Deep learning with synthetic diffusion MRI data for free-water elimination in glioblastoma cases
|
|
- Justin Nash
- 5 years ago
- Views:
Transcription
1 Deep learning with synthetic diffusion MRI data for free-water elimination in glioblastoma cases Miguel Molina-Romero,2[ ], Benedikt Wiestler 3[ ], Pedro A. Gómez,2[ ], Marion I. Menzel 2[ ], and Bjoern H. Menze [ ] Computer Science, Technischen Universität München, Munich, Germany 2 GE Healthcare Global Research Organization, Munich, Germany 3 Department of Neuroradiology, Klinikum rechts der Isar der Technischen Universität München, Munich, Germany. Abstract. Glioblastoma is the most common and aggressive brain tumor. In clinical practice, diffusion MRI (dmri) enables tumor infiltration assessment, tumor recurrence prognosis, and identification of whitematter tracks close to the resection volume. However, the vasogenic edema (free-water) surrounding the tumor causes partial volume contamination, which induces a bias in the estimates of the diffusion properties and limits the clinical utility of dmri. We introduce a voxel-based deep learning method to map and correct free-water partial volume contamination in dmri. Our model learns from synthetically generated data a non-parametric forward model that maps free-water partial volume contamination to volume fractions. This is independent of the diffusion protocol and can be used retrospectively. We show its benefits in glioblastoma cases: first, a gain of statistical power; second, quantification of free-water and tissue volume fractions; and third, correction of free-water contaminated diffusion metrics. Freewater elimination yields more relevant information from the available data. Keywords: Glioblastoma, Brain Tumor, DTI, Deep learning, Fractional Anisotropy, Free-water elimination, Data harmonization. Introduction Glioblastomas are the most common primary brain tumor. Ninety percent of these are IDH-wild-type and have a dismal prognosis, with a 5-year survival rate of less than %. Diffusion magnetic resonance imaging (dmri) and the diffusion tensor model (DTI) are used clinically for surgical planing. DTI yields quantitative estimates of the tissue diffusivity, e.g. mean diffusivity (MD), and fractional anisotropy (), an index of tissue microstructure organization. Previous research focused on as indicator of tumor grade, tumor cellularity, tumor infiltration and edema assessment [], and tumor recurrence [2]. However, results are controversial due to reproducibility issues derived from differences in methology, image acquisition, or post-processing [].
2 II Data harmonization in dmri is attracting attention to overcome these problems [3]. One way to reduce uncontrolled variability is to eliminate the free-water signal [4]. Free-water elimination (FWE) uses a two-compartments tissue model composed by tissue (or parencma) and free-water [5]. Fitting the diffusion tensor in a two-compartments model is an ill-posed problem that has been solved using spatial regularization [6] or optimized acquisition protocols [7]. Harmonization of image resolution and diffusion directionality can be achieved by image quality transfer (IQT) with non-linear learninig algorithms, such as convolutional neural networks [8] or random forest [9]. IQT offers a new dimension in dmri, enabling learning complex diffusion model on high quality data, to then transfer information captured in signal patterns to low quality data. However, this approach is limited by the availability of rich datasets. We propose an new method for free-water elimination based on an artificial neural network (ANN), trained with synthetically generated data, that is independent of the number of diffusion shells (b-values) and can be applied retrospectively to any dmri data. Instead of regularizing the FWE ill-posed inverse problem, we teach a non-parametric forward model to learn the mapping between partial volume contamination and free-water volume fraction from synthetic data. Besides, unlike IQT, our approach works only in the diffusion dimension, enabling an important simplification of the ANN model. We further show the advantages of FWE in glioblastoma cases: ) a gain of statistical power through data harmonization, 2) complementary information of the tissue microstructure composition, and 3) better assessment of edema, tumor, and tumor infiltrated areas. The source code can be found in 2 Methods Diffusion signal modeling: Following previous work on free-water elimination we modeled the diffusion signal of a single voxel, along the diffusion directions (b, g), as the contribution of tissue and free-water compartments: S(T E, b, g) = S ( ˆf t e T E T 2t ) T E S t (b, g) + ˆf T fw e 2fw Sfw (b, g), () where b and g summarize the gradient effects; S is a scaling factor proportional to the proton density; ˆf t, ˆf fw, T 2t, and T 2fw are the volume fraction and T 2 values of tissue and free-water respectively. Since T 2t < T 2fw, measurements at different echo-times (TE) yield distinct contributions of tissue and free-water. Thus, disentangling the volume fractions from the T 2 effects requires measurements at least two different TEs []. However, in clinical routine only one TE is acquired simplifying Eq.: S(b, g) = S (f t S t (b, g) + f fw S fw (b, g)), (2) where the T 2i and ˆf i effects are integrated in f i, inducing a positive bias towards the new free-water volume fraction (f fw ) as TE increases. The volume
3 <latexit sha_base64="oqbiqlm8watplguqtbznti4ee/u=">aaab9hicbvdlsgnbeoynrxhfuy9ebomqqcjgbpuw8kk3ik4jjguyncwmq2yfzprgwpl/8ojbxasf482/czlsqrmlgoqqbrq7vfgkjbb9bewwlldw/lrhy3nre2d4u7eg44bjdihmppkclylkdgquvbkrtgnp8oy3ujr4jsfxwkthpy5i7gafwpfmipgerwreyftiwdpb3zcwu6xzffskcgiqwakbbnqnejxuxuxjoahmkmblxtgn2ukhrm8nghnwgeuzagpd4ynkqb246vxpmjozsjx6ktivipurvizqgwo8cz3qgfpt63pui/3mtbplnxvhncap2wyrnicezleqlpccyzyzahlsphbcettrrmaoaomhor8y4veoacvuzbsltjksjdwdwcgwowjnu4brq4aadbc/wcm/wk/vivvsfs9aclc3swx9ynz8rzhe</latexit> <latexit sha_base64="oqbiqlm8watplguqtbznti4ee/u=">aaab9hicbvdlsgnbeoynrxhfuy9ebomqqcjgbpuw8kk3ik4jjguyncwmq2yfzprgwpl/8ojbxasf482/czlsqrmlgoqqbrq7vfgkjbb9bewwlldw/lrhy3nre2d4u7eg44bjdihmppkclylkdgquvbkrtgnp8oy3ujr4jsfxwkthpy5i7gafwpfmipgerwreyftiwdpb3zcwu6xzffskcgiqwakbbnqnejxuxuxjoahmkmblxtgn2ukhrm8nghnwgeuzagpd4ynkqb246vxpmjozsjx6ktivipurvizqgwo8cz3qgfpt63pui/3mtbplnxvhncap2wyrnicezleqlpccyzyzahlsphbcettrrmaoaomhor8y4veoacvuzbsltjksjdwdwcgwowjnu4brq4aadbc/wcm/wk/vivvsfs9aclc3swx9ynz8rzhe</latexit> <latexit sha_base64="oqbiqlm8watplguqtbznti4ee/u=">aaab9hicbvdlsgnbeoynrxhfuy9ebomqqcjgbpuw8kk3ik4jjguyncwmq2yfzprgwpl/8ojbxasf482/czlsqrmlgoqqbrq7vfgkjbb9bewwlldw/lrhy3nre2d4u7eg44bjdihmppkclylkdgquvbkrtgnp8oy3ujr4jsfxwkthpy5i7gafwpfmipgerwreyftiwdpb3zcwu6xzffskcgiqwakbbnqnejxuxuxjoahmkmblxtgn2ukhrm8nghnwgeuzagpd4ynkqb246vxpmjozsjx6ktivipurvizqgwo8cz3qgfpt63pui/3mtbplnxvhncap2wyrnicezleqlpccyzyzahlsphbcettrrmaoaomhor8y4veoacvuzbsltjksjdwdwcgwowjnu4brq4aadbc/wcm/wk/vivvsfs9aclc3swx9ynz8rzhe</latexit> <latexit sha_base64="cukpqo6o7jpdhjxjvcrpawgevo=">aaab8nicbvbns8naej3urq/qh69lbzbeeoqgnorenfbbwmltsib7azdutme3ylqqv+gfw8qxv3vw3btsctpxbwoo9gwbmhakubl332ymtrk6tb5q3kvbo7t7f2dr5nkmngpjtlrnzaaloxihgquvjnqtunq8ny4upn67seujujua45thsroeqkgeur+t6kmbss9zcc9aot+7oqjzjoyakndqvb/8fskymctkkhrtbbgpbjnvkjjkk4qfgz5snqid3rvuubsqygc3t8ijvfoksrqthwsm/p7iawzmoa5tzxxaba9qfif8wugpyodimuwlzrvemcszkggdpc8zyrellglhbyvssdvlagoq2baaiy8ve++8fl37y9qzbsijticwtgcqgmuoqm3aipgktwdk/w5mtoi/pufmxbs4xcwh/4hz+ajblko4=</latexit> <latexit