Advanced hepatic vasculobiliary imaging segmentation and 3D reconstruction as an aid in the surgical management of high biliary stenosis

Pe ei adaSil ae al. BMC Med Imaging (2020) 20:120 h ps://doi.o g/10.1186/s12880-020-00520-0 CASE REPORT Ad anced hepa ic asculobilia y imaging segmen a ion and3D econs uc ion asanaid in hesu gical managemen o high bilia y s enosis Nuno Pe ei a da Sil a1* , Inês Ab eu1, Ma co Se ôdio2, Luís Fe ei a2, Hen ique Alexand ino2,3,4,5 and Paulo Dona o1,3 Abs ac Backg ound: Th ee-dimensional (3D) models a e inc easingly used o help su geons, guiding hem h ough he complex hepa ic asculobilia y ana omy. The bilia y ac is a ela i ely un apped e i o y wi h only a ew case epo s desc ibed in medical li e a u e. Ou aim is o p esen an inno a i e 3D econs uc ion me hodology o bilia y imaging and su gical planning, applied o a case o ia ogenic bilia y s ic u e, wi h usion o segmen ed CT and MRI images. Case p esen a ion: A selec ed case o Bismu h ype III ia ogenic bilia y s enosis o 3D planning. CT and MR s udies we e acqui ed wi h dedica ed p o ocols o segmen a ion. Two adiologis s pe o med segmen a ion and 3D model pos -p ocessing, using bo h imaging echniques o ai h ully ende he ana omical s uc u es. Measu emen s o ana omical landma ks we e aken in bo h he CT/MRI and he 3D model o assu e i s accu acy and di e ences in measu emen we e calcula ed. The 3D model eplica es ana omical s uc u es and pa hology wi h high accu acy, wi h only 2.2% a ia ion be ween STL, CT and MRI measu emen s. The model was discussed wi h he su gical eam and used in he su gical planning, imp o ing con idence in his delica e p ocedu e, due o he de ailed p io knowledge o he pa ien ’s ana omy. Conclusion: Th ee-dimensional econs uc ions a e a apidly g owing a ea o esea ch wi h a signi ican impac in he pe sonalized and p ecision medicine. The cons uc ion o 3D models ha combine ascula and bilia y ana omy, using di e en imaging echniques, espec i ely CT and MRI, will p edic ably con ibu e o a mo e igo ous planning o complex li e su ge ies. Keywo ds: Th ee-dimensional (3D) model, Bilia y s ic u e, Hepa obilia y imaging, P e-ope a i e planning, Pe sonalized medicine, Image pos -p ocessing, Compu ed omog aphy, Magne ic esonance imaging, Case epo © The Au ho (s) 2020. Open Access This a icle is licensed unde a C ea i e Commons A ibu ion 4.0 In e na ional License, which pe mi s use, sha ing, adap a ion, dis ibu ion and ep oduc ion in any medium o o ma , as long as you gi e app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o he C ea i e Commons licence, and indica e i changes we e made. The images o o he hi d pa y ma e ial in his a icle a e included in he a icle’s C ea i e Commons licence, unless indica ed o he wise in a c edi line o he ma e ial. I ma e ial is no included in he a icle’s C ea i e Commons licence and you in ended use is no pe mi ed by s a u o y egula ion o exceeds he pe mi ed use, you will need o ob ain pe mission di ec ly om he copy igh holde . To iew a copy o his licence, isi h p://c ea i eco mmons .o g/licen ses/by/4.0/. The C ea i e Commons Public Domain Dedica ion wai e (h p://c ea i eco mmons .o g/publi cdoma in/ze o/1.0/) applies o he da a made a ailable in his a icle, unless o he wise s a ed in a c edi line o he da a. Backg ound Th ee-dimensional (3D) p in ing is a apidly g owing a ea o esea ch wi h a signi ican impac in pe sonalized and p ecision medicine. The design o 3D models and 3D p in ing ha e been de eloped o di e en medical pu poses. 