Segment 3DPrint Instructions For Use - English

Transcription

1 Segment 3DPrint Instructions For Use - English December 3, 2018 Software platform v2.2 R6839

2 MEDVISO AB Griffelvägen 3 SE Lund Sweden Tel: ii

3 Contents 1 Terms and conditions Regulatory status Conditions for use Indications for use Intended patient population Applicable medical conditions Safety Instructions Conventions and Abbreviations Typographic conventions Trademarks Abbreviations Getting started System requirements Installation Installing Matlab Compiler Runtime Installing Segment 3DPrint Starting Segment 3DPrint Uninstallation Software overview First time running Segment 3DPrint Setting preferences Setting window positions PACS connection Loading and storing Data - Step by step Loading data iii

4 CONTENTS Loading data from database Loading data from PACS Loading data from network disc Storing data Image viewing tools Viewing options Crop image stack Zoom Adjust contrast Manual adjustment Automatic adjustment Measure distance Hide/Show overlay Resample image stack D Segmentation module D Segmentation - Step by step Segmentation tools (panel a) Object tools (panel k) View modes (panel b) Sort out tool Split tool Smoothing tool Vessel tool Preview / generate STL file Storing segmentation CT Cardiac analysis - Step by step Cardiac reconstruction LV analysis Automatic LV segmentation Manual LV segmentation Erase LV segmentation Validation of LV segmentation RV analysis Manual RV segmentation Erase RV segmentation iv

5 CONTENTS Validation of RV segmentation MR Cardiac analysis - Step by step LV analysis Automatic LV segmentation [1] Manual LV segmentation Erase LV segmentation Copying LV segmentation Validation of LV segmentation RV analysis Automatic RV segmentation [1] Manual RV segmentation Erase RV segmentation Validation of RV segmentation ROI analysis - Step by step Manual ROI analysis Erase ROI segmentation Validation of the ROI analysis Summary of ROI validation v

6

7 1 Terms and conditions This manual contains Instructions for Use, to safely be able to use Segment 3DPrint. 1.1 Regulatory status Segment 3DPrint bears the CE marking of conformity and is certified according to the ISO standard. Users are required to investigate the regulatory requirements pertinent to their country or location prior to using Segment 3DPrint. It is in the users responsibility to obey these statues, rules and regulations. 1.2 Conditions for use The software may only be used by medically trained professionals. The software is a tool to provide relevant clinical data. The clinician is solely responsible for interpretation of clinical data and taking decisions how to manage the patients. The software should be used to read medical images in DICOM format. Images can be supplied either by CD s or by using DICOM standard to transfer files directly from the scanner or PACS systems. The user is responsible for installing Segment 3DPrint. Medviso provide online assistance in the installation process if needed. To install Segment 3DPrint you need to be Administrator on the computer to install on. Users are required to participate in an online training session held by Medviso prior to use Segment 3DPrint. This is provided by Medviso as a part of the software trial process. 1.3 Indications for use Segment 3DPrint is a software for visualization and image segmentation of medical images. The software aims to transfer imaging information from a 1

8 CHAPTER 1. TERMS AND CONDITIONS medical imaging system such as a Computer Tomography (CT) scanner or Magnetic Resonance (MR) scanner to an output file or for visualization. It is intended to generate anatomical 3D models for education or evaluating surgical treatment options. Segment 3DPrint is intended to be used by medically trained professionals for the patient population and medical conditions given in the Instructions for Use for the device. 1.4 Intended patient population The device can be used on the following patient population Age Weight Height Health From children to adult (radiation inappropriateness versus the clinical need is the limiting age factor) No limitations. The lower limitations lay in the resolution of the image and the upper limitations is to fit in the medical scanner No limitations. The lower limitations lay in the resolution of the image and the upper limitations is to fit in the medical scanner Possible contrast agent contraindications and general inappropriateness of radiation or magnetic field dependt on used medical scanner for image acquisition Patient state Patient is not user 1.5 Applicable medical conditions The medical purpose of Segment 3DPrint is to generate patient-specific anatomical models for increased understanding of anatomy to be used for 2 Education Communication Surgical planning / procedures evaluate surgical treatment options