sha_base64="ci6cegujdqx/gqmxvxrl96ermq=">aaab6nicbvbns8naej3urq/qh69lbbrulfug8fl3qrygyhdwwz3brld5owoxfk6f/w4khfq7/im//gtzudtj4yelw3w8y8ijhcoot+o6wvbxjfjmzwt7z3evun/waojum+6xwma6eddpyi4hwil7ysauxvi3g7gn7nffulaidh6wencfuwhkqgfo5hl4wkf+9waw3dnimukuzaafgjq+9qcxsnkkhrtbbgj+hnvkjjkovntyhbeyhvgtprbu3fja7dupordigyaxtruhm6u+jjcpjjiqwnyriycx6ufifxvpizesup8ojnf4wpjbit/heyejozlbnlknpc3kryigrkmztssefl9ejt55/bru3l/umndfgmu4gmm4gwzcqhnuoquembjbm7zcm6ocf+fd+zi3lpxi5hd+wpn8asxmjdw=</latexit> <latexit sha_base64="ci6cegujdqx/gqmxvxrl96ermq=">aaab6nicbvbns8naej3urq/qh69lbbrulfug8fl3qrygyhdwwz3brld5owoxfk6f/w4khfq7/im//gtzudtj4yelw3w8y8ijhcoot+o6wvbxjfjmzwt7z3evun/waojum+6xwma6eddpyi4hwil7ysauxvi3g7gn7nffulaidh6wencfuwhkqgfo5hl4wkf+9waw3dnimukuzaafgjq+9qcxsnkkhrtbbgj+hnvkjjkovntyhbeyhvgtprbu3fja7dupordigyaxtruhm6u+jjcpjjiqwnyriycx6ufifxvpizesup8ojnf4wpjbit/heyejozlbnlknpc3kryigrkmztssefl9ejt55/bru3l/umndfgmu4gmm4gwzcqhnuoquembjbm7zcm6ocf+fd+zi3lpxi5hd+wpn8asxmjdw=</latexit> <latexit sha_base64="ci6cegujdqx/gqmxvxrl96ermq=">aaab6nicbvbns8naej3urq/qh69lbbrulfug8fl3qrygyhdwwz3brld5owoxfk6f/w4khfq7/im//gtzudtj4yelw3w8y8ijhcoot+o6wvbxjfjmzwt7z3evun/waojum+6xwma6eddpyi4hwil7ysauxvi3g7gn7nffulaidh6wencfuwhkqgfo5hl4wkf+9waw3dnimukuzaafgjq+9qcxsnkkhrtbbgj+hnvkjjkovntyhbeyhvgtprbu3fja7dupordigyaxtruhm6u+jjcpjjiqwnyriycx6ufifxvpizesup8ojnf4wpjbit/heyejozlbnlknpc3kryigrkmztssefl9ejt55/bru3l/umndfgmu4gmm4gwzcqhnuoquembjbm7zcm6ocf+fd+zi3lpxi5hd+wpn8asxmjdw=</latexit> <latexit sha_base64="hu/uvp4kssc8ziyybylx5knguk=">aaab+xicbvbns8naen3urq/uj6wcbsmpcoqt4evvfyttcfstpn26ead3ulifkpxjyoepwfeppfugznyha4/3zpiz58wcswvz3zhzxvtfao4wdra3tndm8v7jzjkbawbrjwsby9i4cwewzhforliihhoeunr6d+awrcsih8ujmynidq+yzspswxln8x/vouyogat87cvn/nllmxapzm+blus9jbevouuzxtxfrjibquu6k7nstwdkpeyprdlmpmiicr2spnqdukaklnp2f4wcs97edcv6jwtp9kzjaykng6c6aqifc9kbif4nuf6lk7iwthsedl7itzhwez7mghtmafv8ogmhgulbmrqqajsazvcpxfl5ejfva7qll355xgbz5ger2iirfdxsbguggnzgnkbqjz/sk3own48v4nz7mrqujnzlaf2b8/ga4yjoo</latexit> <latexit sha_base64="hu/uvp4kssc8ziyybylx5knguk=">aaab+xicbvbns8naen3urq/uj6wcbsmpcoqt4evvfyttcfstpn26ead3ulifkpxjyoepwfeppfugznyha4/3zpiz58wcswvz3zhzxvtfao4wdra3tndm8v7jzjkbawbrjwsby9i4cwewzhforliihhoeunr6d+awrcsih8ujmynidq+yzspswxln8x/vouyogat87cvn/nllmxapzm+blus9jbevouuzxtxfrjibquu6k7nstwdkpeyprdlmpmiicr2spnqdukaklnp2f4wcs97edcv6jwtp9kzjaykng6c6aqifc9kbif4nuf6lk7iwthsedl7itzhwez7mghtmafv8ogmhgulbmrqqajsazvcpxfl5ejfva7qll355xgbz5ger2iirfdxsbguggnzgnkbqjz/sk3own48v4nz7mrqujnzlaf2b8/ga4yjoo</latexit> <latexit sha_base64="hu/uvp4kssc8ziyybylx5knguk=">aaab+xicbvbns8naen3urq/uj6wcbsmpcoqt4evvfyttcfstpn26ead3ulifkpxjyoepwfeppfugznyha4/3zpiz58wcswvz3zhzxvtfao4wdra3tndm8v7jzjkbawbrjwsby9i4cwewzhforliihhoeunr6d+awrcsih8ujmynidq+yzspswxln8x/vouyogat87cvn/nllmxapzm+blus9jbevouuzxtxfrjibquu6k7nstwdkpeyprdlmpmiicr2spnqdukaklnp2f4wcs97edcv6jwtp9kzjaykng6c6aqifc9kbif4nuf6lk7iwthsedl7itzhwez7mghtmafv8ogmhgulbmrqqajsazvcpxfl5ejfva7qll355xgbz5ger2iirfdxsbguggnzgnkbqjz/sk3own48v4nz7mrqujnzlaf2b8/ga4yjoo</latexit> s s s >t >t >t et et et al al al /< /< /< =w =w =w dj dj dj MX MX MX SA SA SA 8n 8n 8n Pw Pw Pw +D +D +D h5 h5 h5 pl pl pl 3i 3i 3i Z+ Z+ Z+ df df df +F +F +F co co co 6m 6m 6m Cz Cz Cz 7M 7M 7M Bj Bj Bj BM BM BM eu eu eu Qo Qo Qo un un un hq hq hq cz cz cz wg wg wg 4M 4M 4M mg mg mg 4U 4U 4U mg mg mg Fd Fd Fd nm nm nm U/ U/ U/ l3 l3 l3 ur ur ur b/ b/ b/ tj tj tj e9 e9 e9 lf lf lf E E E Ss Ss Ss TZ TZ TZ M M M Kr Kr Kr Gi Gi Gi Yr Yr Yr k3 k3 k3 CP CP CP NK NK NK LN LN LN Bl Bl Bl zo zo zo ye ye ye H/ H/ H/ Ti Ti Ti BJ BJ BJ pw pw pw 4F 4F 4F Nj Nj Nj o8 o8 o8 pu pu pu SE SE SE zi zi zi Pv Pv Pv x x x fi fi fi fu fu fu 6x 6x 6x Cy Cy Cy ir ir ir Yn Yn Yn wq wq wq ij ij ij jp jp jp Cj Cj Cj J+ J+ J+ u6 u6 u6 mh mh mh UR UR UR tx tx tx ay ay ay gi gi gi Dr Dr Dr Op Op Op Ud Ud Ud 7a 7a 7a jf jf jf 3U 3U 3U BR BR BR pt pt pt Gv Gv Gv Hy Hy Hy Eb Eb Eb TN TN TN vo vo vo kj kj kj JK JK JK Vn Vn Vn h+ h+ h+ Jg Jg Jg bb bb bb Tr Tr Tr hk hk hk kn kn kn cq cq cq q q q j j j GF GF GF aa aa aa ZU ZU ZU ku ku ku MI MI MI nd nd nd 3W 3W 3W aw aw aw fk fk fk W4 W4 W4 Lh Lh Lh 5o 5o 5o Fg Fg Fg Qk Qk Qk HW HW HW Uf Uf Uf Cn Cn Cn Ew Ew Ew 6h 6h 6h Di Di Di al al al uf uf uf fn fn fn 7N 7N 7N G7 G7 G7 g3 g3 g3 IV IV IV xu xu xu as as as y7 y7 y7 li li li wh wh wh 4i 4i 4i Yp Yp Yp DD DD DD E E E 6a 6a 6a MW MW MW x6 x6 x6 +M +M +M UJ UJ UJ Oa Oa Oa nu nu nu ve ve ve 3Z 3Z 3Z 7t 7t 7t WZ WZ WZ mj mj mj fj fj fj X X X b b b VW VW VW 6O 6O 6O +t +t +t Oo Oo Oo CH CH CH JI JI JI 8y 8y 8y 8w 8w 8w 3w 3w 3w Le Le Le Y4 Y4 Y4 jt jt jt Du Du Du ZT ZT ZT G/ G/ G/ /m /m /m I/ I/ I/ 7q 7q 7q FH FH FH k4 k4 k4 F6 F6 F6 KF KF KF xo xo xo wo wo wo 5d 5d 5d LR LR LR b3 b3 b3 zw zw zw WD WD WD GY GY GY rq rq rq 3L 3L 3L F8 F8 F8 GU GU GU Fl Fl Fl U U U Rb Rb Rb BL BL BL hq hq hq /q /q /q r r r U3 U3 U3 AN AN AN 8S 8S 8S NB NB NB Vb Vb Vb ci ci ci n6 n6 n6 BA BA BA AA AA AA >" >" >" =Q =Q =Q mr mr mr e6 e6 e lr lr lr XV XV XV XM XM XM QG QG QG /x /x /x Qd Qd Qd Ju Ju Ju ge ge ge C6 C6 C6 ic ic ic "= "= "= es es es ab ab ab _ ah ah ah ti ti ti xe xe xe ta ta ta l< l< l< <latexit sha_base64="/kbhdtgqqoehupbed5hkjhk6za=">aaab93icbvbns8naej34wethqx69lbbbuleua9cwyvekhhbaepybdftshuxolhv4slx5uvppxvplv3ly5aoudgcd7m8zmcipnnr2t7wvlk6tl7akg9ube9uqrt79zrjfomus2si2ghvxiqyuyhq8naqoicyvvb8gritx64ikj73cuci+i/vieglekl+tdffexjpqz8phmbnqzw7bk9bfoltkbouaprvr24vyvney2ssatx7bs9ncoutpjxuztpnli2ph3emtsmzpuxtw8fkyoj9eiykfmxkqn6eyknkdajkdcdecwbnvcm4n9ej8pw3mtfngbiyzzbfgasyeimkzceujyhhblcmrlmvsigvfggjquyccgzf3mrucfi7p9epr3brploaadueyhdidblxdexgkmezvmkb9ws9wo/wx6xyspm9ueprm8ft5gssa==</latexit> <latexit sha_base64="/kbhdtgqqoehupbed5hkjhk6za=">aaab93icbvbns8naej34wethqx69lbbbuleua9cwyvekhhbaepybdftshuxolhv4slx5uvppxvplv3ly5aoudgcd7m8zmcipnnr2t7wvlk6tl7akg9ube9uqrt79zrjfomus2si2ghvxiqyuyhq8naqoicyvvb8gritx64ikj73cuci+i/vieglekl+tdffexjpqz8phmbnqzw7bk9bfoltkbouaprvr24vyvney2ssatx7bs9ncoutpjxuztpnli2ph3emtsmzpuxtw8fkyoj9eiykfmxkqn6eyknkdajkdcdecwbnvcm4n9ej8pw3mtfngbiyzzbfgasyeimkzceujyhhblcmrlmvsigvfggjquyccgzf3mrucfi7p9epr3brploaadueyhdidblxdexgkmezvmkb9ws9wo/wx6xyspm9ueprm8ft5gssa==</latexit> <latexit sha_base64="/kbhdtgqqoehupbed5hkjhk6za=">aaab93icbvbns8naej34wethqx69lbbbuleua9cwyvekhhbaepybdftshuxolhv4slx5uvppxvplv3ly5aoudgcd7m8zmcipnnr2t7wvlk6tl7akg9ube9uqrt79zrjfomus2si2ghvxiqyuyhq8naqoicyvvb8gritx64ikj73cuci+i/vieglekl+tdffexjpqz8phmbnqzw7bk9bfoltkbouaprvr24vyvney2ssatx7bs9ncoutpjxuztpnli2ph3emtsmzpuxtw8fkyoj9eiykfmxkqn6eyknkdajkdcdecwbnvcm4n9ej8pw3mtfngbiyzzbfgasyeimkzceujyhhblcmrlmvsigvfggjquyccgzf3mrucfi7p9epr3brploaadueyhdidblxdexgkmezvmkb9ws9wo/wx6xyspm9ueprm8ft5gssa==</latexit> <latexit sha_base64="penpjuyelbdic/vn5lohllrqyku=">aaab6xicbvbns8naej3urq/qh69lbbbuleug8fl3qragyhdwwz3brln5uwoxfk6e/w4khfq//im//gbzudtj4yelw3w8y8mjxcoot+o6wvbxjfjmzwt7z3evun/wajjmm+6zrca6hvldpvdcr4gstpnarxk3gpho/9cseyl6whhkg5goligeo2il+6ihvwrnrbszkgxifaqgbzq96le3n7as5gqzpmzpdffikcabzn8uulmhqeujeiadyxvnoymygentsijvfoksrqthwsm/p7iawzmoa5tzxxaba9qfif8kwugxyodimuwlzrvemcszk+jfpc8zyrellglhbyvssdvlanop2bc8xzexix9wv6q7d+exm2rrhmo4bhowymlamannmehbgn4hld4c6tz4rw7h/pwklpmhmifoj8/xs2nqw==</latexit> <latexit sha_base64="penpjuyelbdic/vn5lohllrqyku=">aaab6xicbvbns8naej3urq/qh69lbbbuleug8fl3qragyhdwwz3brln5uwoxfk6e/w4khfq//im//gbzudtj4yelw3w8y8mjxcoot+o6wvbxjfjmzwt7z3evun/wajjmm+6zrca6hvldpvdcr4gstpnarxk3gpho/9cseyl6whhkg5goligeo2il+6ihvwrnrbszkgxifaqgbzq96le3n7as5gqzpmzpdffikcabzn8uulmhqeujeiadyxvnoymygentsijvfoksrqthwsm/p7iawzmoa5tzxxaba9qfif8kwugxyodimuwlzrvemcszk+jfpc8zyrellglhbyvssdvlanop2bc8xzexix9wv6q7d+exm2rrhmo4bhowymlamannmehbgn4hld4c6tz4rw7h/pwklpmhmifoj8/xs2nqw==</latexit> <latexit