3D models, o en p in ed, a e widely used as su gical aids in o hopedic and maxillo acial su ge y due o he ela i e ease in bone segmen a ion [1]. Li e models a e also inc easingly used o help su geons, guiding hem h oughou he complex li e ana omy, namely in umo esec ions and p e- ansplan s udies [2–6]. Open Access *Co espondence: nunoapsil [email protected] 1 Medical Imaging Depa men , Coimb a Uni e si y Hospi al Cen e , P ace a P o . Mo a Pin o, 3000-075 Coimb a, Po ugal Full lis o au ho in o ma ion is a ailable a he end o he a icle Page 2 o 9 Pe ei adaSil ae al. BMC Med Imaging (2020) 20:120 A sys ema ic e iew by Pe ica e al. [7] con i med he added alues o he models on hese su gical p oce- du es. The bilia y ac howe e is, as o ye , a ela i ely un apped e i o y wi h only a ew case epo s desc ibed in medical li e a u e [4, 8, 9]. Comple e bilia y s enosis esul ing om ia ogenic inju y can be li e- h ea ening. Su gical epai wi h Roux- en-Y hepa icojejunos omy is he gold s anda d he apy as i ensu es bilia y ou low and econs i u es bilio-en e ic d ainage. Gi en he high densi y and equen ana omic a iabili y o asculo-bilia y s uc u es in he li e hilum, me iculous p eope a i e planning is c ucial. This equi es excellen p eope a i e imaging, usually wi h mul iple modali ies, such as compu ed omog aphy (CT) and magne ic esonance—wi h cholangiopanc ea og aphy (MRCP). Howe e , hese di e en modali ies will p o ide dis inc ea u es ha need h ee-dimensional in eg a ion by he su gical eam o plan he app oach [10]. Fo ins ance, he a e ial and po al ascula ana omy a e eliably demons a ed by he CT-angiog aphy, while bilia y ana omy is bes depic ed by MRCP. An in eg a ed app oach o hese images would be ex emely help ul as i would p o ide a oadmap o his challenging ana omical a ea, u he compounded wi h adhesions and in lamma- o y changes. We p esen a case o a pa ien wi h an ia ogenic bil- ia y s ic u e in which we used an inno a i e 3D econ- s uc ion me hodology, using images om wo dis inc modali ies, one o ascula illus a ion (CT) and o he o bilia y ee ana omy demons a ion (MRCP). We belie e ha his app oach can be used in o he con- ex s o complex su gical, endoscopic and in e en- ional adiology p ocedu es, pa icula ly in cases o benign o malignan bilia y s enosis, as well as in li e ansplan a ion. Case p esen a ion A 50-yea -old male pa ien p esen ed o he eme gency depa men wi h jaundice, e e (39°C) and cholu ia in he p e ious 2days. The pa ien had unde gone cholecys- ec omy o cholecys oli hiasis and Mi izzi synd ome wo yea s ea lie (Fig.1). He epo ed simila episodes in he p e ious mon hs, wi h milde symp oms. He also had pe sonal medical his o y o G a es’ dis- ease, ulce a i e coli is and clea cell ca cinoma o he kidney (ha ing unde gone lapa oscopic igh pa ial neph ec omy 10yea s ea lie ). The labo a o y s udies showed ele a ion o li e enzymes, mainly gamma glu amyl anspep idase (GGT) and o al and di ec bili ubin, o 789 U/L [N: < 55U/L], 10.1mg/dL [N:0.3–1.2mg/dL] and 6.7mg/dL [N:0.1– 0.3mg/dL], espec i ely, and a sligh ele a ion o aspa - a e amino ans e ase (AST), alanine amino ans e ase (ALT) and alkaline phospha ase (ALP). C- eac i e p o- ein was also ele a ed (7.12mg/dL; N:0–0,.mg/dL). The pa ien was admi ed o medical ea men o cholangi is and unde wen imaging es s o cha ac e ize he bile duc s, wi h suspicion o ia ogenic/pos in lam- ma o y bilia y s enosis. A MRI was pe o med wi h MRCP p o ocol and wi h hepa obilia y con as , con i ming dila a ion o he in ahepa ic bile duc s wi h an ab up s op nea he bil- ia y con luence, in close con ac wi h he i s po ion o