9 1.6. SAFETY INSTRUCTIONS 1.6 Safety Instructions The computer where Segment 3DPrint is installed should have anti-virus protection installed to protect against cyber attacks login to access Segment 3DPrint to ensure that only accredited users use the device backup on patient data to ensure no data is lost In the preferences for Segment 3DPrint the user can activate user login to the device and user logging for the usage of the device. 3

10

11 2 Conventions and Abbreviations This chapter describes the typographic conventions in this manual and in the program. 2.1 Typographic conventions A Key A at the keyboard. Ctrl-A Control key. Hold down Ctrl key and A simultaneously. Icon in toolbar. *.mat Filename extension. C:/Program Folder. File Menu, e.g. File menu. File Save As Sub menu, e.g. under the File menu the item Save As is found. Close Push/Toggle button in the graphical user interface. Endocardium Radiobutton in the graphical user interface. Single frame Checkbox in the graphical user interface. 2.2 Trademarks Below are some of the trademarks used in this manual. Segment 3DPrint is a trademark of Medviso AB. Segment DICOM Server is a trademark of Medviso AB. Sectra PACS is a trademark of Sectra Imtec AB, ( Matlab is a trademark of the Mathworks Inc, ( 2.3 Abbreviations CT Computed Tomography LV Left Ventricle MR Magnetic Resonance ROI Region of Interest RV Right Ventricle 5

12

13 3 Getting started 3.1 System requirements Operating System: 64-bit Windows 10 Computer with 16 GB of memory or more Harddisk with at least 1 GB of available space CUDA enabled graphics card with 4GB memory or more Internet connection for software license management 3-button computer mouse Systems with two screens is recommended Using SSD disk for reading data is recommended 3.2 Installation For first time installation, start with installing Matlab Compiler Runtime (Section 3.2.1) and thereafter installing Segment 3DPrint (Section 3.2.2). For upgrading, only install Segment 3DPrint (Section 3.2.2). To be able to perform the installation you need to have administrator privileges on the machine Installing Matlab Compiler Runtime 1. Download the MCR file and double click it. The file can be downloaded from Medviso AB homepage ( 2. Follow the instructions in Figures Reboot the computer Installing Segment 3DPrint 1. The latest Segment 3DPrint version can be downloaded from Medviso AB homepage ( Download the Software file, double click it and follow the instructions. You need a 7

14 CHAPTER 3. GETTING STARTED Figure 1: Click on Yes. Figure 2: Click on Next. Figure 3: Select Yes and click on Next. 8

15 3.2. INSTALLATION Figure 4: Click on Next. Figure 5: Click on Install. 9

16 CHAPTER 3. GETTING STARTED password to download the software, which you should have received at the time of purchase. If you have lost the password, please contact 2. For new installations you need to add your license code for Segment 3DPrint. Add your license by enter your license code in the installation process. You can also add your license code after installation by starting Segment 3DPrint and select Generate License under the Help menu in Segment 3DPrint. Note that you have to run the software as Administrator to be able to add the license code in Segment 3DPrint. A third way of adding your license is to add a license file (named code.lic) to the same folder as where Segment 3DPrint is installed. If you do not have a license code, please contact sales@medviso.com. 3.3 Starting Segment 3DPrint To start the program, double click the file C:/Program Files/Segment 3DPrint/segment3dp. or your shortcut to it. When starting the software, the image in Figure 6 should be displayed. If it is not displayed, then the software is not correctly installed. 3.4 Uninstallation To uninstall Segment 3DPrint, remove all files in the folder C:/Program/Segment 3DPrint or C:/Program Files/Segment 3DPrint. User preferences are stored in the Application Data and the subfolder Segment 3DPrint under each user account. To uninstall the Matlab Compiler Runtime, use the Windows functionality Install or Remove Programs in the control panel menu. 3.5 Software overview An overview of Segment 3DPrint is given in Figure 7. The letters in the figure will be used as references throughout this manual. To learn more about each tool, hold the mouse over the icon in the software and a help text will be displayed. 10

17 3.6. FIRST TIME RUNNING SEGMENT 3DPRINT Figure 6: Startup window for Segment 3DPrint. 3.6 First time running Segment 3DPrint The first time Segment 3DPrint is started, it runs a setup process which can take a while, so be patient. To complete the setup, set preferences and window positions as described in Sections and Setting preferences It is recommended to set the preferences of which folders to use to avoid browsing each time you want to load or save a file. It is invoked by using the icon (p). Set Data, Export and CD folders Setting window positions The position of the main window for Segment 3DPrint can be set by dragging the window to an optional position and size. The size and position will be saved so that next time Segment 3DPrint is launched the same position will be used. In case where one have switched to another monitor, Segment 3DPrint may move outside the screen. In this case you could press 11