sha_base64="penpjuyelbdic/vn5lohllrqyku=">aaab6xicbvbns8naej3urq/qh69lbbbuleug8fl3qragyhdwwz3brln5uwoxfk6e/w4khfq//im//gbzudtj4yelw3w8y8mjxcoot+o6wvbxjfjmzwt7z3evun/wajjmm+6zrca6hvldpvdcr4gstpnarxk3gpho/9cseyl6whhkg5goligeo2il+6ihvwrnrbszkgxifaqgbzq96le3n7as5gqzpmzpdffikcabzn8uulmhqeujeiadyxvnoymygentsijvfoksrqthwsm/p7iawzmoa5tzxxaba9qfif8kwugxyodimuwlzrvemcszk+jfpc8zyrellglhbyvssdvlanop2bc8xzexix9wv6q7d+exm2rrhmo4bhowymlamannmehbgn4hld4c6tz4rw7h/pwklpmhmifoj8/xs2nqw==</latexit> III a ANN architecture.9 Estimates and ground truth correlation (testing data) c.7 S(b,~g) <latexit sha_base64="wazrdy4u8w68k2z7hakvupvyc=">aaab8nicbvbns8naej3urq/qh69bitqquoqgnorenfbrwmltsib7ardutme3u2hhp4nlx5uvpprvplv3ly5aoudgcd7m8zmcxlolhacb6uwsrq2vlhclgt7+zulfcpnlscsooujxks2wfryjlavzpnszijfhasrumb6z+a4rssvg86ngcfkt6gowmemk76eanhkjpfl/ctotv5yam4o9too5qucozrf85fvimkyonovequ7dsbsfeakz5tgpeanchnah6wphueeivh42u3linxilz4exncwpvn/t2qkumocbayzinqgfr2p+j/xsxv45wdmjklgqeelwptboranadg9jpfqpjaeumnmrtydeemonjgvtajxzexixteu6459xevxl2erhgo4biquidlamatnmefcgk8wyu8wanyrbh/pwgpxphmifwj8/mhwq9w==</latexit> <latexit sha_base64="wazrdy4u8w68k2z7hakvupvyc=">aaab8nicbvbns8naej3urq/qh69bitqquoqgnorenfbrwmltsib7ardutme3u2hhp4nlx5uvpprvplv3ly5aoudgcd7m8zmcxlolhacb6uwsrq2vlhclgt7+zulfcpnlscsooujxks2wfryjlavzpnszijfhasrumb6z+a4rssvg86ngcfkt6gowmemk76eanhkjpfl/ctotv5yam4o9too5qucozrf85fvimkyonovequ7dsbsfeakz5tgpeanchnah6wphueeivh42u3linxilz4exncwpvn/t2qkumocbayzinqgfr2p+j/xsxv45wdmjklgqeelwptboranadg9jpfqpjaeumnmrtydeemonjgvtajxzexixteu6459xevxl2erhgo4biquidlamatnmefcgk8wyu8wanyrbh/pwgpxphmifwj8/mhwq9w==</latexit> <latexit sha_base64="wazrdy4u8w68k2z7hakvupvyc=">aaab8nicbvbns8naej3urq/qh69bitqquoqgnorenfbrwmltsib7ardutme3u2hhp4nlx5uvpprvplv3ly5aoudgcd7m8zmcxlolhacb6uwsrq2vlhclgt7+zulfcpnlscsooujxks2wfryjlavzpnszijfhasrumb6z+a4rssvg86ngcfkt6gowmemk76eanhkjpfl/ctotv5yam4o9too5qucozrf85fvimkyonovequ7dsbsfeakz5tgpeanchnah6wphueeivh42u3linxilz4exncwpvn/t2qkumocbayzinqgfr2p+j/xsxv45wdmjklgqeelwptboranadg9jpfqpjaeumnmrtydeemonjgvtajxzexixteu6459xevxl2erhgo4biquidlamatnmefcgk8wyu8wanyrbh/pwgpxphmifwj8/mhwq9w==</latexit> Nb Nb/2 Nb/4 f t f t.5.3. =.9923 b S(b,~g) t S(b,~g)fw f t S(b,~g) <latexit sha_base64="wazrdy4u8w68k2z7hakvupvyc=">aaab8nicbvbns8naej3urq/qh69bitqquoqgnorenfbrwmltsib7ardutme3u2hhp4nlx5uvpprvplv3ly5aoudgcd7m8zmcxlolhacb6uwsrq2vlhclgt7+zulfcpnlscsooujxks2wfryjlavzpnszijfhasrumb6z+a4rssvg86ngcfkt6gowmemk76eanhkjpfl/ctotv5yam4o9too5qucozrf85fvimkyonovequ7dsbsfeakz5tgpeanchnah6wphueeivh42u3linxilz4exncwpvn/t2qkumocbayzinqgfr2p+j/xsxv45wdmjklgqeelwptboranadg9jpfqpjaeumnmrtydeemonjgvtajxzexixteu6459xevxl2erhgo4biquidlamatnmefcgk8wyu8wanyrbh/pwgpxphmifwj8/mhwq9w==</latexit> <latexit sha_base64="wazrdy4u8w68k2z7hakvupvyc=">aaab8nicbvbns8naej3urq/qh69bitqquoqgnorenfbrwmltsib7ardutme3u2hhp4nlx5uvpprvplv3ly5aoudgcd7m8zmcxlolhacb6uwsrq2vlhclgt7+zulfcpnlscsooujxks2wfryjlavzpnszijfhasrumb6z+a4rssvg86ngcfkt6gowmemk76eanhkjpfl/ctotv5yam4o9too5qucozrf85fvimkyonovequ7dsbsfeakz5tgpeanchnah6wphueeivh42u3linxilz4exncwpvn/t2qkumocbayzinqgfr2p+j/xsxv45wdmjklgqeelwptboranadg9jpfqpjaeumnmrtydeemonjgvtajxzexixteu6459xevxl2erhgo4biquidlamatnmefcgk8wyu8wanyrbh/pwgpxphmifwj8/mhwq9w==</latexit> <latexit sha_base64="wazrdy4u8w68k2z7hakvupvyc=">aaab8nicbvbns8naej3urq/qh69bitqquoqgnorenfbrwmltsib7ardutme3u2hhp4nlx5uvpprvplv3ly5aoudgcd7m8zmcxlolhacb6uwsrq2vlhclgt7+zulfcpnlscsooujxks2wfryjlavzpnszijfhasrumb6z+a4rssvg86ngcfkt6gowmemk76eanhkjpfl/ctotv5yam4o9too5qucozrf85fvimkyonovequ7dsbsfeakz5tgpeanchnah6wphueeivh42u3linxilz4exncwpvn/t2qkumocbayzinqgfr2p+j/xsxv45wdmjklgqeelwptboranadg9jpfqpjaeumnmrtydeemonjgvtajxzexixteu6459xevxl2erhgo4biquidlamatnmefcgk8wyu8wanyrbh/pwgpxphmifwj8/mhwq9w==</latexit> f t+ ffw = <latexit sha_base64="cukpqo6o7jpdhjxjvcrpawgevo=">aaab8nicbvbns8naej3urq/qh69lbzbeeoqgnorenfbbwmltsib7azdutme3ylqqv+gfw8qxv3vw3btsctpxbwoo9gwbmhakubl332ymtrk6tb5q3kvbo7t7f2dr5nkmngpjtlrnzaaloxihgquvjnqtunq8ny4upn67seujujua45thsroeqkgeur+t6kmbss9zcc9aot+7oqjzjoyakndqvb/8fskymctkkhrtbbgpbjnvkjjkk4qfgz5snqid3rvuubsqygc3t8ijvfoksrqthwsm/p7iawzmoa5tzxxaba9qfif8wugpyodimuwlzrvemcszkggdpc8zyrellglhbyvssdvlagoq2baaiy8ve++8fl37y9qzbsijticwtgcqgmuoqm3aipgktwdk/w5mtoi/pufmxbs4xcwh/4hz+ajblko4=</latexit> Fig.. Artificial neural network model. The ANN architecture (a). Training data was synthetically generated following Eq. 2 (b). Free-water partial volume contamination effects were visible (b, red arrows). We ran a correlation analysis for 3 randomly generated samples (c) reaching a factor of.99. fraction indexes are ratios relative to the signal contribution of each compartment and thus f fw + f t =. Synthetic signal generation: The diffusion properties of free-water at body temperature are well characterized [5], presenting isotropic behavior and a diffusion coefficient D fw = 3 3 mm 2 /s, thus S fw (b, g) = e bd fw. On the other hand, S t is unknown since it depends on the tissue microstructure organization and the orientation of the diffusion gradients, g. Thus, we modeled its behavior with a random variable, S t R N b, following an uniform distribution, U(, ), where N b is the number of diffusion measures, including non-diffusion-weighted volumes. Furthermore, we also represented the tissue volume fraction of a voxel, f t [, ], as a random uniform variable, U(, ). Based on the prior knowledge of D fw, the models of S t and f t, and knowing the diffusion protocol, it is possible to generate unlimited synthetic diffusion signals, S(b, g), containing free-water partial volume effects (Eq. 2 and Fig. b). ANN architecture and training: We designed a regression fully connected ANN capable of estimating the tissue volume fraction, f t, directly from the diffusion signal S(b, g) (Eq. 2). The input layer contained as many units as the number of acquired diffusion measures (including non-diffusion-weighted volumes), N b, and a single output unit yielding the estimate of f t. We compared ANN architectures containing from one to five hidden layers, finding an optimum in performance for two hidden layers with N b /2 and N b /4 respectively (Fig. a). To train the ANN we used 2 synthetic signals generated as explained above (7% training, 5% validation, and 5% testing). Convergence for N b = 33 (Fig. ) was reached after nine epochs and 4.7 seconds in a consumers laptop (Apple