he duodenum. In he hepa obilia y phase, con as was exc e ed o he duodenal bulb and no con as was p esen in he choledochal duc , seemingly con i ming Fig. 1 Co onal and Axial Abdominal in he po al enous phase. A calci ied gallbladde s one is isible in he in undibulum (a ow) causing comp ession and dila ion o he common bile duc (a owhead), he Mi izzi synd ome Page 3 o 9 Pe ei adaSil ae al. BMC Med Imaging (2020) 20:120 hepa icoduodenal is ula. S enosis o he main bilia y ac was con i med, wi h app oxima ely 3cm in leng h (Figs.2, 3). The pa ien was diagnosed wi h Bismu h ype III s e- nosis a e cholecys ec omy, associa ed wi h hepa i- coduodenal is ula. Gi en he young pa ien age and a e discussion in a mul idisciplina y se ing, he pa ien was o e ed su gical econs uc ion. Abdominal iphasic CT was also pe o med o de ec po en ial associa ed ascu- la inju y and o ob ain a p eope a i e ana omical map- ping. The imaging p o ocols used in MRI and CT a e de ailed below. 3D planning‑CT andMR p o ocols P e-con as and iphasic pos -con as abdomi- nal CT was pe o med on a 64-slice CT scanne (GE Ligh Speed VCT) wi h he ollowing scanning p o- ocol: de ec o collima ion 64-slice × 0.625 mm wi h gan y o a ion o 500 ms, 120 kVp and 134 mAs, a slice hickness o 0.625mm wi h esul ing oxel size o 0.625 × 0.625 × 0.625 mm3 (Fig.4). The MRI was pe o med in a 3T machine (Magne om p isma i -Siemens Medical Solu ions, E langen, Ge - many) using axial and co onal HASTE; axial in-phase and ou -o -phase T1-weigh ed imaging; axial T2-weigh ed a Fig. 2 MRI one yea a e cholecys ec omy. Co onal HASTE (a) shows in a-hepa ic bile duc dila ion wi h an ab up s op a he con luence, in close con ac wi h he duodenum (a ow). In he hepa obilia y phase (b) con as was exc e ed o he duodenal bulb (a ow) and no con as was depic ed in he choledochus (a owhead), con i ming is ulisa ion Fig. 3 MR cholangiog aphy. a Volume ende ing (VR); b maximum in ensi y p ojec ion (MIP) 50 mm. Bo h e o ma ions show in a-hepa ic bile duc dila ion wi h ab up s op a he con luence. On B, he dis al choledochus is iden i ied (a ow) wi h no signs o choledocholi hiasis. A s enosis o he common bile duc wi h a leng h o 3 cm was diagnosed Page 4 o 9 Pe ei adaSil ae al. BMC Med Imaging (2020) 20:120 sa u a ed sequence; 3D colangiog aphic sequence; axial DWI b-50-100-800 and a p econ as and pos con as T1-weigh ed 3D VIBE a e 10mL endo enous gadox- e ic acid injec ion on a e ial, po al, ansi ional and hepa obilia y phases. The segmen a ion was pe o med in he hepa obilia y phase, a 3D Dixon sequence, 20min a e injec ion, wi h a 320 × 320 ma ix and ield o iew (FOV) o 320 × 280 (Figs.2,3). 3D planning: image pos ‑p ocessing andsegmen a ion O iginal CT and MRCP images in Digi al Imaging and Communica ions in Medicine (DICOM) o ma we e expo ed o a sepa a e wo ks a ion wi h In elliSpace Po - al, e sion 9 (Phillips Heal hca e, Bes , The Ne he lands) o pos -p ocessing and segmen a ion. Two adiologis s ained in 3D p in ing pe o med seg- men a ion o MR and CT images. The li e , po al and hepa ic eins we e segmen ed using in he po al enous phase, he coeliac unk and hepa ic a e y using he a e ial phase imaging. The in ahepa ic bile duc s we e segmen ed in bo h he CT (po al enous phase) and he hepa obilia y phase o he hepa ic MR. Rega ding he segmen a ion echniques, bo h manual and semi-au o- ma ic segmen a ion echniques we e used wi h di e - en me hods in o de o ai h ully ende he ana omical s uc u es including h esholding echnique, edge de ec- ion, egion g owing and mul iple slice in e pola