18 CHAPTER 3. GETTING STARTED Figure 7: Main graphical user interface. The letters in the figure will be used as references throughout this document. 12

19 3.6. FIRST TIME RUNNING SEGMENT 3DPRINT Shift-Ctrl-R to reset GUI positions. This is also available under the File menu (a) PACS connection Setting up PACS connection and Segment Server usually requires help from your local PACS support, and we recommend that you contact us to setup a telephone / web-based video conference to make this process as smooth as possible. The Database and PACS connection manual and the Sectra PACS plugin manual is found at Medviso AB homepage ( The Sectra PACS plugin may require additional Microsoft Visual C++ components that can be downloaded from Medviso AB homepage ( 13

20

21 4 Loading and storing Data - Step by step 4.1 Loading data To load patient data from the database, follow Section To load patient data from PACS, follow Section To load patient data from network disc, follow Section Loading data from database 1. Click on the tool (a), Figure 8 is shown. Figure 8: Graphical patient selector. 2. Select patient. 3. Patients can be stored in two formats; DICOM, or Segment 3DPrint format. For images in DICOM format, click Select & Load, Figure 9 is shown. For images in Segment 3DPrint format click Load all. 15

22 CHAPTER 4. LOADING AND STORING DATA - STEP BY STEP 4. Select image series to load in Figure 9. Figure 9: Graphical image series selector. 5. Click on Load Loading data from PACS 1. Select Import from PACS in the File menu, Figure 10 is shown. Figure 10: Interface to import image stacks from PACS. 16

23 4.2. STORING DATA 2. Enter patient or study details and perform search. 3. Select study and series to load. 4. Load the study into Segment 3DPrint by select Open files Loading data from network disc 1. Select Open from Disc in the File menu, Figure 11 is shown. Figure 11: Image stack loading selector. 2. Select image stack to load. 3. For DICOM files, select crop size for the image stack and crop the image. 4. Click on Load Selected File(s) or Folder(s). 4.2 Storing data 1. To store images including delineation to Patient Database, click on the tool (b). 17

24 CHAPTER 4. LOADING AND STORING DATA - STEP BY STEP 2. To store images including delineation to Disc, select Save to Disc in the File menu (a). 3. To store images including delineation to PACS, select Save to PACS in the File menu (a). 18

25 5 Image viewing tools 5.1 Viewing options Choose panel viewing option by using,,,,,,, (c) To play a movie of the image stacks, use the tools, (d) To synchronize all image stacks, use the tool To change between single frame mode or all frames mode, use the tools, To hide the icons and the result panel, use the tools, 5.2 Crop image stack 1. Click on mode (h). 2. Select the tool (i). 3. Crop an image stack by mark the region in the image stack. 5.3 Zoom 1. Select image stack. 2. Use the tools and (f) to zoom in and zoom out in the image stack. 5.4 Adjust contrast Manual adjustment 1. Select image stack. 2. Select the tool (i) under mode (h). 19

26 CHAPTER 5. IMAGE VIEWING TOOLS 3. Adjust the contrast and brightness by right click and drag the mouse in the image stack (right-left for contrast and up-down for brightness) Automatic adjustment The automatic contrast and brightness adjustment is based on the intensity values inside the LV segmentation. If no LV segmentation exist, default HU values are used in the adjustment. To automatically adjust contrast and brightness in the selected image stack, select (i) under mode (h). To automatically adjust contrast and brightness in all loaded image stacks, select (i) under mode (h). 5.5 Measure distance 1. Click on mode (h). 2. Select the tool (i). 3. Measure a distance by dragging the mouse from start to end point in the image stack. 5.6 Hide/Show overlay 1. Select image stack with any type of overlay. 2. Use the tool to show or hide all overlays like segmentation, text, points, measurements. 5.7 Resample image stack Select image stack you want resample and choose the following resampling type: 20 Select Upsample/Downsaple Image (In Plane) under menu Resample image stack under menu Image tools to increase/decrease image resolution (in plane).