4 IV 8 a Fractional Anisotropy 4 Mean Diffusivity 2 Standard DTI, shell b 6 c Standard DTI, 2 shells 5 6 ANN, shell.5 ANN, 2 shells Pasternak et al., shell 4 Hoy et al., 2 shells Tissue volume fraction ANN, shell ANN, 2 shells Pasternak et al., shell Hoy et al., 2 shells.5.5 MD [mm 2 /s] -3 Fig. 2. Comparison of FWE with ANN and state of the art methods for two diffusion protocols. The correction effects on of ANN for one shell were equivalent to Pasternak s et al. algorithm (a), while ANN MD estimates were less over-regularized (b). Tissue volume fraction estimates were also in agreement although Pasternak s method failed to estimate f t < (c). The ANN for two shells and Hoy s et al. algorithm were in good agreement for (a) but larger differences were visible for MD and f t (b and c). See the supplementary material for more information. MacBook Pro, Intel Core i5, 8GB RAM; MATLAB, Mathworks, Natwick, MA). The training process depended on the diffusion protocol prescribed. Thus, we trained four networks to match the DWI data used in the experiments below. 3 Experiments and Results Comparison with state of the art: Methods from Pasternak et al. [6] and Hoy et al. [7] are the state of the art for one and two shell acquisitions. For comparison we measured data from a volunteer in a GE 3T MR75w (GE Healthcare, Milwaukee, WI). The protocol comprised first, one diffusion weighted imaging (DWI) acquisition for 3 diffusion direction (b=5 s/mm 2 ) and two b= volume (N b =32); and second a DWI for two shells (b=5 and s/mm 2 ) with 3 diffusion directions for each shell and four non-diffusion-weighted volumes (N b =64). The data was processed with a pipeline including steps for: ) head motion and eddy current corrections (FSL eddy); 2) denoising based on random matr theory []; and 3) free-water elimination based on ANN, Pasternak s, and Hoy s methods (Fig. 2). The ANN results were comparable to those of the state of the art methods. Data harmonization: The T 2 effects described in Eq. and Eq. 2 were investigated using data from a volunteer scanned in the same scanner as before. The protocol consisted of a DWI acquisition for 3 direction (b= s/mm 2 ) and one non-diffusion-weighted volume (N b = 3). This was repeated for seven equispaced TE = ms. The data was processed as described before, but only ANN FWE was computed. For comparison two processing lines were created with and without ANN based FWE. Both were fitted with robust DTI (RESTORE) [2] to extract diffusion metrics. The multi-echo diffusion data showed an increase of free-water (Fig. 3c) and its effects (Fig. 3a, larger low peak) with the TE, which agrees with the
5 V a Standard DTI Free-water eliminated DTI b TE = 74.9 ms TE 2 = 84.9 ms TE 3 = 94.9 ms TE 4 = 4.9 ms TE 5 = 4.9 ms TE 6 = 24.9 ms TE 7 = 34.9 ms c Free-water volume fraction Fig. 3. Histogram comparison at several TEs. from standard DTI (a) showed larger variability across TEs for <, than the free-water corrected (b). The tissue volume fraction estimates are influenced by T 2 decay (Eq. ) and thus sensitive to TE (c). The free-water elimination step has a TE harmonization effect on the diffusion signal (b), shifting the TE variability into the volume fraction estimate (c). two-compartments tissue model (Eq. ) accounting for T 2 effects, simplified in Eq. 2. Multi-center studies are often carried out in data acquired with heterogeneous protocols. The prescribed TEs are a source of variability that mostly depend on the gradient strength of the scanner and the image resolution. Data harmonization is important to remove uncontrolled variability and achieve a good statistical power. Free-water elimination plays a double role. First, it accounts for T 2 effects in the f i, shifting this variability from the diffusion metrics to the volume fraction estimates (Fig. 3). And second, it removes the diffusion isotropic noise from the signal showing the actual tissue anisotropy, eliminating the variability induced by the presence of free-water. Glioblastoma analysis: Data from 25 patients affected by glioblastoma (IDH wild-type, WHO 26 classification) were provided by the Department of Neuroradiology, Klinikum rechts der Isar der Technischen Universität München, Munich, Germany. All patients are part of a prospective glioma database, approved by the local ethics committee, and gave written informed consent. They were scanned in a 3T whole-body scanner (Achieva, Philips Medical Systems, Best, The Netherlands). The protocol included DWI for 32 directions (b=8 s/mm 2 ) and one non-diffusion-weighted volume (N b = 33). Furthermore, T 2 turbo spin echo (T2w), T 2 -FLAIR, and T contrast enhanced (CE-Tw) were acquired. The DWI data was processed as described before, and T2w, CE-Tw, and FLAIR volumes were registered to the DWI space (Fig. 4). Tissue and free-water volume fraction estimates: The free-water and tissue maps computed with ANN (Fig. 4d and h) showed tumor and edema areas that were in agreement with well established methods: T2w, CE-Tw, and FLAIR (Fig. 4e, f, and g). The estimation of tissue and free-water volume fraction maps is an important feature of FWE. They provide complementary information of the underlying tissue organization. Cytotoxic edema and necrosis areas that are not distinguishable can be better identified with the knowledge of the amount of tissue in the voxel (Fig. 4d, e, f, g, and h, light-blue arrows). Furthermore, in clinical routine tumor delineation is based on CE-Tw per-intensities (Fig. 4f
6 VI Fig. 4. Comparison of metrics and contrasts for one patient. Fractional anisotropy from standard DTI (a) showed a dimmed region corresponding to edema (white arrows). This area was recovered after free-water elimination (b and c). We observed alterations of the white matter integrity compared to the NAWM (b, white asterisks). Free-water (d) and tissue (h) volume fraction maps were estimated by the ANN. Edema regions were well defined and distinguishable from the tumor (red contour). Pools compatible with cytotoxic edema were observable inside the tumor (light-blue arrows). These findings agreed with the observations based on T2w (e), CE-Tw (f), and FLAIR (g). Extended tumor infiltration was derived from the comparison of CE-Tw and tissue volume fraction map (h, green arrow and contour). red contour). When this is compared with tissue volume fraction maps (Fig. 4h zoomed area), we observed an increased tumor region of up to four millimeters that is compatible with tumor infiltration. This agrees with the radiated area after resection (Fig. 4h, green arrow and contour). Fractional anisotropy recovery: The comparison between standard and FWE maps (Fig. 4a and b) exhibited a recovery of the anisotropic information in the edema region and around the ventricles (Fig. 4c, white arrows). The elimination of the isotropic compartment from the diffusion signal leads to a recovery of the tissue anisotropy captured by the protocol, leading to an enhancement of the maps, especially in areas with large partial volume contamination like edema and the border of the ventricles (Fig. 4a, b, and c, white arrows). The correction of the maps provides new information of the tissue microstructure integrity hidden by the edema (Fig. 4b asterisks). Edema and infiltration unming: To assess the impact of FWE in edema, tumor, and tumor infiltrated areas, regions of interests (ROI) were defined by a neuroradiologist for each subject using FLAIR, T2w, CE-Tw, ft, and FWE