ion. The la ge olume s uc u es, such as he li e and po al ein, we e ini ially segmen ed wi h semi-au oma ic selec- ion ools o edge de ec ion and h esholding, he con- ou s we e de ined and adjus ed manually. The smalle calibe essels and bilia y duc s we e segmen ed s ep-by- s ep, wi h high p ecision selec ion ools equi ing manual delinea ion using egion g owing echnique. The inal da ase was expo ed in o S anda d Tessella- ion Language (STL) o be iewed and p ocessed wi h he Meshmixe 3.5 (Au odesk, Inc). Using he CT STL as a e e ence, he MRI STL was ini ially supe imposed using h ee-dimensional Ca esian coo dina es and hen manu- ally adjus ed o achie e a eliable o e lap, using e e ence poin s placed along he igh and le hepa ic duc s, he pos e io b anch o he igh hepa ic duc and on he con e gences o mo e dis al sec o al duc s on bo h CT and MR STLs. The o iginal (CT) model and he used model (CT + MR) we e e iewed join ly by he adiologis s and he su gical eam, concluding ha he bile duc s we e mo e accu a ely segmen ed using MRCP han CT, due o i s supe io con as esolu ion. Du ing he edi ing p o- cess, some de o mi ies and ee objec s we e emo ed and he bo de s we e smoo hed (Fig.5). 3D planning: esul s Measu emen s o ana omical landma ks we e aken in bo h he CT/MRI and he 3D model o assu e i s accu- acy. The s uc u es we e measu ed h ee imes by bo h Radiologis s o each echnique and he mean alue was used. The s uc u es measu ed we e: he igh and le hepa ic duc , main po al ein, p ope hepa ic a e y and igh and le b anches o he hepa ic a e y. Di e ences in measu emen s o he diame e o he a o emen ioned s uc u es we e calcula ed be ween bo h MR, CT and STL (Fig. 6), using he ollowing o mula: % di e ence = (A − B) (A+B) 2 × 100 . The pe cen age o di e ence in he measu emen using CT only and used CT/MR we e low bu he ag eemen was sligh ly be e when using used imaging (5.2% and 2.1% espec i ely) (Table1). Fig. 4 Abdominal CECT wi h co onal e o ma ion. a A e ial phase; b po al enous phase. The in ica e ana omical ela ions be ween he s enosis, he duodenal loop and essels a e appa en bu i s cha ac e iza ion is somewha insu icien . Close con ac be ween he igh hepa ic a e y and s enosis is depic ed Page 5 o 9 Pe ei adaSil ae al. BMC Med Imaging (2020) 20:120 Fig. 5 3D models on Meshmixe (STL o ma ). The segmen ed issues a e isualized in a on al iew (abo e) and a bo om iew (below). On he igh , he li e was excluded o easie isualiza ion o he ana omy o he in e nal s uc u es. Colou legend: anspa en —li e ; da k blue—hepa ic eins and in e io ena ca a; ligh blue—po al ein; ed—ao a and hepa ic a e y; g een—bile duc s Fig. 6 Measu emen s o he hepa ic a e y ( op ow) and main po al ein (bo om ow) in bo h CT and STL Page 6 o 9 Pe ei adaSil ae al. BMC Med Imaging (2020) 20:120 The pa ien had an in e es ing ana omical a ia ion, an accesso y hepa ic ein appa en ly d aining S6, which was also segmen ed and included in ou model. This howe e did no p esen any clinical consequences. The ensuing 3D models we e discussed and analyzed wi h he su gical eam and used in he planning o he su ge y, emphasizing i s use ulness o he mapping o ascula ela ions wi h he s enosis and bilia y ee (Fig.7). Du ing su ge y, a e dissec ion o adhesions, he hepa icoduodenal is ula was iden i ied and he duo- denal o i ice closed. Thanks o he p eope a i e 3D econs uc ions, he igh hepa ic a e y was ound and only minimally dissec ed away om he bilia y con e - gence, hus a oiding de ascula iza ion o he bile duc s. The hila pla e was lowe ed, and he bilia y con e gence