27 5.7. RESAMPLE IMAGE STACK Select Upsample/Downsaple Slices under menu Resample image stack under menu Image tools to increase/decrease number of image slices.. Select Upsample/Downsaple Temporal under menu Resample image stack under menu Image tools to increase/decrease number of time frames. 21

28

29 6 3D Segmentation module The module allows to perform a 3D segmentation for 3D printing purposes, or to export to CAD/CFD analysis software for patient-specific modelling. The segmentation process is to convert the image volume into a set of distinct objects by assigning which voxels in the image belongs to which objects. Before printing or exporting, the objects are converted to surfaces. 3D Segmentation module open after clicking 3D PRINT tab. An overview of the module is given in Figure 12. To learn more about each tool, hold the mouse over the icon in the software and a help text will be displayed. The standard 4 panel view dispaly the image with segmentation overlay in the three standard views; (transversal view, coronal view, sagittal view), and the fourth panel showing object likelihood. Figure 12: Overview of graphical user interface. The letters (a-k) in the figure will be used as references throughout this chapter D Segmentation - Step by step This section outlines the typical steps involved when creating a 3D model. Details on how each of the tool works is given in Section

30 CHAPTER 6. 3D SEGMENTATION MODULE 1. Create a vision. The first step is to create a vision on how the finished printed object should look. What parts of the image volume should be included? How should the objects be connected so that the printed model does not fall apart? Creating a vision together with the colleages that will used the finished model is essential, for sucessful usage of the technology. 2. Crop the image volume. This makes the entire workflow faster and easier. Cropping is performed with the crop tool. Crop relatively tight around object of interest. Objects from different cropped regions can be combined in exporting step as the software remembers position of objects also after crop operations. 3. Adjust object likelihood function. Object likelihood is a mapping function that maps voxel intensities to likelihood of the object to be segmented. Use either predefined mapping functions or manually adjust the curve. Type of mapping function can be changed by selecting one of the three mappint functions: Pos - Positive, Neg - Negative, Gauss - Gaussian. Object likelihood is adjusted in panel i, and object likelihood view is presented in panel (g). 4. Create object. There are both automatic, semi-automatic and manual methods to segment an object using,., and. All segmentation tools are described in Section Separate objects. Often objects are connected and needs to be separated. There are two main methods to split an object. A) Use the split and erase tool one object from another object., or B) Use set operations of objects to remove 6. Manual adjustments. Manual segmentation tools are described in Section 6.2. The suggested workflow is to work with one object at a time. Do your manual adjustment for one object before you start processing the next object. In the end of the process you can merge and split objects as desired. 7. Fill objects. In general, it is not necessary to have solid objects for 3D printing, however filling is highly recommended as A) Smoothing of non filled objects may erase sections of the object, B) filled object 24

31 6.2. SEGMENTATION TOOLS (PANEL A) generates simpler surfaces and better control of support structures for 3D printing. Fill of objects is done using the fill tool. 8. Smooth object. Before exporting the object smoothing is desirable. For details, see Section Export / preview. When the segmentation of the desired anatomical objects is complete, start the export/preview tool. For details, see Section Segmentation tools (panel a) Segmentation tools are generally applied to current object shown in panel i, except for some tools that creates new objects. 2D/3D. (hot-key 2, 3) Selects if tools are applied in 2D or 3D. Crosshair tool. Put crosshair to mark point in 3D. Wand tool. Magic wand from mouse click. Clicking on the area performs automatic segmentation according to probability map threshold. Threshold image volume. This creates a new object by thresholding. Threshold is adjusted by object likelihood function. 3D paint tool. Hold down mouse and drag small sections of the object at a time. Local threshold. A local threshold operation is performed within the defined pen radius. This tool can be used to locally paint regions according to object probability. Pen radius in mm is adjusted in panel g. Draw tool, add region to segmentation. adjusted in panel g. Pen radius in mm is Erase tool, remove region from segmentation. Pen radius in mm is adjusted in panel g. Active contour tool (hot-key a). Applicable in 2D. Adjusts and smooth contour using active contour algorithm. Use Ctrl-mouse wheel to go to next slide and apply active contour tool. Fill tool (hot-key 2D: f, 3D: F), fill object in 2D/3D. 25