7 VII a Standard DTI comparison P=2.6e-4 P=8.7e-4 P=6.4e-6 ROI NAWM b FWE DTI comparison P=3.6e-2 P=6.e-5 P=6.4e-5 c Tissue volume fraction Tissue volume fraction comparison P=.2e- P=5.6e-2 P=7.4e- Edema Tumour Infiltration Edema Tumour Infiltration Edema Tumour Infiltration Fig. 5. Edema, tumor, and tumor infiltrated regions differentiation. The values for standard DTI (a) were statistically different (α =.) between NAWM areas. After free-water elimination (b) values in edema and tumor infiltrated areas were significantly larger than those for tumor (family-wise error =.3), indicating a better organized tissue microstructure. Complementary, the tissue volume fractions in tumor and tumor infiltrated regions were statistically larger than in edema areas, suggesting that the loss in in edema was mostly due to free-water infiltration. maps (Fig. 4). To minimize the influence of outliers we used the median for each ROI, and ran t-test comparison for and tissue volume fraction values across subjects for the three type of ROIs (Fig. 5). For reference, extra ROIs were drawn in normal appearing white mater (NAWM) mostly contralateral. Comparison of NAWM with tumor and edema areas for the 25 patients showed a statistically significant difference (α =.) in for standard DTI (Fig. 5a). Tumor infiltration and edema are the driving factors behind the loss in. After free-water elimination we compared FWE (Fig. 5b) and tissue volume fraction (Fig. 5c) in edema, tumor and tumor infiltrated areas (family-wise error =.3). Significantly larger was found in edema and tumor infiltrated areas compared to tumor regions. While, tissue volume fraction in edema was statistically lower than in tumor infiltrated and tumor sections. We pothesize that the combination of FWE and tissue volume fraction yields a better understanding of tissue microstructure integrity (Fig. 5). High and low tissue volume fraction might indicate well organized microstructure infiltrated by free-water (vasogenic edema). Low and high tissue volume fraction is compatible with the unstructured cellularity found in tumors. Finally, areas with high and tissue volume fraction might be caused by tumor infiltration in highly structured white matter bundles. Discussion The presented ANN model combines simplicity with high accuracy. We introduce for the first time an ANN design capable of learning partial volume effect features from synthetic data, and reach comparable results with the state of the art methods but at least 55-fold faster. Besides, our approach, unlike [6] is voxel-based, avoiding blurring artifacts induced by the use of patch based
8 VIII regularization, and can be applied to any diffusion protocol beyond DTI. The robustness of our approach was shown with 25 patients, 2 volunteers, 4 protocols, and 2 scanners from different manufacturers. We attribute these results to the good understanding of the psical signal model mimicked by the ANN. The addition of free-water elimination in processing pipeline extracts more information from the data, which has a potential benefit for glioblastoma patients in three aspects: diagnosis, surgical planning, and guided radiotherapy. The diagnosis potentially improves with the quantification of the free-water volume fraction map, yielding the severity of the vasogenic edema. The precision of the surgical planning benefits from the corrected maps in edema areas providing a better definition of white matter fiber bundles and limiting the resection of healt tissue. Finally, guided radiotherapy profits from identification of tumor infiltrated areas. Acknowledgments The authors want to thank Dr. Ofer Pasternak for his support in the comparison of the methods. This work was supported by the TUM Institute of Advanced Study, funded by the German Excellence Initiative, and the European Commission (Grant Agreement Number 6562). References. A.S. Field. Diffusion imaging in brain tumors. In D.K. Jones, editor, Diffusion MRI, chapter 33, pages Oxford University Press, S. Bette et al. Local Fractional Anisotropy Is Reduced in Areas with Tumor Recurrence in Glioblastoma. Radiology, 283(2):499 57, Z. Eaton-Rosen et al. Beyond the Resolution Limit: Diffusion Parameter Estimation in Partial Volume. MICCAI, pages 65 62, C. Metzler-Baddeley et al. How and how not to correct for CSF-contamination in diffusion MRI. Neuroimage, 59(2):394 43, C. Pierpaoli and D.K. Jones. Removing CSF Contamination in Brain DT-MRIs by Using a Two-Compartment Tensor Model. In ISMRM, Kyoto, page 25, O. Pasternak et al. Free Water Elimination and Mapping from Diffusion MRI. Magn. Reson. Med., 73:77 73, A.R. Hoy et al. Optimization of a Free Water Elimination Two-Compartment Model for Diffusion Tensor Imaging. Neuroimage, (3): , R. Tanno et al. Bayesian image quality transfer with CNNs: Exploring uncertainty in dmri super-resolution. volume 433 LNCS, pages 6 69, D.C. Alexander et al. Image quality transfer and applications in diffusion MRI. Neuroimage, 52(March): , 27.. M. Molina-Romero et al. A diffusion model-free framework with echo time dependence for free-water elimination and brain tissue microstructure characterization. Magn. Reson. Med., mar 28.. J. Veraart et al. Denoising of diffusion MRI using random matr theory. Neuroimage, 42:394 46, nov L. Chang et al. Restore: Robust estimation of tensors by outlier rejection. Magnetic Resonance in Medicine, 53(5):88 95, 25.
In vivo diffusion tensor imaging (DTI) of articular cartilage as a biomarker for osteoarthritis
In vivo diffusion tensor imaging (DTI) of articular cartilage as a biomarker for osteoarthritis Jose G. Raya 1, Annie Horng 2, Olaf Dietrich 2, Svetlana Krasnokutsky 3, Luis S. Beltran 1, Maximilian F.
More informationPatterns of Brain Tumor Recurrence Predicted From DTI Tractography
Patterns of Brain Tumor Recurrence Predicted From DTI Tractography Anitha Priya Krishnan 1, Isaac Asher 2, Dave Fuller 2, Delphine Davis 3, Paul Okunieff 2, Walter O Dell 1,2 Department of Biomedical Engineering
More informationDiffusion Restriction Precedes Contrast Enhancement in Glioblastoma Multiforme
Diffusion Restriction Precedes Contrast Enhancement in Glioblastoma Multiforme Adil Bata 1, Jai Shankar 2 1 Faculty of Medicine, Class of 2017 2 Department of Diagnostic Radiology, Division of Neuroradiology,
More informationVisualization strategies for major white matter tracts identified by diffusion tensor imaging for intraoperative use
International Congress Series 1281 (2005) 793 797 www.ics-elsevier.com Visualization strategies for major white matter tracts identified by diffusion tensor imaging for intraoperative use Ch. Nimsky a,b,
More informationPOC Brain Tumor Segmentation. vlife Use Case
Brain Tumor Segmentation vlife Use Case 1 Automatic Brain Tumor Segmentation using CNN Background Brain tumor segmentation seeks to separate healthy tissue from tumorous regions such as the advancing tumor,
More informationDiffusion Tensor Imaging in brain tumours
Diffusion Tensor Imaging in brain tumours @MarionSmits, MD PhD Associate Professor of Neuroradiology Dept. of Radiology, Erasmus MC, Rotterdam (NL) Honorary Consultant and Reader UCLH National Hospital
More informationDTI fiber tracking at 3T MR using b-1000 value in the depiction of periprostatic nerve before and after nervesparing prostatectomy
DTI fiber tracking at 3T MR using b-1000 value in the depiction of periprostatic nerve before and after nervesparing prostatectomy Poster No.: C-2328 Congress: ECR 2012 Type: Scientific Paper Authors:
More informationGroup-Wise FMRI Activation Detection on Corresponding Cortical Landmarks
Group-Wise FMRI Activation Detection on Corresponding Cortical Landmarks Jinglei Lv 1,2, Dajiang Zhu 2, Xintao Hu 1, Xin Zhang 1,2, Tuo Zhang 1,2, Junwei Han 1, Lei Guo 1,2, and Tianming Liu 2 1 School
More informationRECENT ADVANCES IN CLINICAL MR OF ARTICULAR CARTILAGE
In Practice RECENT ADVANCES IN CLINICAL MR OF ARTICULAR CARTILAGE By Atsuya Watanabe, MD, PhD, Director, Advanced Diagnostic Imaging Center and Associate Professor, Department of Orthopedic Surgery, Teikyo
More informationCover Page. The handle holds various files of this Leiden University dissertation.