was dissec ed in a densely in lamed hepa ic hilum, wi h he usual ana omical landma ks obscu ed (Fig.8). A e cholangioscopic explo a ion con i med all main bilia y duc s we e pa en , a Roux-en-Y hepa icojejunos omy was pe o med. Gi en he pa ien ’s medical his o y o ulce a i e coli is a hepa ic biopsy was pe o med o exclude p ima y scle osing cholangi is (PSC). Table 1 Di e ences inmeasu emen inMR, CT andSTL N/A no applicable Landma ks Measu emen s (mm) Di e ences (%) inmeasu emen s (STL/MRI) Di e ences (%) inmeasu emen s (STL/CT) MRI CT STL Righ hepa ic duc 5.9 5.1 5.7 3.448 11.111 Le hepa ic duc 7.9 7.1 7.9 0 10.667 Main po al ein N/A 16.2 16.8 N/A 3.636 Hepa ic a e y N/A 4.8 4.9 N/A 2.062 Righ hepa ic a e y N/A 4.5 4.6 N/A 2.198 Le hepa ic a e y N/A 4.8 4.9 N/A 2.062 Di e ences in measu emen s using CT as e e ence (%) 5.2 Di e ences in measu emen s using MR as e e ence o bile duc s and CT o o he s uc u es (%) 2.2 Fig. 7 3D models on Meshmixe (STL o ma ). A de ailed iew o he bilia y ee on he le and wi h he a e ial ela ions on he igh . On he bo om, a zoomed caudal iew o he close ela ionship o he bilia y con e gence and he igh hepa ic a e y. The o he ana omical s uc u es we e hidden o be e depic he in ica e ana omy. All issues can be o a ed 360º. Colou legend: ed—ao a and hepa ic a e y; g een—bile duc s Page 7 o 9 Pe ei adaSil ae al. BMC Med Imaging (2020) 20:120 The pa ien made an une en ul eco e y and was dis- cha ged i e days a e su ge y. On pa hology, he li e biopsy was nega i e o PSC. Th ee mon hs a e su ge y he pa ien is well and symp om ee, wi h no mal li e biochemis y. Discussion andconclusion Mos benign bilia y s ic u es a e ia ogenic, ollow- ing cholecys ec omy o li e ansplan a ion. The main symp oms include jaundice, p u i us, and da kened u ine, weigh loss, e e , nausea and omi ing. Compli- ca ions such as ascending cholangi is, abscess o ma ion and sepsis can occu [10]. Benign bilia y s ic u es may be loca ed acco ding o he Bismu h classi ica ion (Fig.9). Type I s ic u es a e loca ed > 2cm dis al o he con luence o igh and le hepa ic duc s; ype II s ic u es a e loca ed < 2cm om he con luence; ype III s ic u es in ol e he con luence bu he ceiling o he bilia y con luence is in ac ; ype IV s ic u es cause in e up ion a he con luence; ype V co espond o a lesion o an abe an igh hepa ic duc alone o in associa ion wi h inju y o he common hepa ic duc [11]. Bile duc inju y esul ing in s enosis can be ea ed by i e a i e pe cu aneous anshepa ic dila ion, wi h las ing pa ency o he bilia y duc . Biodeg adable s en s may also p o e o be an in e es ing ea men op ion o benign bilia y s ic u es, po en ially o e ing be e echnical esul s and quali y o li e o pa ien s [10, 12]. Endoscopic e og ade dila ion and s en ing is ano he op ion. Howe e , in his pa icula case, due o he young age o he pa ien and he p esence o a long s enosis and Fig. 8 Ope a i e ield a e he comple ion o he dissec ion, demons a ing he bilia y con e gence (whi e a ow). Su ounding issue is densely in lamed, making dissec ion di icul and po en ially haza dous. In his case, he 3D econs uc ion p o ided he ope a i e eam wi h an accu a e ep esen a ion o pa hology and ana omical ela ionships Fig. 9 The Bismu h classi ica ion o bilia y s ic u es—3D models Page 8 o 9 Pe ei adaSil ae al. BMC Med Imaging (2020) 20:120 associa ed is ula, a su gical app oach was p e e ed a e mul idisciplina y discussion. The p esence o he is ula may explain he la e p esen a ion a e cholecys ec omy, and he p e ious episodes o cholangi is, since i would allow an in