11 m/s o A1 and 0.047 m/s o A2). Fo A3 he median alues o bo h modali ies we e simila (median 0.179 m/s o 4D low MRI and 0.163 m/s o CFD), whe eas o A4, he measu ed alues we e sligh ly highe (median 0.115 m/s o 4D low MRI and 0.09 m/s o CFD). Wi h Con ou S ong MRI a i ac s o igina ing om he adiopaque ma ke o he Con ou ( igu e 2C, black a ows) we e obse ed on 4D low MRI images. All 4D low MRI da a ob ained in he icini y o s ong a i- ac s mus be analyzed ca e ully, as he a i ac s a ec he phase- based low encoding. Fu he mo e, he s ong low educ ion inside he IA leads o nea - ze o eloci ies in he IA sac, which a e close o he 4D low MRI eloci y noise limi . All o his p e en s he use o 4D low MRI o assess pos - ea men IA hemodynamics, especially o small IAs. As a esul , he analysis o eloci y ields measu ed wi h MRI close o he Con ou is impossible and limi ed only o he a i ac - ee a ea. Figu e 3 Vi ual ascula model wi h and wi hou Con ou (A, D),oscilla o y shea index (B, E, OSI), and oscilla o y eloci y index (C, F, OVI). No e he a ia ion in posi ion and o ien a ion o he de ice. G een a ows highligh he highe OSI in he egions close o he Con ou loca ions. Zei sch i enab eilung. P o ec ed by copy igh . on Sep embe 9, 2024 a Uni e si a sbiblio hekh p://jnis.bmj.com/J Neu oIn e en Su g: i s published as 10.1136/jnis-2023-020403 on 18 Oc obe 2023. Downloaded om 820 Ko eJ, e al. J Neu oIn e en Su g 2024;16:815–823. doi:10.1136/jnis-2023-020403 Basic science Fo case A4 C10, low di ision on he le and igh PCAs e ealed highe low h ough he le PCA wi h bo h modali ies ( igu e 2C, co onal plane, g een ci cle). This is o e ing he pe spec i e ha measu ing low by 4D low MRI close o he implan is no comple ely impossible. O e all, he s ong low educ ion induced by he Con ou was de ec ed by CFD and 4D low MRI. Rema kably, he p essu e cu es acqui ed om CFD and senso measu emen s showed excellen ag eemen h oughou he whole ca diac cycle o all cases (mean de ia ion o 6% wi hou and 8% wi h Con ou ). In a-aneu ysmal de ice e icacy The majo i y o he de ices we e loca ed cen ally in he pa en essel ( igu e 3A/D). A2 C5 and A4 C10 we e sligh ly shi ed owa ds he igh and A3 C7 o he le PCA bi u ca ion. Compa ed wi h a e ical middle line h ough he aneu ysm model, he implan ed de ices we e shi ed by an angle be ween 1° and 5°. Conce ning he low pa ame e s OSI ( igu e 3B and E) and OVI ( igu e 3C and F), he in a- aneu ysmal educ ion a e deploymen was isible o C1–4 and o C6–9. Conce ning C5 and C10, OSI and OVI we e inc eased, espec i ely. Fo all cases wi h Con ou s, excep C5 and C10, OSI and OVI we e highe , especially a ound he loca ion o he Con ou (see g een a ows), and lowe wi hin he aneu ysm dome compa ed wi h hose wi hou Con ou . E ec o de ice size A g ea e educ ion in NIR, TAWSS, V, KE, and ICI was ound a e deploying a smalle - sized (83–99%, de ice size 5 mm, A1/ A2 C1–5) han a la ge - sized Con ou (48–95%, de ice size 11 mm, A3 C6–9), as ep esen ed in igu e 4A. Fo A3 (C6–9) a high educ ion o 86–95% is isible only o KE, and o C10 one o 87–100% o TAWSS, V, and KE. S ill, an o e all educ- ion o mo e han 40% is isible o C1–10 o hese pa ame- e s, excep o ICI wi hin C10. The educ ion in OSI and OVI was s able a mo e han 58% o A3 (C6–9). In con as , his Figu e 4 Quan i