Cover Page The handle http://hdl.handle.net/1887/35124 holds various files of this Leiden University dissertation. Author: Wokke, Beatrijs Henriette Aleid Title: Muscle MRI in Duchenne and Becker muscular
More informationClustering of MRI Images of Brain for the Detection of Brain Tumor Using Pixel Density Self Organizing Map (SOM)
IOSR Journal of Computer Engineering (IOSR-JCE) e-issn: 2278-0661,p-ISSN: 2278-8727, Volume 19, Issue 6, Ver. I (Nov.- Dec. 2017), PP 56-61 www.iosrjournals.org Clustering of MRI Images of Brain for the
More informationDifferentiating Tumor and Edema in Brain Magnetic Resonance Images Using a Convolutional Neural Network
Original Article Differentiating Tumor and Edema in Brain Magnetic Resonance Images Using a Convolutional Neural Network Aida Allahverdi 1, Siavash Akbarzadeh 1, Alireza Khorrami Moghaddam 2, Armin Allahverdy
More information/13/$ IEEE
Multivariate Discriminant Analysis of Multiparametric Brain MRI to Differentiate High Grade and Low Grade Gliomas - A Computer- Aided Diagnosis Development Study *, Zeynep Firat, Ilhami Kovanlikaya, Ugur
More informationComparative Study of K-means, Gaussian Mixture Model, Fuzzy C-means algorithms for Brain Tumor Segmentation
Comparative Study of K-means, Gaussian Mixture Model, Fuzzy C-means algorithms for Brain Tumor Segmentation U. Baid 1, S. Talbar 2 and S. Talbar 1 1 Department of E&TC Engineering, Shri Guru Gobind Singhji
More informationarxiv: v1 [cs.cv] 13 Jul 2018
Multi-Scale Convolutional-Stack Aggregation for Robust White Matter Hyperintensities Segmentation Hongwei Li 1, Jianguo Zhang 3, Mark Muehlau 2, Jan Kirschke 2, and Bjoern Menze 1 arxiv:1807.05153v1 [cs.cv]
More informationTumor cut segmentation for Blemish Cells Detection in Human Brain Based on Cellular Automata
Tumor cut segmentation for Blemish Cells Detection in Human Brain Based on Cellular Automata D.Mohanapriya 1 Department of Electronics and Communication Engineering, EBET Group of Institutions, Kangayam,
More informationMethods of MR Fat Quantification and their Pros and Cons
Methods of MR Fat Quantification and their Pros and Cons Michael Middleton, MD PhD UCSD Department of Radiology International Workshop on NASH Biomarkers Washington, DC April 29-30, 2016 1 Disclosures
More informationBrain diffusion tensor imaging changes in cerebrotendinous xanthomatosis reversed with
Brain diffusion tensor imaging changes in cerebrotendinous xanthomatosis reversed with treatment Claudia B. Catarino, MD, PhD, 1*, Christian Vollmar, MD, PhD, 2,3* Clemens Küpper, MD, 1,4 Klaus Seelos,
More informationT2, T2*, ute. Yeo Ju Kim. Radiology, Inha University Hospital, Incheon, Korea
SY28-1 T2, T2*, ute Yeo Ju Kim Radiology, Inha University Hospital, Incheon, Korea T2 relaxation times relate to the rate of transverse magnetization decay, caused by the loss of phase coherence induced
More informationTissue-engineered medical products Evaluation of anisotropic structure of articular cartilage using DT (Diffusion Tensor)-MR Imaging
Provläsningsexemplar / Preview TECHNICAL REPORT ISO/TR 16379 First edition 2014-03-01 Tissue-engineered medical products Evaluation of anisotropic structure of articular cartilage using DT (Diffusion Tensor)-MR
More informationMRI Applications in Radiation Oncology:
MRI Applications in Radiation Oncology: Physician s Perspective Jeff Olsen, MD Department of Radiation Oncology Washington University, St. Louis, MO Disclosures Washington University has research and service
More informationRepeatability of 2D FISP MR Fingerprinting in the Brain at 1.5T and 3.0T
Repeatability of 2D FISP MR Fingerprinting in the Brain at 1.5T and 3.0T Guido Buonincontri 1,2, Laura Biagi 1,3, Alessandra Retico 2, Michela Tosetti 1,3, Paolo Cecchi 4, Mirco Cosottini 1,4,5, Pedro
More information3D Deep Learning for Multi-modal Imaging-Guided Survival Time Prediction of Brain Tumor Patients
3D Deep Learning for Multi-modal Imaging-Guided Survival Time Prediction of Brain Tumor Patients Dong Nie 1,2, Han Zhang 1, Ehsan Adeli 1, Luyan Liu 1, and Dinggang Shen 1(B) 1 Department of Radiology
More informationOverview. Fundamentals of functional MRI. Task related versus resting state functional imaging for sensorimotor mapping
Functional MRI and the Sensorimotor System in MS Nancy Sicotte, MD, FAAN Professor and Vice Chair Director, Multiple Sclerosis Program Director, Neurology Residency Program Cedars-Sinai Medical Center
More informationLiver Fat Quantification
Liver Fat Quantification Jie Deng, PhD, DABMP Department of Medical Imaging May 18 th, 2017 Disclosure Research agreement with Siemens Medical Solutions 2 Background Non-alcoholic fatty liver diseases
More informationSpeed, Comfort and Quality with NeuroDrive
Speed, Comfort and Quality with NeuroDrive Echelon Oval provides a broad range of capabilities supporting fast, accurate diagnosis of brain conditions and injuries. From anatomical depiction to vascular
More informationMRI/MRS Biomarkers. Robert E. Lenkinski, Ph.D.
MRI/MRS Biomarkers Robert E. Lenkinski, Ph.D. Disclosure GE Healthcare-Research Grant Aspect MR-Scientific Advisor Aposense-Scientific Advisor Brainwatch-Scientific Advisor I will be discussing off-label
More informationTracking the language pathways in edema patients: Preliminary results.
Tracking the language pathways in edema patients: Preliminary results. Sarah M. E. Gierhan 1,2, Peter Rhone 3,4, Alfred Anwander 1, Isabel Jost 3, Clara Frydrychowicz 3, Karl-Titus Hoffmann 4, Jürgen Meixensberger
More informationClassification and Statistical Analysis of Auditory FMRI Data Using Linear Discriminative Analysis and Quadratic Discriminative Analysis
International Journal of Innovative Research in Computer Science & Technology (IJIRCST) ISSN: 2347-5552, Volume-2, Issue-6, November-2014 Classification and Statistical Analysis of Auditory FMRI Data Using
More informationA deep learning model integrating FCNNs and CRFs for brain tumor segmentation
A deep learning model integrating FCNNs and CRFs for brain tumor segmentation Xiaomei Zhao 1,2, Yihong Wu 1, Guidong Song 3, Zhenye Li 4, Yazhuo Zhang,3,4,5,6, and Yong Fan 7 1 National Laboratory of Pattern
More informationActivated Fibers: Fiber-centered Activation Detection in Task-based FMRI
Activated Fibers: Fiber-centered Activation Detection in Task-based FMRI Jinglei Lv 1, Lei Guo 1, Kaiming Li 1,2, Xintao Hu 1, Dajiang Zhu 2, Junwei Han 1, Tianming Liu 2 1 School of Automation, Northwestern
More informationExperimental Assessment of Infarct Lesion Growth in Mice using Time-Resolved T2* MR Image Sequences
Experimental Assessment of Infarct Lesion Growth in Mice using Time-Resolved T2* MR Image Sequences Nils Daniel Forkert 1, Dennis Säring 1, Andrea Eisenbeis 2, Frank Leypoldt 3, Jens Fiehler 2, Heinz Handels
More informationWHAT DOES THE BRAIN TELL US ABOUT TRUST AND DISTRUST? EVIDENCE FROM A FUNCTIONAL NEUROIMAGING STUDY 1
SPECIAL ISSUE WHAT DOES THE BRAIN TE US ABOUT AND DIS? EVIDENCE FROM A FUNCTIONAL NEUROIMAGING STUDY 1 By: Angelika Dimoka Fox School of Business Temple University 1801 Liacouras Walk Philadelphia, PA
More informationDong Gyu Na, Vincent N. Thijs, Gregory W. Albers, Michael E. Moseley, and Michael P. Marks. AJNR Am J Neuroradiol 25: , September 2004
AJNR Am J Neuroradiol 25:1331 1336, September 2004 Diffusion-Weighted MR Imaging in Acute Ischemia: Value of Apparent Diffusion Coefficient and Signal Intensity Thresholds in Predicting Tissue at Risk
More informationExploring Peritumoral White Matter Fibers for Neurosurgical Planning
Exploring Peritumoral White Matter Fibers for Sonia Pujol, Ph.D. Ron Kikinis, M.D. Surgical Planning Laboratory Harvard University Clinical Goal Diffusion Tensor Imaging (DTI) Tractography has the potential
More informationAutomatic Definition of Planning Target Volume in Computer-Assisted Radiotherapy
Automatic Definition of Planning Target Volume in Computer-Assisted Radiotherapy Angelo Zizzari Department of Cybernetics, School of Systems Engineering The University of Reading, Whiteknights, PO Box
More informationDeclaration of conformity Adapter
Declaration of conformity acc. to VO (EG) 1935/2004 and VO (EU) 10/2011 as well as acc. to FDA Document ID: 34466 Editing status: 2017-05-19 2 CFR FDA stands for Food and Drug Administration, a U.S. authority.
More informationASL BASICS II. Learning Objectives. Outline. Acquisition. M. A. Fernández-Seara, Ph. D. Arterial spin labeled perfusion MRI: basic theory
Acquisition ASL BASICS II M. A. Fernández-Seara, Ph. D. Neuroimaging Laboratory Center for Applied Medical Research University of Navarra Pamplona, Spain Outline Arterial spin labeled perfusion MRI: basic
More informationCorrelation of quantitative proton MR spectroscopy with local histology from stereotactic brain biopsy to evaluate heterogeneity of brain tumors
Correlation of quantitative proton MR spectroscopy with local histology from stereotactic brain biopsy to evaluate heterogeneity of brain tumors Steve H. Fung, MD 1, Edward F. Jackson, PhD 2, Samuel J.