e mi en d ainage o bile o he duodenum. As usual in ia ogenic bile duc inju ies, associa ed ascu- la inju y should be suspec ed and can also be esponsi- ble o la e s enosis, due o ischemic cholangi is [13]. When p epa ing o a complex bilia y econs uc ion, he su gical eam is dependen upon op imal p eope a- i e imaging. This should p o ide he ascula and bilia y ana omy, he ela ionship o hese s uc u es, as well as de ec po en ial equen ana omical a ia ions. How- e e , in eg a ing he ascula and bilia y ana omy is di - icul since bo h a e bes depic ed in wo dis inc exams, CT and MR. This complex case beau i ully illus a es he way in which in es men in 3D models o su gical plan- ning p o ed o be essen ial. The su geons epo ed ha he addi ional planning eased he su gical app oach and imp o ed con idence in his delica e p ocedu e, due o he de ailed p io knowledge o he pa ien ’s ana omy, especially in he close ela ions be ween he bilia y ac and he ascula s uc u es a he le el o a densely in lamed hepa ic hilum. In Bismu h ype III inju ies he main bile duc is usually only supplied by he hila com- ponen o he epicoledochal a e ial plexus, since he usual 3 and 9 o’clock ma ginal a e ies a e in e up ed [13]. This makes he a e ial supply o he bile duc s ex emely agile in hese ci cums ances. Since success- ul su gical econs uc ion wi h a Roux-en-Y hepa icoje- junos omy elies on a well ascula ized bile duc s ump, excessi e dissec ion be ween he bilia y con e gence and he hepa ic a e y b anches should be a oided. By e iewing he 3D model he su gical eam managed o keep dissec ion a he li e hilum o a minimum o sa ely pe o m he anas omosis. This is pa icula ly ele an since ana omy was highly dis o ed by dense adhesions and local in lamma ion. Ou 3D model eplica es ana omical s uc u es and pa hology wi h high accu acy be ween STL, CT and MRI wi hou signi ican disc epancy be ween measu emen s o he selec ed ana omical landma ks. A ecen sys emic e iew by Pe ica e al. [7] highligh ed he limi ed s udies in ol ing quan i a i e assessmen o he accu acy o 3D p in ed li e models and indica es di e ences be ween he 3D p in ed li e model and o iginal CT da a o be ween 0.20% and 20.8%. Ou measu emen s demon- s a ed high accu acy, wi h only 2.2% a ia ion, al hough we mus ake in conside a ion ha ou model was no p in ed, a ac ha migh ha e con ibu ed o he lowe disc epancy epo ed. The design o 3D models and 3D p in ing ha e been de eloped o di e en medical pu poses, s eng hened by he g ow h o imaging echniques and so wa e o he acquisi ion, p ocessing and segmen a ion o he image in Radiology, especially applied o CT and MRI [14]. The cons uc ion o 3D models ha combine ascu- la and bilia y ana omy, using di e en imaging ech- niques, espec i ely CT and MRI, will p edic ably con ibu e o a mo e igo ous planning o complex li e su ge ies, such as hepa ic esec ion, li ing-dono ansplan a ion o , as in he p esen case, bilia y econ- s uc ion su ge y. In hepa obilia y su ge y, ana omical a ia ions a e he no m ins ead o he excep ion, and su gical mishaps can ha e se ious consequences. The e is scien i ic e idence ha suppo s he bene i s o 3D models in clinical p ac ice, wi h educed su gi- cal ime, and subsequen ly anes he ic ime; educ ion in he numbe and se e i y o complica ions, educed hospi aliza ion ime and, in u n, lowe cos s wi h each pa ien ; imp o emen in he p ecision o su ge ies by be e p eope a i e planning, esul ing in be e ou - come; echnical imp o emen h ough he possibili y o simula ion wi h app op ia e models and also imp o ed communica ion