ica ion o low educ ion o he Con ou . (A) T ea men e ec o chosen hemodynamic pa ame e s in %. (B)No malized mean ou le low di ision wi h low da a de i ed om senso measu emen s in % and he ea men e ec on o he ou le low di ision o he le and igh PCAs. ICI, in low concen a ion index; IR, neck in low a e (mL/s); KE, kine ic ene gy (Pa); OSI, oscilla o y shea index; OVI, oscilla o y eloci y index; PCA, pos e io ce eb al a e y; TAWSS, ime- a e aged wall shea s ess (Pa); V, eloci y (m/s); w/, wi h Con ou ; w/o, wi hou Con ou . Zei sch i enab eilung. P o ec ed by copy igh . on Sep embe 9, 2024 a Uni e si a sbiblio hekh p://jnis.bmj.com/J Neu oIn e en Su g: i s published as 10.1136/jnis-2023-020403 on 18 Oc obe 2023. Downloaded om 821 Ko eJ, e al. J Neu oIn e en Su g 2024;16:815–823. doi:10.1136/jnis-2023-020403 Basic science educ ion was lowe in A1/A2/A4 han in all cases o A3 excep C9 o OVI/OSI e en inc eased. E ec o he aneu ysm heigh The heigh o he aneu ysm sac A2 was almos wo imes la ge han A1 (6.9 mm s 3.5 mm), while he size o he neck and dome was compa able. The e o e, he same Con ou size was implan ed. Rema kably, no speci ic di e ence in he ea men e ec a ibu ed o he di e en aneu ysm heigh s was ound. E ec o he de ice posi ioning Fou de ices (C6–9) we e placed in A3. He e, low educ ion anged wi hin a di e ence o Δ 8.63% (KE) and Δ 13.53% (NIR) o Δ 20.06% (OVI). De ailed TE alues can be ound in online supplemen al able S4. Signi icance Mean NIR, TAWSS, V, KE, ICI, and TOT we e signi ican ly highe (P<0.01) o cases wi hou Con ou , while OVI (P=0.16) and OSI (P=0.11) did no di e signi ican ly be ween g oups (see able 1 and online supplemen al igu e S5). Flow changes in he pos e io ce eb al a e ies The le ou low a ied be ween 44% and 52% and he igh be ween 47% and 56% ( igu e 4B). The TE on he ou low was low a 0.1–5.7%; only in C5 was he e ec s onge (TE=10%). C5 was also mos shi ed o he le PCA (see igu e 3). C6/C7 as compa ed wi h C8/C9 had a smalle e ec on he low al e a ion and he Con ou was placed highe inside he aneu ysm so ha he PCAs we e less a ec ed. DISCUSSION Endo ascula ea men o WNBAs wi h Con ou is a no el echnique ha has no been s udied well ye , bu he ea men esul s in high IA occlusion a es and sa e y.11 14 16 In his s udy, he in a- aneu ysmal low educ ion as well as low al e a ions in he PCAs, ha we e a ec ed by he Con ou , we e analyzed. Aneu ysm models ep esen ing di e en shapes and sizes, oge he wi h he e ec o di e en Con ou deploymen s in he same geome y, we e in es iga ed. In con as o exis ing mini- mally in asi e echniques, Con ou can ea IAs wi h complex shapes ega dless o aneu ysm heigh and does no equi e pos - in e en ional an ipla ele he apy.7 8 Compa ison o 4D low MRI and CFD Due o s ong me al a i ac s wi hin 4D low MRI da a caused by Con ou , i is ad an ageous o use CFD o analyzing he e ec o Con ou deploymen ( igu es 2 and 3). Being he i s - e e nume ical s udy analyzing low al e a ions by Con ou , i was necessa y o compa e he CFD esul s quali a i ely wi h measu ed 4D low MRI eloci y ields in he me al a i ac s- ee egions and quan i a i ely wi h measu ed p essu e senso da a and 4D low MRI da a be o e Con ou deploymen ( igu e 2). Veloci y- encoded s eamlines and his og am plo s showed he highes di e ences be ween bo