More informationBrain Tumour Detection of MR Image Using Naïve Beyer classifier and Support Vector Machine
International Journal of Scientific Research in Computer Science, Engineering and Information Technology 2018 IJSRCSEIT Volume 3 Issue 3 ISSN : 2456-3307 Brain Tumour Detection of MR Image Using Naïve
More informationPublication for the Philips MRI Community Issue 39 December 2009
FieldStrength Publication for the Philips MRI Community Issue 39 December 2009 32-channel coil boosts 3.0T neuro imaging at Kennedy Krieger Kennedy Krieger Institute sees significantly better fmri, DTI,
More informationConfidence-based Ensemble for GBM brain tumor segmentation
Confidence-based Ensemble for GBM brain tumor segmentation Jing Huo 1, Eva M. van Rikxoort 1, Kazunori Okada 2, Hyun J. Kim 1, Whitney Pope 1, Jonathan Goldin 1, Matthew Brown 1 1 Center for Computer vision
More informationFunctional Magnetic Resonance Imaging with Arterial Spin Labeling: Techniques and Potential Clinical and Research Applications
pissn 2384-1095 eissn 2384-1109 imri 2017;21:91-96 https://doi.org/10.13104/imri.2017.21.2.91 Functional Magnetic Resonance Imaging with Arterial Spin Labeling: Techniques and Potential Clinical and Research
More informationInvestigations in Resting State Connectivity. Overview
Investigations in Resting State Connectivity Scott FMRI Laboratory Overview Introduction Functional connectivity explorations Dynamic change (motor fatigue) Neurological change (Asperger s Disorder, depression)
More informationEarly Diagnosis of Autism Disease by Multi-channel CNNs
Early Diagnosis of Autism Disease by Multi-channel CNNs Guannan Li 1,2, Mingxia Liu 2, Quansen Sun 1(&), Dinggang Shen 2(&), and Li Wang 2(&) 1 School of Computer Science and Engineering, Nanjing University
More informationA quality control program for MR-guided focused ultrasound ablation therapy
JOURNAL OF APPLIED CLINICAL MEDICAL PHYSICS, VOLUME 3, NUMBER 2, SPRING 2002 A quality control program for MR-guided focused ultrasound ablation therapy Tao Wu* and Joel P. Felmlee Department of Radiology,
More informationCSM is the most common cause of spinal cord dysfunction
Published December 22, 2011 as 10.3174/ajnr.A2756 ORIGINAL RESEARCH T. Sato T. Horikoshi A. Watanabe M. Uchida K. Ishigame T. Araki H. Kinouchi Evaluation of Cervical Myelopathy Using Apparent Diffusion
More informationBioMatrix Tuners: CoilShim
MAGNETOM Vida special issue Head and Neck Imaging Clinical 11 BioMatrix Tuners: CoilShim Miriam R. Keil, Ph.D.; Jörg Rothard; Carmel Hayes, Ph.D. Siemens Healthineers, Erlangen, Germany A cervical spine
More informationFindings of DTI-p maps in comparison with T 2 /T 2 -FLAIR to assess postoperative hyper-signal abnormal regions in patients with glioblastoma
Beigi et al. Cancer Imaging (2018) 18:33 https://doi.org/10.1186/s40644-018-0166-4 REGULAR ARTICLE Open Access Findings of DTI-p maps in comparison with T 2 /T 2 -FLAIR to assess postoperative hyper-signal
More informationTwelve right-handed subjects between the ages of 22 and 30 were recruited from the
Supplementary Methods Materials & Methods Subjects Twelve right-handed subjects between the ages of 22 and 30 were recruited from the Dartmouth community. All subjects were native speakers of English,
More informationThe Low Sensitivity of Fluid-Attenuated Inversion-Recovery MR in the Detection of Multiple Sclerosis of the Spinal Cord
The Low Sensitivity of Fluid-Attenuated Inversion-Recovery MR in the Detection of Multiple Sclerosis of the Spinal Cord Mark D. Keiper, Robert I. Grossman, John C. Brunson, and Mitchell D. Schnall PURPOSE:
More informationUNIVERSITY OF CALGARY. A New Method for Assessing Tissue Alignment using Clinical MRI in Multiple Sclerosis. Shrushrita Sharma A THESIS
UNIVERSITY OF CALGARY A New Method for Assessing Tissue Alignment using Clinical MRI in Multiple Sclerosis by Shrushrita Sharma A THESIS SUBMITTED TO THE FACULTY OF GRADUATE STUDIES IN PARTIAL FULFILMENT
More informationPINPOINTING RADIATION THERAPY WITH THE PRECISION OF MR.
GE Healthcare PINPOINTING RADIATION THERAPY WITH THE PRECISION OF MR. MR Radiation Oncology Suite MAXIMIZE YOUR PRECISION. HELP MINIMIZE PATIENT COMPLICATIONS. Our goal in MR radiation oncology is to
More informationSupplemental Information
Supplemental Information Prediction of Prostate Cancer Recurrence using Quantitative Phase Imaging Shamira Sridharan 1, Virgilia Macias 2, Krishnarao Tangella 3, André Kajdacsy-Balla 2 and Gabriel Popescu
More informationImaging and radiotherapy physics topics for project and master thesis
Imaging and radiotherapy physics topics for project and master thesis Supervisors: Assoc. Professor Kathrine Røe Redalen and PhD candidates Franziska Knuth and Kajsa Fridström. Contact: kathrine.redalen@ntnu.no,
More informationConvolutional Neural Networks for Estimating Left Ventricular Volume
Convolutional Neural Networks for Estimating Left Ventricular Volume Ryan Silva Stanford University rdsilva@stanford.edu Maksim Korolev Stanford University mkorolev@stanford.edu Abstract End-systolic and
More informationHigh spatial resolution reveals excellent detail in pediatric neuro imaging
Publication for the Philips MRI Community Issue 46 2012/2 High spatial resolution reveals excellent detail in pediatric neuro imaging Achieva 3.0T with 32-channel SENSE Head coil has become the system
More informationPearls and Pitfalls of MR Diffusion in Clinical Neurology
Pearls and Pitfalls of MR Diffusion in Clinical Neurology Dr. Alberto Bizzi Neuroradiology Unit Fondazione IRCCS Istituto Neurologico Carlo Besta Milan, Italy Email: alberto_bizzi@fastwebnet.it Diffusion
More informationANALYSIS AND DETECTION OF BRAIN TUMOUR USING IMAGE PROCESSING TECHNIQUES
ANALYSIS AND DETECTION OF BRAIN TUMOUR USING IMAGE PROCESSING TECHNIQUES P.V.Rohini 1, Dr.M.Pushparani 2 1 M.Phil Scholar, Department of Computer Science, Mother Teresa women s university, (India) 2 Professor
More informationImpaired Regional Myocardial Function Detection Using the Standard Inter-Segmental Integration SINE Wave Curve On Magnetic Resonance Imaging
Original Article Impaired Regional Myocardial Function Detection Using the Standard Inter-Segmental Integration Ngam-Maung B, RT email : chaothawee@yahoo.com Busakol Ngam-Maung, RT 1 Lertlak Chaothawee,
More informationITERATIVELY TRAINING CLASSIFIERS FOR CIRCULATING TUMOR CELL DETECTION
ITERATIVELY TRAINING CLASSIFIERS FOR CIRCULATING TUMOR CELL DETECTION Yunxiang Mao 1, Zhaozheng Yin 1, Joseph M. Schober 2 1 Missouri University of Science and Technology 2 Southern Illinois University
More informationDeveloping ML Models for semantic segmentation of medical images
Developing ML Models for semantic segmentation of medical images Raj Jena University of Cambridge Microsoft Research Cambridge Disclosures I perform consultancy work for InnerEye team at Microsoft Research
More informationA Sound Track to Reading
A Sound Track to Reading Blending Flashcards Prepared by Donald L Potter June 1, 2018 Mr. Potter prepared these cards to be used with Sister Monica Foltzer s advanced intensive phonics program and reader,
More informationDeclaration of conformity VEGABAR 82
Declaration of conformity acc. to VO (EG) 1935/2004 and VO (EU) 10/2011 as well as acc. to FDA and USP Class VI as well as ADI-free Document ID: 47480 Editing status: 2018-09-11 2 CFR FDA stands for Food
More informationSupplementary Information Methods Subjects The study was comprised of 84 chronic pain patients with either chronic back pain (CBP) or osteoarthritis
Supplementary Information Methods Subjects The study was comprised of 84 chronic pain patients with either chronic back pain (CBP) or osteoarthritis (OA). All subjects provided informed consent to procedures
More informationThe Paul Evans Memorial Lecture Functional radiotherapy targeting using focused dose escalation. Roberto Alonzi Mount Vernon Cancer Centre
The Paul Evans Memorial Lecture Functional radiotherapy targeting using focused dose escalation Roberto Alonzi Mount Vernon Cancer Centre Overview Introduction and rationale for focused dose escalation
More informationA Reliable Method for Brain Tumor Detection Using Cnn Technique
IOSR Journal of Electrical and Electronics Engineering (IOSR-JEEE) e-issn: 2278-1676,p-ISSN: 2320-3331, PP 64-68 www.iosrjournals.org A Reliable Method for Brain Tumor Detection Using Cnn Technique Neethu
More informationMammography is a most effective imaging modality in early breast cancer detection. The radiographs are searched for signs of abnormality by expert
Abstract Methodologies for early detection of breast cancer still remain an open problem in the Research community. Breast cancer continues to be a significant problem in the contemporary world. Nearly
More informationTo test the possible source of the HBV infection outside the study family, we searched the Genbank
Supplementary Discussion The source of hepatitis B virus infection To test the possible source of the HBV infection outside the study family, we searched the Genbank and HBV Database (http://hbvdb.ibcp.fr),
More informationThe risk of MR-detected carotid plaque hemorrhage on recurrent or first-time stroke: a meta-analysis of individual patient data
The risk of MR-detected carotid plaque hemorrhage on recurrent or first-time stroke: a meta-analysis of individual patient data Schindler A 1, Bonati LH 2, Schinner R 1, Altaf N 3, Hosseini AA 3, Esposito-Bauer
More informationMagnetic resonance (MR) diffusion-weighted imaging (DWI) is
Diagn Interv Radiol DOI 10.4261/1305-3825.DIR.3892-10.1 Turkish Society of Radiology 2010 ABDOMINAL IMAGING ORIGINAL ARTICLE Diffusion tensor imaging of the kidney at 3 Tesla: normative values and repeatability
More information2D-Sigmoid Enhancement Prior to Segment MRI Glioma Tumour
2D-Sigmoid Enhancement Prior to Segment MRI Glioma Tumour Pre Image-Processing Setyawan Widyarto, Siti Rafidah Binti Kassim 2,2 Department of Computing, Faculty of Communication, Visual Art and Computing,
More information1) Diffusion weighted imaging DWI is a term used to describe moving molecules due to random thermal motion. This motion is restricted by boundaries
1) Diffusion weighted imaging DWI is a term used to describe moving molecules due to random thermal motion. This motion is restricted by boundaries such as ligaments, membranes and macro molecules. Diffusion
More informationFibre orientation dispersion in the corpus callosum relates to interhemispheric functional connectivity
Fibre orientation dispersion in the corpus callosum relates to interhemispheric functional connectivity ISMRM 2017: http://submissions.mirasmart.com/ismrm2017/viewsubmissionpublic.aspx?sei=8t1bikppq Jeroen
More informationMR Functional Imaging to Guide Radiotherapy: Challenges and Opportunities
Abstract No. 1234 MR Functional Imaging to Guide Radiotherapy: Challenges and Opportunities Michael Milosevic, MD Department of Radiation Oncology, University of Toronto Radiation Medicine Program, Princess
More informationBasics of Quantification of ASL
Basics of Matthias van Osch Contents Thoughts about quantification Some quantification models Bloch equations Buxton model Parameters that need to be estimated Labeling efficiency M Bolus arrival time
More informationThe Evolution of SBRT and Hypofractionation in Thoracic Radiation Oncology
The Evolution of SBRT and Hypofractionation in Thoracic Radiation Oncology (specifically, lung cancer) 2/10/18 Jeffrey Kittel, MD Radiation Oncology, Aurora St. Luke s Medical Center Outline The history
More informationDiffusion tensor imaging of spinal cord as an emerging tool in neuroradiology!!!