be ween heal hca e p o essionals and pa ien s and amily membe s [14–19]. Resea ch on 3D gene a ed models in bilia y pa hol- ogy is s ill limi ed, wi h only a ew cases epo ed in he li e a u e and no la ge-scale s udies o accu a ely assess i s ue clinical alue. Howe e , ou 3D model was con- side ed o be o g ea alue o su gical planning. Ou 3D model eplica es ana omical s uc u es and pa hology wi h high accu acy, wi h only 2.2% a ia ion be ween STL,CT and MRI measu emen s. A 3D p in ed model could no be ob ained as we do no ye possess he adequa e p in ing equipmen in ou ins i u ion. Howe e , i is he au ho s opinion ha he added alue o a p in ed model would no be signi i- can , in his case. Fu he mo e, he de elopmen and a ailabili y o new i ual eali y echnologies in he nea u u e may limi he need o p in ed models, in some ins ances. We belie e ha 3D models o he bilia y ac can become aluable ools in daily su gical p ac ice. In he cu en medical landscape, whe e a “one size i s all” app oach is inc easingly ou da ed, hese ools can help ake in o conside a ion pa ien -speci ic needs and p o- ide a much mo e pe sonalized app oach o hepa obil- ia y in e en ions. Abb e ia ions 3D: Th ee-dimensional; ALP: Alkaline phospha ase; ALT: Alanine amino ans- e ase; AST: Aspa a e amino ans e ase; CT: Compu ed omog aphy; GGT : Gamma-glu amyl ans e ase; MRCP: Magne ic esonance cholangiopan- c ea og aphy; MRI: Magne ic esonance imaging; PSC: P ima y scle osing cholangi is; STL: S anda d essella ion language. Page 9 o 9 Pe ei adaSil ae al. BMC Med Imaging (2020) 20:120 • as , con enien online submission • ho ough pee e iew by expe ienced esea che s in you ield • apid publica ion on accep ance • suppo o esea ch da a, including la ge and complex da a ypes • gold Open Access which os e s wide collabo a ion and inc eased ci a ions maximum isibili y o you esea ch: o e 100M websi e iews pe yea • A BMC, esea ch is always in p og ess. Lea n mo e biomedcen al.com/submissions Ready o submi you esea ch Ready o submi you esea ch ? Choose BMC and bene i om: ? Choose BMC and bene i om: Acknowledgemen s The au ho s would like o acknowledge “Expe 3D—Comp ehensi e T ans- e sal P og amme in Medical Image Pos P ocessing: 3D P in sing and A i icial In elligence” and hank co-di ec o s D . Josep Munue a and A nau Valls o hei excellen eaching cou se. Au ho s’ con ibu ions NPS: p ojec de elopmen , da a collec ion and managemen , manusc ip w i ing, image segmen a ion and pos -p ocessing. IA: p ojec de elopmen , da a collec ion and managemen , manusc ip w i ing, image segmen a ion and pos -p ocessing. MS: clinical and su gical managemen . LF: clinical and su gical managemen . HA: clinical and su gical managemen , manusc ip edi ing, p ojec de elopmen . PD: manusc ip edi ing, p ojec de elopmen . All au ho s ead and app o ed he inal manusc ip . Funding This publica ion was unded by Cen o Hospi ala e Uni e si á io de Coimb a. A ailabili y o da a and ma e ials The da ase s used and/o analysed du ing he cu en s udy a e a ailable om he co esponding au ho on easonable eques . E hics app o al and consen o pa icipa e No applicable. Consen o publica ion W i en in o med consen was ob ained om he pa ien o publica ion o his case epo and any accompanying images. A copy o he w i en consen is a ailable o e iew by he Edi o o his jou nal. Compe ing in e es s The au ho s decla e ha hey ha e no compe ing in e es s. Au ho de ails 1 Medical Imaging Depa men , Coimb a Uni e si y Hospi al Cen e , P ace a P o . Mo a Pin o, 3000-075 Coimb a, Po ugal. 2 Depa men o Su ge y, Coimb