h modali ies in he smalle IAs (A1 and A2). This is a ibu ed o he ela i ely highe in luence o measu emen noise, 3D- p in ing inhomogenei ies o egis a- ion e o s ha in luence he acqui ed alues. Fu he mo e, he SCAs, which we e pa ly p esen in he 3D p in ed models bu emo ed om he i ual CFD model, migh cause a de ia ion. Howe e , o he la ge IA models (A3 and A4) he eloci y ields a e mo e simila and he main cha ac e is ics ( low je o median alues) a e nea ly in acco dance. The indings a e in ag eemen wi h Sindee e al, who showed he compa ibili y o MRI and CFD wi hin IAs.23 In a-aneu ysmal low The subsequen in- dep h analysis e ealed a s ong low educ ion in all cases o NIR, ICI, and TAWSS and he in a- aneu ysmal low (KE and V). Acco ding o Oua ed e al,24 o low di e e s en (FDS) deploymen , a educ ion in eloci y g ea e han 35% can be conside ed as a success ul occlusion o an aneu ysm. In he p esen s udy, eloci y educ ion was highe han 60% o all cases and hus indica es a po en ially e ec i e occlusion, con i ming he indings o ecen in i o s udies.11 14 16 Aneu ysm occlusion is also ensu ed by he use o a WEB.25 26 Howe e , he sizing o he de ice depends on he aneu ysm wid h and heigh , whe e heigh is usually limi ed o 10 mm. The Con ou is cha ac e ized by i s heigh - independen implemen a- ion as i is placed di ec ly a he neck,15 and aneu ysm heigh did no a ec he e icacy o he Con ou . Compa ing he TE be ween Con ou and FDS, he la e shows a lowe educ ion in NIR (ΔNIR >29%), TAWSS (ΔTAWSS >23%), and V (ΔV >20%) wi hin he aneu ysm.27 Kulcsá e al epo ed ha TAWSS and V educ ion cause IA occlusion, bu hey could no de e mine a p edic i e h eshold alue.28 FDS deploymen ca ies he isk o occluding small la e al b anches and FDS a e di icul o use in bi u ca ion aneu ysms,29 which does no apply o he Con ou .11 Ne e heless, he Coun ou is no well sui ed o small- neck aneu ysms in con as o in- essel de ices such as FDS.8 Compa ed wi h in asaccula coiling, he Con ou showed a simila ly e ec i e low educ ion. S ill, his educ ion a e coiling is no signi ican ly ela ed o aneu ysm occlusion.30 Coiling ca ies he ad an age o conse ing pa en essel low, bu implemen a ion is mo e complex and no sui able o WNBAs wi hou he use o addi ional s en s o balloons.10 In con as o he NIR, TAWSS, and V educ ions, Con ou deploymen en iches he oscilla o y e ec s inside he aneu ysm in some cases ( igu e 4A and able 1). This inding is in line wi h a p e ious s udy in which OVI inc eased inside he IA a e implan ing an FDS.31 Rolo e al31 ound ha FDS malposi- ioning inc eases OSI, bu has no majo e ec on low educ ion. This leads o he assump ion ha adequa e posi ioning o he Con ou could ensu e a dec ease in OSI/OVI ( igu es 3 and 4, able 1). Mo eo e , high OSI co ela ed wi h ecanaliza ion a e coil emboliza ion.32 Table 1 Compa ison o hemodynamic mean esul s be ween wi h (w/) and wi hou (w/o) Con ou including SD and P alue Mean w/ ±SD w/ Mean w/o ±SD w/ P alue NIR* (mL/s) 0.53 ±0.62 1.46 ±1.28 0.002 TAWSS* (Pa) 0.17 ±0.17 1.35 ±0.73 0.002 V* (m/s) 0.03 ±0.03 0.12 ±0.05 0.002 KE* (J) 1.06 ±1.29 16.03 ±11.9 0.002 ICI* 0.64 ±0.84 1.43 ±1.23 0.002 OVI 0.04 ±0.04 0.06 ±0.01 0.16 OSI 0.05 ±0.04 0.08 ±0.02 0.105 TOT* (s) 3.46 ±3.9 0.28 ±0.15 0.002 Signi ican di e ences a e ma ked wi h an as e