2D and 3D Medical Ima 2D and 3D Medical Ima ages for Anatomy ages - PowerPoint PPT Presentation

2D and 3D Medical Ima 2D and 3D Medical Ima ages for Anatomy ages for Anatomy Education using a cloud d computing platform • Lih ‐ Shyang Chen • Department of Electrical En ngineering • National Cheng Kung Unive N ti l Ch K U i ersity it

Outline Outline • Introduction • Related backgroun • Related backgroun nd nd • System architectur y re • System functions f for anatomy education education • Cloud platform for p r education • Current status of t the system 2017/3/22

troduction troduction uman anatomy is the basic scientific study fo y y or students in medical schools to learn. • Name • Shape • Position • Size • Organ structures • Functionality he traditional way to learn anatomy • Anatomical atlas book • Static 2D images St ti 2D i • Laboratory dissection 2017/3/22

I t Introduction d ti • The difficulty of learning an Th diffi lt f l i natomy t • There is a huge amount o g f information that students should understand and me emorize – more than one thousand of organs. g • The students still can not visualize the internal structures of a given orga structures of a given orga an an. • It is still difficult for the s students to visualize or imagine the spatial relatio imagine the spatial relatio onship between various onship between various organs since the organs m might obscure each other. 2017/3/22

Related background Related background d • The computerized 3D an natomical atlases can be used to create arbitrary used to create arbitrary views of the human views of the human anatomy, provide a “loo k and feel” close to real organs, and support an organs and support an “ “interactive learning by interactive learning by doing” mechanism. • The basic data we use • Anatomical cross ‐ section • Anatomical cross ‐ section nal image datasets nal image datasets • The Visible Human Pro oject (VHP) 2017/3/22

The Visible Human P The Visible Human P Project Project • The project is run by the U Th j i b h U .S. National Library of S N i l Lib f Medicine (NLM) under the e direction of Michael J. Ackerman. Ackerman • It is an effort to create a de etailed data set of cross ‐ sectional photographs of t ti l h t h f t h h he human body b d • A male and a female cadav ver were cut into thin slices which were then photogra hi h h h phed and digitized. h d d di i i d 2017/3/22

Data collect Data collect tion tion • Photographic cross-sections: Volume Dimensions: 1760 x 1 Volume Dimensions: 1760 x 1 1024 x 1878 1024 x 1878 Pixel Dimensions: .33 mm x .3 33 mm x 1 mm Pixel Depth: 24-bit (8-bits x R p ( RGB) Volume Size: 9.5 GB • Computed Tomography (CT) – in frozen state: Volume Dimensions: 512 x 51 12 x 1878 Pixel Dimensions: variable x v variable x 1 mm Pixel Depth: 16 bit Pixel Depth: 16-bit Volume Size: 939 MB

The Visible Human P The Visible Human Project Project 2017/3/22

D organ Reconstruct tion process 2D object segmentation bj i • We have developed a semi ‐ automatic method to automatically generate all y g the potential object edges. • The non ‐ expert persons can The non expert persons can easily and efficiently draw the contour of an organ the contour of an organ without carefully following the boundary of an organ the boundary of an organ 2017/3/22

3D organ Reconstr 3D organ Reconstr ruction process ruction process • Assignment of organ name Assignment of organ name es es • Each contour in an image e should be assigned a correct organ name for t correct organ name for t the students to learn. the students to learn • The same organ in the re est of the images can be easily assigned an organ il i d name automatically. t ti ll • An expert can double ‐ ch heck the correctness of the name assignment. • 3D object reconstruction 3D object reconstruction • The 3D organs are recon structed from a series of 2D images images. 2017/3/22

3D organ Reconstr 3D organ Reconstr ruction process ruction process • Correction of the 2D orga Correction of the 2D orga an segmentation and an segmentation and assignments of organ na mes 2017/3/22

S System functions for t f ti f anatomy education t d ti • 2D and 3D display • 2D cross ‐ sectional images of an orga g g n are useful for the students to learn where they are in the human body a nd the detailed internal structures of the organ. 2017/3/22

S System functions for t f ti f anatomy education t d ti • the 3D images are useful for the stud dents to visualize its shape in a 3D space and its spatial relationships with oth er related organs. 2017/3/22

System functions for anatomy e ducation • The students want to be ab The students want to be ab ble to see the spatial ble to see the spatial relationships between the 2 2D images and 3D images. 2017/3/22

System functions for anatomy e ducation • The students can also click a at a point in a 2D image and the system will show the co the system will show the co orresponding point in the 3D orresponding point in the 3D image. 2017/3/22

The 3D kidney and some 2D cross-sectional images

The 3D trachea and some 2D cross-sectional images

System functions for anatomy e ducation • Learning CT and MRI image es • In general, different orga I l diff t ans have different image h diff t i features in the same typ e of images. • The same organ may hav ve different image features in different types of imag in different types of imag ges such as T1 T2 and ges, such as T1, T2, and proton density MRI imag ges. • Find the corresponding p Fi d th di points in different images. i t i diff t i 2017/3/22

tem functions for anatomy educat tion (corresponding nts) nts) 2017/3/22

System functions for anatomy e ducation • The s stem • The system will show the 3D ill sho the 3D D image of the organ and all D image of the organ and all the 2D image planes that ha ave CT and MRI images in the 3D 3D scene. 2017/3/22

System functions for anatomy e ducation • Manipulation of the 3D object • Move, scale, and rotate the object , , j 2017/3/22

System functions for anatomy e ducation • The user can further use e some tools to manipulate the 3D object interactive the 3D object interactive ely ely. 2017/3/22

Cl Cloud computing plat d ti l t tform for education tf f d ti • This kind of anatomic learning system needs a relatively “big” hard l i l “bi ” h d ware platform system l f • Need at least 8G byt • Need at least 8G byt te memory te memory • Need at least 100G byte disk for storing y g the data • Need a PC level GPU U to display 3D image • It can not be used by a It t b d b a mobile device. bil d i 2017/3/22

oud computing platfor p g p rm for education • A cloud computing platf A l d ti l tf orm with a RemoteApp ith R t A can be used to allow a m mobile device, which is , affordable for most stud dents, to use the system. • Currently, we are workin C l ki ng in this direction. i hi di i • Azure (GPU was not supp Azure (GPU was not supp ported) + RemoteApp (will ported) + RemoteApp (will not be supported) • Azure + ZenApp essentia A + Z A ti als (in the future) l (i th f t ) • Simple demo. p 2017/3/22

rrent status of the system y • Currently, the system has been y, y developed for some initial p trials. • Both the instructors and studen Both the instructors and studen nts have found that the nts have found that the system indeed can provide muc ch more anatomical information for the students to o learn the anatomic structure and related information. • It is much easier for the students and instructors to share the images and disc ss ario s problems sim ltaneo discuss various problems simultaneou usly. sl • Students can interactively man ipulate the organs over and over again until they fully unde over again until they fully unde erstand the anatomical erstand the anatomical structures. • The system is still under develo Th t i till d d l opment. t 2017/3/22