a Uni e si y Hospi al Cen e , P ace a P o . Mo a Pin o, 3000-075 Coim- b a, Po ugal. 3 Facul y o Medicine, Uni e si y o Coimb a, Rua La ga, 3004-504 Coimb a, Po ugal. 4 Uni e si y o Coimb a, Coimb a Ins i u e o Clinical and Biomedical Resea ch (iCBR), Rua La ga, 3004-504 Coimb a, Po ugal. 5 Uni e si y o Coimb a, Cen e o Inno a i e Biomedicine and Bio- echnology (CIBB), Rua La ga, 3004-504 Coimb a, Po ugal. Recei ed: 30 June 2020 Accep ed: 13 Oc obe 2020 Re e ences 1. Mi sou as D, Liacou as P, Imanzadeh A, Giannopoulos A, e al. Medical 3D p in ing o he adiologis . RadioG aphics. 2015;35:1965–88. 2. Baimakhano Z, Soyama A, Taka suki M, Hidaka M, e al. P eope a i e simula ion wi h a 3-dimensional p in ed solid model o one-s op econs uc ion o mul iple hepa ic eins du ing li ing dono li e ans- plan a ion. Li e T anspl. 2015;21:266–8. 3. Pe ica E, Sun Z. Pa ien -speci ic h ee-dimensional p in ing o p e- su gical planning in hepa ocellula ca cinoma ea men . Quan Imaging Med Su g. 2017;7:668–77. 4. Zein NN, Hanouneh IA, Bishop PD, Samaan M, e al. Th ee-dimensional p in o a li e o p eope a i e planning in li ing dono li e ansplan a- ion. Li e T anspl. 2013;19:1304–10. 5. Wi owski J, Wake N, G ochowska A, Sun Z, e al. In es iga ing accu acy o 3D p in ed li e models wi h compu ed omog aphy. Quan Imaging Med Su g. 2019;9:43–52. 6. Souzaki R, Kinoshi a Y, Iei i S, Hayashida M, e al. Th ee-dimensional li e model based on p eope a i e CT images as a ool o assis in su gical planning o hepa oblas oma in a child. Pedia Su g In . 2015;31:593–6. 7. Pe ica ER, Sun Z. A sys ema ic e iew o h ee-dimensional p in ing in li e disease. J Digi Imaging. 2018;31:692–701. 8. Ja an R, Zeman M. A p o o ype educa ional model o hepa obilia y in e en ions: un eiling he ole o g aphic designe s in medical 3D p in - ing. J Digi Imaging. 2018;31:133–43. 9. Allan A, Kealley C, Squelch A, Wong YH, e al. Pa ien -speci ic 3D p in ed model o bilia y duc s wi h congeni al cys . Quan Imaging Med Su g. 2019;9(1):86–93. 10. Kapoo BS, Mau i G, Lo enz JM. Managemen o bilia y s ic u es: s a e-o - he-a e iew. Radiology. 2018;289(3):590–603. 11. Bismu h H, Majno PE. Bilia y s ic u es: classi ica ion based on he p inci- ples o su gical ea men . Wo ld J Su g. 2001;25(10):1241–4. 12. Mau i G, C iado E. Pe cu aneous managemen o benign bilia y s ic u es: is i ime o ocus on educing p ocedu e in asi eness? J Vasc In e Radiol. 2016;27(6):934–6. 13. S asbe g SM, Hel on WS. An analy ical e iew o asculobilia y inju y in lapa oscopic and open cholecys ec omy. HPB (Ox o d). 2011;13(1):1–14. 14. Ma sumo o JS, Mo is JM, Foley TA, e al. Th ee-dimensional physical modeling: applica ions and expe ience a Mayo clinic. Radiog aphics. 2015;35:1989–2006. 15. Ven ola CL. Medical applica ions o 3D p in ing: cu en and p ojec ed uses. P&T. 2014;39:704–11. 16. Wilson CA, A hu s OJ, Black AE, e al. P in ed h ee-dimensional ai way model assis s planning o single-lung en ila ion in a small child. B J Anaes h. 2015;115:616–20. 17. Rehde R, Abd-El-Ba M, Hoo en K, e al. The ole o simula ion in neu o- su ge y. Childs Ne Sys . 2015;32:43–54. 18. Jones DB, Sung R, Weinbe g C, e al. Th ee-dimensional modeling may imp o e su gical educa ion and clinical p ac ice. Su g Inno . 2015;23:189–95. 19. Be nha d JC, Iso ani S, Ma sugasumi T, e al. Pe sonalized 3D p in ed model o kidney and umo ana omy: a use ul ool o pa ien educa ion. Wo ld J U ol. 2015;34:337–45. Publishe ’s No e Sp inge Na u e emains neu al wi h ega d o ju isdic ional claims in pub- lished maps and ins i u ional a ilia ions.