isk*. ICI, in low concen a ion index; KE, kine ic ene gy; NIR, neck in low a e; OSI, oscilla o y shea index; OVI, oscilla o y eloci y index; TAWSS, ime- a e aged wall shea s ess; TOT, aneu ysm u no e ime ; V, eloci y. Zei sch i enab eilung. P o ec ed by copy igh . on Sep embe 9, 2024 a Uni e si a sbiblio hekh p://jnis.bmj.com/J Neu oIn e en Su g: i s published as 10.1136/jnis-2023-020403 on 18 Oc obe 2023. Downloaded om 822 Ko eJ, e al. J Neu oIn e en Su g 2024;16:815–823. doi:10.1136/jnis-2023-020403 Basic science The calcula ed aneu ysm TOT ( able 1) is a measu e o low s asis and has been s udied as a po en ial ma ke o h ombus o ma ion, and subsequen ly IA occlusion.33–36 The mean TOT in all models we e signi ican ly highe a e Con ou deploymen , hus con i ming he posi i e ea men e ec o he Con ou and inc easing he chance o h ombus o ma ion. These esul s a e in acco dance wi h he washou ime calcula ions obse ed wi h DSA in he same aneu ysm models.17 Ou le low al e a ion Fu he mo e, due o i s shape, acco ding o ou indings, he Con ou dis u bs ou le essel low by al e ing low di ision in o he PCAs, depending on he posi ioning ( igu e 4B). S ill, he in a- aneu ysmal low educ ion emained independen om he de ice posi ioning. Thus, an angula shi o 5° a ec s low al e - a ion h ough he PCAs bu no in a- aneu ysmal low educ ion. Howe e , i emains unce ain i his esul can be eplica ed in i o. Likely, he deploymen o he Con ou would no al e he low demand in he dis al ascula beds, esul ing in simila ou low bounda y condi ions. As a esul , minimal changes in low di ision among he PCAs can be expec ed ollowing Con ou deploymen . The e o e, hese indings equi e u he in es iga ion. In compa ison, pa en essel low is conse ed by deploying WEB and coiling as no pa o he de ice is placed ou side he IA.25 37 Limi a ions This s udy has se e al limi a ions. Fi s , 4D low MRI migh be a ec ed by insu icien spa ial esolu ion,38 especially wi hin small IAs. Mo eo e , MR images we e impai ed by me al a i- ac s, mos ly due o he adiopaque ma ke . The me al a i ac s caused by Con ou seem o be d as ically s onge han o FDS.39 Thus, compa ison o he CFD simula ions o 4D low was mainly limi ed o he cases wi hou Con ou and o he p essu e senso measu emen s. Second, while e o s we e made o ealis ically mimic he Con ou placemen , mino misalignmen s we e obse ed. S ill, i was he bes app oach a ailable since di ec use o µCT segmen ed Con ou s was no possible due o segmen a ion a i- ac s (online supplemen al ile S4). Ne e heless, his is he i s s udy o i ually mimic Con ou placemen ealis ically, enabling nume ical analysis o ele an hemodynamic pa ame e s. Thi d, he inpu low wa e o ms o CFD we e no pa ien - speci ic bu mimicked he shape o a ca diac cycle and mean low a e epo ed in a basila a e y in i o (see online supplemen al igu es S2 and S3). Fou h, o compa e hemodynamic pa ame e s be ween he cu en nume ical s udy and a p e ious expe imen al s udy17 (speci ically TOT s washou ime (WOT)), a saline solu- ion was u ilized o bo h expe imen s and CFD simula ions. The iscosi y o saline di e s om ha o blood, po en ially hinde ing a di ec compa ison o he esul s in his s udy o in i o IA hemodynamics. Howe e , addi ional simula ions did no e eal a subs an ial impac o iscosi y on he ea