Diffusion tensor imaging of spinal cord as an emerging tool in neuroradiology!!! Award: Cum Laude Poster No.: C-1049 Congress: ECR 2016 Type: Educational Exhibit Authors: K. Agawane, M. R. Verma, D. P.
More informationClinically focused workflow with unique ability to integrate fmri, DTI, fiber tracks and perfusion in a single, multi-layered 3D rendering
Clinically focused workflow with unique ability to integrate fmri, DTI, fiber tracks and perfusion in a single, multi-layered 3D rendering Neurosurgeons are demanding more from neuroradiologists and increasingly
More informationarxiv: v1 [stat.ml] 23 Jan 2017
Learning what to look in chest X-rays with a recurrent visual attention model arxiv:1701.06452v1 [stat.ml] 23 Jan 2017 Petros-Pavlos Ypsilantis Department of Biomedical Engineering King s College London
More informationSYSTEM OVERVIEW MINIMALLY INVASIVE ROBOTIC LASER THERMOTHERAPY
SYSTEM OVERVIEW MINIMALLY INVASIVE ROBOTIC LASER THERMOTHERAPY Software NeuroBlate Fusion Software is the exclusive software intelligence behind the NeuroBlate System. NeuroBlate Fusion allows neurosurgeons
More informationA characteristic feature of acute haematomas in the brain on echo-planar diffusion-weighted imaging
Neuroradiology (2002) 44: 907 911 DOI 10.1007/s00234-002-0860-5 DIAGNOSTIC NEURORADIOLOGY N. Morita M. Harada K. Yoneda H. Nishitani M. Uno A characteristic feature of acute haematomas in the brain on
More informationInternational Conference on Biological Sciences and Technology (BST 2016)
International Conference on Biological Sciences and Technology (BST 2016) A Better Characterization of Brain Damage in Carbon Monoxide Intoxication Assessed in Vivo Using Diffusion Kurtosis Imaging Wen-Yao
More informationPDFlib PLOP: PDF Linearization, Optimization, Protection. Page inserted by evaluation version
PDFlib PLOP: PDF Linearization, Optimization, Protection Page inserted by evaluation version www.pdflib.com sales@pdflib.com Principal Diffusion Direction in Peritumoral Fiber Tracts Color Map Patterns
More informationA Comparative Evaluation of Voxel-based Spatial Mapping in Diffusion Tensor Imaging
A Comparative Evaluation of Voel-based Spatial Mapping in Diffusion Tensor Imaging Ryan P. Cabeen 1, Mark E. Bastin 2, David H. Laidlaw 1 1 Department of Computer Science, Brown University, Providence,
More informationMR-Radiomics in Neuro-Oncology
Klinik für Stereotaxie und funktionelle Neurochirurgie Institut für Neurowissenschaften und Medizin MR-Radiomics in Neuro-Oncology M. Kocher Klinik für funktionelle Neurochirurgie und Stereotaxie Forschungszentrum
More informationNeuroimaging. BIE601 Advanced Biological Engineering Dr. Boonserm Kaewkamnerdpong Biological Engineering Program, KMUTT. Human Brain Mapping
11/8/2013 Neuroimaging N i i BIE601 Advanced Biological Engineering Dr. Boonserm Kaewkamnerdpong Biological Engineering Program, KMUTT 2 Human Brain Mapping H Human m n brain br in m mapping ppin can nb
More informationMR Tumor Staging for Treatment Decision in Case of Wilms Tumor
MR Tumor Staging for Treatment Decision in Case of Wilms Tumor G. Schneider, M.D., Ph.D.; P. Fries, M.D. Dept. of Diagnostic and Interventional Radiology, Saarland University Hospital, Homburg/Saar, Germany
More informationarxiv: v1 [cs.cv] 17 Aug 2017
Deep Learning for Medical Image Analysis Mina Rezaei, Haojin Yang, Christoph Meinel Hasso Plattner Institute, Prof.Dr.Helmert-Strae 2-3, 14482 Potsdam, Germany {mina.rezaei,haojin.yang,christoph.meinel}@hpi.de
More informationCSE Introduction to High-Perfomance Deep Learning ImageNet & VGG. Jihyung Kil
CSE 5194.01 - Introduction to High-Perfomance Deep Learning ImageNet & VGG Jihyung Kil ImageNet Classification with Deep Convolutional Neural Networks Alex Krizhevsky, Ilya Sutskever, Geoffrey E. Hinton,
More informationCOMPUTER DISCRIMINATION BETWEEN DISEASES OF THE BRAIN BASED ON MR IMAGE FEATURES
COMPUTER DISCRIMINATION BETWEEN DISEASES OF THE BRAIN BASED ON MR IMAGE FEATURES DU BOULAY G. 1,TEATHER D. 2,TEATHER B. 2, DU BOULAY B. 3, SHARPLES M. 4, JEFFERY N. 2, HINKLEY L. 5, WHITCOMBE E. 3 Institute
More informationRelated Symposia in AAPM 2007
Related Symposia in AAPM 7 Functional and Physiological MR Imaging for Therapy Assessment Yue Cao,, Ph.D. Departments of Radiation Oncology and Radiology, University of Michigan President s s symposium:
More informationARTIFICIAL INTELLIGENCE FOR DIGITAL PATHOLOGY. Kyunghyun Paeng, Co-founder and Research Scientist, Lunit Inc.
ARTIFICIAL INTELLIGENCE FOR DIGITAL PATHOLOGY Kyunghyun Paeng, Co-founder and Research Scientist, Lunit Inc. 1. BACKGROUND: DIGITAL PATHOLOGY 2. APPLICATIONS AGENDA BREAST CANCER PROSTATE CANCER 3. DEMONSTRATIONS
More informationImaging can play an important role in determining benignancy
ORIGINAL RESEARCH A. Srinivasan C.J. Galbán T.D. Johnson T.L. Chenevert B.D. Ross S.K. Mukherji Utility of the K-Means Clustering Algorithm in Differentiating Apparent Diffusion Coefficient Values of Benign
More informationImage Fusion, Contouring, and Margins in SRS
Image Fusion, Contouring, and Margins in SRS Sarah Geneser, Ph.D. Department of Radiation Oncology University of California, San Francisco Overview Review SRS uncertainties due to: image registration contouring
More informationMR Advance Techniques. Vascular Imaging. Class II
MR Advance Techniques Vascular Imaging Class II 1 Vascular Imaging There are several methods that can be used to evaluate the cardiovascular systems with the use of MRI. MRI will aloud to evaluate morphology
More informationarxiv: v2 [cs.cv] 19 Dec 2017
An Ensemble of Deep Convolutional Neural Networks for Alzheimer s Disease Detection and Classification arxiv:1712.01675v2 [cs.cv] 19 Dec 2017 Jyoti Islam Department of Computer Science Georgia State University
More informationBRAIN STATE CHANGE DETECTION VIA FIBER-CENTERED FUNCTIONAL CONNECTIVITY ANALYSIS
BRAIN STATE CHANGE DETECTION VIA FIBER-CENTERED FUNCTIONAL CONNECTIVITY ANALYSIS Chulwoo Lim 1, Xiang Li 1, Kaiming Li 1, 2, Lei Guo 2, Tianming Liu 1 1 Department of Computer Science and Bioimaging Research
More informationOriginal Article Age-related changes of normal prostate: evaluation by MR diffusion tensor imaging
Int J Clin Exp Med 2015;8(7):11220-11224 www.ijcem.com /ISSN:1940-5901/IJCEM0009871 Original Article Age-related changes of normal prostate: evaluation by MR diffusion tensor imaging Ji Zhang 1,2, Wei-Zhong
More informationEffect of intravenous contrast medium administration on prostate diffusion-weighted imaging
Effect of intravenous contrast medium administration on prostate diffusion-weighted imaging Poster No.: C-1766 Congress: ECR 2015 Type: Authors: Keywords: DOI: Scientific Exhibit J. Bae, C. K. Kim, S.
More information