32 CHAPTER 6. 3D SEGMENTATION MODULE Split and erase tool, splits the current object and cuts away undesired part. For details, see Section 6.6. Pick tool (hot-key p), left click in image to pick an object and add to current object. In 2D mode use shift and mouse wheel to pick objects at crosshair position from multiple slices. Keep tool (hot-key k), left click in image to keep a part of the current object and create a new object. Typically useful with vessel tool. Vessel tool, opens a new user interface for vessel segmentation. For details, see Section 6.8. Support structure tool, creates support structure between two consecutive points. Add Text tool, adds label to object. Box tool, add segmentation / remove segmentation inside 3D box. This tool can be used to crop objects and to add support structures for models. Place annotation point tool (hot-key p), right click at the desired position to set a point. This tool can be used to set up the support structure. Clear point tool, clears all annotation points. Cut in 3D menu, drop down menu to select cut plane. Cut in transversal plane tool, cuts segmentation in transversal plane and creates two cut planes: lower and upper. Cut in sagittal plane tool, cuts segmentation in sagittal plane and creates two cut planes: left and right. Cut in coronal plane tool, cuts segmentation in coronal plane and creates two cut planes: anterior and posterior. Cut in obliquel plane tool, cuts segmentation in oblique plane and creates two cut planes.it is needed first to set up at least three points to place plane for cutting. Crop tool, crop image volume. 26

33 6.3. OBJECT TOOLS (PANEL K) 6.3 Object tools (panel k) Working with objects is a good way to divide the segmentation task in smaller steps. The segmentation process typically creates many objects, and it is highly recommended to rename objects with understandable names and delete temporarily created objects as it may be hard to remember the content of each object. Objects may also be of use in the segmentation process by using set operations,,, to combine objects algebraically. In particular this is extra powerful when the objects have been created using different object likelihood mappings. Sort out object. Removes non connected parts. For details, see Section 6.5. Smoothes object. For details, see Section 6.7. Erode object. Creates a new object that is one voxel thinner than the parent object. Dilate object. Creates a new object that is one voxel thicker than the parent object. Close object. Closes the borders of the object. Hollow. Makes a filled object hollow. Minus. Takes first selected object minus the second selected object, and creates a new object. Not. Inverts the object, and creates a new object. Union. object. Joins two or more selected objects, and creates a new Cut. Takes the intersection of two or more selected objects, and creates a new object. Move up. Moves object up in list. Move down. Moves object down in list. New object. Creates a new object. Copy object. Copies the selected object to a new object. Rename object. Renames the selected object. Delete object. Deletes the selected object. 27

34 CHAPTER 6. 3D SEGMENTATION MODULE 6.4 View modes (panel b) Standard view with, transversal view, sagittal view, coronal view, and object likelihood image. To switch between active view use, or the hot keys (t = transversal view, s = sagittal view, c = coronal view) Display - transversal view & likelihood image (hot-key T) Display - sagittal view & likelihood image (hot-key S) Display - coronal view & likelihood image (hot key C) Display segmentation countour (green line) Display segmentation as red overlay Display segmentation in a 3D view Preview / generate STL file. For details, see Section 6.9 Delete segmentation Delete all 6.5 Sort out tool Select. This tool is used to remove non connected parts of an object by using of three following options: (1) by keeping only the largest or (2) by keeping N the parts larger then threshold as separate objects or (3) by removing N parts smaller than the threshold. The user interface is shown in Figure 13. The number of parts in the object are shown. Default option is to keep only the largest part. You can select to keep all parts larger than a specified threshold, then each kept part is created as new object. You can also remove N parts smaller then threshold, without creating separate objects for each kept part. 6.6 Split tool Select. Put crosshair on the object you want to keep and right click on object you want to cut away. If possible put cross hair and right click on opposide side of a clear and visible cut region. If possible put the cross hair on solid parts such as the bone to indicate clearly parts that should not be deleted. If you want to keep both the objects, but only split them apart, 28

35 6.7. SMOOTHING TOOL Figure 13: User interface for Sort out. make a copy of the object prior to splitting, work to split out one and use the subtruction function (described in 6.3) to get both desired objects. 6.7 Smoothing tool Select. Smoothing surface window is shown in Figure 14. How to apply smoothing step-by-step: 1. Adjust smoothing with editbox or slider 2. Press Apply to apply the smoothing and create a new smoothed object. 6.8 Vessel tool Select. Start by adjusting the object likelihood to enhance the vessels of interest maximally. Often if you want to enhance vessls, close to bones as illustrated in Figure 15, then use the Gaussian mapping as illustrated in the 29