men e ec o he Con ou analyzed in his s udy (online supplemen al igu e S6). Fi h, only IAs o he basila a e y we e conside ed in his s udy, and all models had iden ical pa en essel and pos e io ce eb al a e ies. This limi s he gene alizabili y o he ob ained esul s. In addi ion, SCAs we e emo ed om he i ual IA model due o small essel diame e s and essel occlusion a e 3D p in ing. This p e en ed he analysis o low h ough SCAs. Ne e heless, he exclusion o he b anches did no a ec he ea men e ec o he Con ou in he IA (see online supple- men al able S2, online supplemen al igu e S1). Six h, due o he complex wo k low and esou ce- in ensi e na u e o he s udy, only 10 cases we e ini ially conside ed. To enhance he obus ness o he indings, a la ge sample size should be included in u u e in es iga ions. Las , his s udy aises impo an clinical ques ions ha canno be ully add essed wi hin he scope o a single s udy. These include he impac o Con ou placemen on low in he PCA, he obus ness o he Con ou e iciency om de ice posi- ioning, namely he e ec o mo e p onounced (>5°) de ice angula ion, and he p edic ion o IA occlusion s a us based on he low educ ion. Add essing hese ques ions will be he ocus o u u e esea ch. CONCLUSIONS In his s udy, he e ec i eness o Con ou was shown and, o he i s ime, low was analyzed in de ail, quan i a i ely add essing IA low educ ion. Con ou educed in a- aneu ysmal eloci y and TAWSS o all cases. O e all compa ison be ween wi h and wi hou Con ou showed a signi ican educ ion in he chosen hemodynamic pa ame e s. De ice size has a g ea e e ec on educing low han does posi ioning. Howe e , posi ioning in lu- ences he low di ision in o he PCAs, which equi es u he in es iga ion. Co ec ion no ice Since his pape i s published, he symbol * has been added o he ca ego y TOT in able 1. X Naomi La sen @naomila sen6 and Ma iya S P a di se a @MP a di se a Acknowledgemen s We acknowledge Philips Heal hca e o suppo in 4D low MRI expe imen s. We hank Ce us Endo ascula o p o iding he Con ou Neu o ascula Sys em o in i o es ing. We acknowledge suppo by Kiel Uni e si y and he Medical Facul y o suppo ing he Molecula Imaging No h Compe ence Cen e (MOIN CC) as co e acili y o imaging in i o. Con ibu o s Concep ualiza ion and design: JK, FG, MSP, NL, PB. Implan a ion o he Con ou s and choice o he aneu ysm models: MSP and FW. Acquisi ion o MRI da a and cons uc ion o he 3D p in ed models: MSP. Acquisi ion o mCT da a: TD. De elopmen o he i ual p e- p ocessing pipeline: JK, FG, ES. Conduc ion o CFD simula ions: JK, FG. Da a analysis and in e p e a ion: NL, JK, FG, MP, PB. Da a isualiza ion: JK, FG. S a is ical analysis: JK. Supe ision and unding acquisi ion: PB, NL, FW, JBH, OJ. W i ing – o iginal d a p epa a ion: JK, FG, MSP. W i ing – e iew and edi ing: JK, FG, MSP, NL, TD, FW, JSH, OJ, PB, GJ. Gua an o : JK. All au ho s ha e ead and ag eed o he published e sion o he manusc ip . JK and FG con ibu ed equally o his pape and sha e i s au ho ship. MP and PB con ibu ed equally o his pape and sha e las au ho ship. Funding This wo k was suppo ed by he Ge man Resea ch Founda ion (GRK2154: P2b, SPP2311: p ojec no. 465189657), he Ge man Fede al Minis y o Educa ion and Resea ch wi hin he Resea ch Campus STIMULATE (g an no. 13GW0473A), and he B uhn Founda ion. MOIN CC was ounded by a g an om he Eu opean Regional De elopmen Fund (ERDF) and he Zukun sp og amm Wi scha o Schleswig- Hols ein (p ojec no. 122- 09- 053). Compe ing in e es s The au ho (s) decla ed he ollowing po en ial con lic s o in e es wi h espec o he esea ch, o his a icle: MSP has ecei ed a esea ch g an om Ce us endo ascula ; FW has ecei ed ees o consul ing and speaking om Bal In e na ional SAS and Ce us Endo ascula . Pa ien consen o publica ion No applicable. E hics app o al No applicable. P o enance and pee e iew No commissioned; ex e nally pee e iewed. Da a a ailabili y s a emen MRI, CFD, low and p essu e senso da a a e a ailable upon easonable eques . Digi al aneu ysm models eady o 3D p in ing a e a ailable o download a he Zenodo esea ch eposi o y [doi: 10.5281/ zenodo.4723100]. Supplemen al ma e ial This con en has been supplied by he au ho (s). I has no been e ed by BMJ Publishing G oup Limi ed (BMJ) and may no ha e been pee - e iewed. Any opinions o ecommenda ions discussed a e solely hose o he au ho (s) and a e no endo sed by BMJ. BMJ disclaims all liabili y and esponsibili y a ising om any eliance placed on he con en . Whe e he con en Zei sch i enab eilung. P o ec ed by copy igh . on Sep embe 9, 2024 a Uni e si a sbiblio hekh p://jnis.bmj.com/J Neu oIn e en Su g: i s published as 10.1136/jnis-2023-020403 on 18 Oc obe 2023. Downloaded om 823 Ko eJ, e al. J Neu oIn e en Su g 2024;16:815–823. doi:10.1136/jnis-2023-020403 Basic science includes any ansla ed ma e ial, BMJ does no wa an he accu acy and eliabili y o he ansla ions (including bu no limi ed o local egula ions, clinical guidelines, e minology, d ug names and d ug dosages), and is no esponsible o any e o and/o omissions a ising om ansla ion and adap a ion o o he wise. Open access This is an open access a icle dis ibu ed in acco dance wi h he C ea i e Commons A ibu ion Non Comme cial (CC BY- NC 4.0) license, which pe mi s o he s o dis ibu e, emix, adap , build upon his wo k non- comme cially, and license hei de i a i e wo ks on di e en e ms, p o ided he o iginal wo k is p ope ly ci ed, app op ia e c edi is gi en, any changes made indica ed, and he use is non- comme cial. See:h p://c ea i ecommons.o g/licenses/by-nc/4.0/. ORCID iDs JanaKo e h p://o cid.o g/0000-0002-0434-712X F anziskaGaidzik h p://o cid.o g/0000-0003-1924-821X NaomiLa sen h p://o cid.o g/0000-0002-8782-8278 F i zWoda g h p://o cid.o g/0000-0003-1413-2699 Gábo Janiga h p://o cid.o g/0000-0002-4560-9640 PhilippBe g h p://o cid.o g/0000-0003-4140-6771 Ma iya SP a di se a h p://o cid.o g/0000-0001-9442-088X REFERENCES 1 Wiebe s DO, Whisnan JP, Hus on J, e al. Un up u ed in ac anial aneu ysms: na u al his o y, clinical ou come, and isks o su gical and endo ascula ea men . Lance 2003;362:103–10. 2 Hackenbe g KAM, Hänggi D, E minan N. Un up u ed in ac anial aneu ysms. S oke 2018;49:2268–75. 3 B isman JL, Song JK, Newell DW. Ce eb al aneu ysms. N Engl J Med 2006;355:928–39. 4 Soldozy S, No a P, Elsa ag M, e al. The biophysical ole o hemodynamics in he pa hogenesis o ce eb al aneu ysm o ma ion and up u e. Neu osu g Focus 2019;47:2019.4.FOCUS19232. 5 Molyneux AJ, Bi ks J, Cla ke A, e al. The du abili y o endo ascula coiling e sus neu osu gical clipping o up u ed ce eb al aneu ysms: 18 yea ollow- up o he UK coho o he In e na ional Suba achnoid Aneu ysm T ial (ISAT). Lance 2015;385:691–7. 6 Fio ella D, A hu AS, Chiacchie ini R, e al. How sa e and e ec i e a e exis ing ea men s o wide- necked bi u ca ion aneu ysms? Li e a u e- based objec i e pe o mance c i e ia o sa e y and e ec i eness. J Neu oin e Su g 2017;9:1197–201. 7 Liebig T, Kille - Obe p alze M, Gal G, e al. The sa e y and e ec i eness o he Con ou Neu o ascula Sys em (Con ou ) o he ea men o bi u ca ion aneu ysms: he CERUS s udy. Neu osu ge y 2022;90:270–7. 8 Akhunbay- Fudge CY, Deniz K, Tyagi AK, e al. Endo ascula ea men o wide- necked in ac anial aneu ysms using he no el Con ou Neu o ascula Sys em: a single- cen e sa e y and easibili y s udy. J Neu oin e Su g 2020;12:987–92. 9 Bhogal P, Chudyk J, Bleise C, e al. T ea men o un up u ed, andem aneu ysms o he ICA wi h a single low di e e . Clin Neu o adiol 2019;29:725–31. 10 Tho mann M, Mpo sa is A, Behme D. T ea men o a middle ce eb al a e y bi u ca ion aneu ysm wi h he no el Con ou Neu o ascula Sys em compa ible wi h 0.021″ ca he e s. Neu o adiol J 2022;35:396–8. 11 Biondi A, P imiki is P, Vi ale G, e al. Endosaccula low dis up ion wi h he Con ou Neu o ascula Sys em: angiog aphic and clinical esul s in a single- cen e s udy o 60 un up u ed in ac anial aneu ysms. J Neu oin e Su g 2023;15:838–43. 12 Ca asin N, Ga a UA, Mag ini S, e al. T ea men o a up u ed e al- ype pos e io communica ing aneu ysm wi h a combined app oach using he new Con ou Neu o ascula Sys em. BMJ Case Rep 2022;15:e248065. 13 Dange NN, Roy JM. Ini ial expe ience wi h he Con ou de ice in he ea men o up u ed in ac anial wide- necked bi u ca ion aneu ysms: a single- ope a o , mul icen e s udy. Neu ol India 2022;70:2059–64. 14 Woda g F, Ozpeyni ci Y, Hensle J, e al. Con ou - assis ed coiling wi h jailed mic oca he e may esul in be e occlusion (CoCoJaMBO) in wide- necked in ac anial aneu ysms: p oo o p inciple and immedia e angiog aphic esul s. In e Neu o adiol 2023;29:79–87. 15 Diana F, de Dios Lascue as M, Peschillo S, e al. In asaccula low dis up o - assis ed coiling o in ac anial aneu ysms using he no el Con ou Neu o ascula Sys ems and NEQSTENT: a single- cen e sa e y and easibili y s udy. B ain Sci 2022;12:991. 16 Ghozy S, Lashin BI, El il M, e al. The sa e y and e ec i eness o he Con ou Neu o ascula Sys em o he ea men o wide- necked aneu ysms: a sys ema ic e iew and me a- analysis o ea ly expe ience. In e Neu o adiol 2022:15910199221139546. 17 P a di se a MS, P a di se AN, Pe e s S, e al. The e ec o he size o he new Con ou neu o ascula de ice o al e ing in aaneu ysmal low. In e Neu o adiol 2023:15910199221145985. 18 P a di se a MS, Peschke E, Lindne T, e al. 3D- p in ed, pa ien - speci ic in ac anial aneu ysm models: om clinical da a o low expe imen s wi h endo ascula de ices. Med Phys 2021;48:1469–84. 19 P a di se a MS. 3D models o basila ip in ac anial aneu ysms (STL) o s udy low dynamics. 2021. 20 Za inkoob L, Amba ki K, Wåhlin A, e al. Blood low dis ibu ion in ce eb al a e ies. J Ce eb Blood Flow Me ab 2015;35:648–54. 21 Adams R, Bischo L. Seeded egion g owing. IEEE T ans Pa e n Anal Machine In ell 1994;16:641–7. 22 Lo ensen WE, Cline HE. Ma ching cubes: a high esolu ion 3D su ace cons uc ion algo i hm. SIGGRAPH Compu G aph 1987;21:163–9. 23 Sindee S, A nold PG, F olo S, e al. Phase- con as MRI e sus nume ical simula ion o quan i y hemodynamical changes in ce eb al aneu ysms a e low di e e ea men . PLoS One 2018;13:e0190696. 24 Oua ed R, La abide I, B ina O, e al. Compu a ional luid dynamics analysis o low educ ion induced by low- di e ing s en s in in ac anial aneu ysms: a pa ien - unspeci ic hemodynamics change pe spec i e. J Neu oin e Su g 2016;8:1288–93. 25 Pie o L, Sziko a I, Ba eau X, e al. Aneu ysm ea men wi h he Wo en EndoB idge (WEB) de ice in he combined popula ion o wo p ospec i e, mul icen e se ies: 5- yea ollow- up. J Neu oin e Su g 2023;15:552–7. 26 Pagano P, Paiusan L, Soize S, e al. In ac anial aneu ysm ea men wi h in asaccula low dis up ion: compa ison o WEB- 21 and WEB- 17 sys ems. J Neu oin e Su g 2022;14:904–9. 27 Damiano RJ, Ma D, Xiang J, e al. Fini e elemen modeling o endo ascula coiling and low di e sion enables hemodynamic p edic ion o complex ea men s a egies o in ac anial aneu ysm. J Biomech 2015;48:3332–40. 28 Kulcsá Z, Augsbu ge L, Reymond P, e al. Flow di e sion ea men : in a- aneu ismal blood low eloci y and WSS educ ion a e pa ame e s o p edic aneu ysm h ombosis. Ac a Neu ochi 2012;154:1827–34. 29 B igan i F, Leone G, Ma seglia M, e al. Endo ascula ea men o ce eb al aneu ysms using low- di e e de ices: a sys ema ic e iew. Neu o adiol J 2015;28:365–75. 30 Goube g i s L, Schalle J, Ke zsche U, e al. Hemodynamic impac o ce eb al aneu ysm endo ascula ea men de ices: coils and low di e e s. Expe Re Med De ices 2014;11:361–73. 31 Rolo C, Be g P. E ec o low di e e s en malposi ion on in ac anial aneu ysm hemodynamics—an expe imen al amewo k using s e eoscopic pa icle image elocime y. PLoS One 2022;17:e0264688. 32 Kim T, Oh CW, Bang JS, e al. Highe oscilla o y shea index is ela ed o aneu ysm ecanaliza ion a e coil emboliza ion in pos e io communica ing a e y aneu ysms. Ac a Neu ochi 2021;163:2327–37. 33 Ngoepe MN, F angi AF, By ne JV, e al. Th ombosis in ce eb al aneu ysms and he compu a ional modeling he eo : a e iew. F on Physiol 2018;9:306. 34 La abide I, Aguila ML, Mo ales HG, e al. In a- aneu ysmal p essu e and low changes induced by low di e e s: ela ion o aneu ysm size and shape. AJNR Am J Neu o adiol 2013;34:816–22. 35 Liebe BB, Sadasi an C. Endoluminal sca olds o ascula econs uc ion and exclusion o aneu ysms om he ce eb al ci cula ion. S oke 2010;41:S21–5. 36 O’kelly CJ, K ings T, Fio ella D, e al. A no el g ading scale o he angiog aphic assessmen o in ac anial aneu ysms ea ed using low di e ing s en s. In e Neu o adiol 2010;16:133–7. 37 Sheng B, Wu D, Yuan J, e al. Hemodynamic cha ac e is ics associa ed wi h pa aclinoid aneu ysm ecu ence in pa ien s a e emboliza ion. F on Neu ol 2019;10:429. 38 P a di se a MS, Gaidzik F, Be g P, e al. In luence o spa ial esolu ion and comp essed SENSE accele a ion ac o on low quan i ica ion wi h 4D low MRI a 3 Tesla. Tomog aphy 2022;8:457–78. 39 Bouillo P, B ina O, Dela e BMA, e al. Neu o ascula s en a i ac s in 3D- TOF and 3D- PCMRI: in luence o s en design on low measu emen . Magn Reson Med 2019;81:560–72. Zei sch i enab eilung. P o ec ed by copy igh . on Sep embe 9, 2024 a Uni e si a sbiblio hekh p://jnis.bmj.com/J Neu oIn e en Su g: i s published as 10.1136/jnis-2023-020403 on 18 Oc obe 2023. Downloaded om