36 CHAPTER 6. 3D SEGMENTATION MODULE Figure 14: Smooth surface window. On the left segmentation before smoothing, on the right segmentation after smoothing. figure. Figure 15: Adjust object likelihood to enhance vessels. Next start the vessel enhancement tool Figure 16.. The user interface is shown in In the user interface select start and end scale for the vessel enhancement. For large vessels it is recommended to use the 3D draw tool. Click Finish to create a new vessel enhancement object. Often from this object it is advantagous to use the keep tool 30 to pick vessels to include in the

37 6.9. PREVIEW / GENERATE STL FILE Figure 16: User interface for vessel enhancement. final object. Then use to merge the kept objects. Often you also may want to use the tool to merge the detected vessels with a larger vessel manually draw as the vessel enhancement may fail precisely when a small vessel branches of a large vessel. 6.9 Preview / generate STL file Select. The user interface to generate STL-files is shown in Figure 17. Step-by-step: 1. Select object to add to preview. 2. Select color with Color. 3. If desired, adjust transparity. 4. Click Add object. 31

38 CHAPTER 6. 3D SEGMENTATION MODULE Figure 17: Preview / generate STL file user interface. 5. Repeat step 1-4 for all desired objects. 6. Inspect the visual output from all angles. 7. If the completed model is acceptable, then click Create STL, otherwise go back and edit the segmentation Storing segmentation When the 3D segmentation module is closed all objects are retained. Use the standard save functions in Segment 3DPrint to store segmentation and image volume to disc or patient database. 32

39 7 CT Cardiac analysis - Step by step 7.1 Cardiac reconstruction To be able to perform automatic segmentation of the left ventricle and right ventricle, as described in 7.2 and 7.4, the original transversal CT image stack must be reconstructed to short-axis image stack. 1. To reconstruct the transversal image stack select LV mode and then the tool (i). A new interface is open, as shown in Figure 18. Figure 18: Interface for cardiac reconstruction. 2. Use left mouse click to define LV AV-plane (two blue points). Use right mouse click to define LV epicadial apex (one red point). Use shift-click on the mouse to define pulmonary valve (one magenta point). 3. Click on RV Slice and LV Slice to toggle between the slices where you have placed the RV point and the LV points, respectively. 33

40 CHAPTER 7. CT CARDIAC ANALYSIS - STEP BY STEP 4. Click on Generate orthogonal image. This generates a orthogonal image, as defined by the dotted red line in the Original image panel. 5. In the orthogonal image, use left mouse click to define LV AV-plane (two blue points) and right mouse click to define LV epicadial apex (one red point). 6. Enter slice thickness for the final generated short-axis image stack. 7. Select if you would like to perform automatic LV segmentation and/or automatic RV segmentation on the generated short-axis image stack. 8. Click on Generate Short-axis Stack to generate the short-axis image stack to the main Segment 3DPrint interface. 9. If you already have been generated a short-axis stack and only would like to store the defined reonstruction points, click on Save Point Positions. 7.2 LV analysis Automatic LV segmentation 1. The automatic LV segmentation is applied to a reconstructed shortaxis stack, generated by cardiac reconstruction as described in Chapter 8. The LV slices to be included in the automatic LV segmentation is defined by the points entered in the reconstruction process. 2. Start the automatic LV segmentation by select LV mode (h) and select (i). 3. The segmentation result is presented in the main window where volumes curve can be reviewed (k) and the measured LV volumes are presented, according to Figure 19 (j). 4. If needed, correction of the LV segmentation can be performed by using the following tools:,,,,,, and (i) Manual LV segmentation For manual delineation of the LV, the following tools can be used:,,, (i). If the checkbox Single Frame Mode (m) is selected, then the operation is only applied to the current time frame. 34

41 7.3. ERASE LV SEGMENTATION Figure 19: LV analysis result. 7.3 Erase LV segmentation To erase the LV segmentation, select Clear All LV Segmentation from LV menu (a) Validation of LV segmentation 1. A. Martinsson, A. Gilland, J. Backe, H. Fransson, E. Heiberg, G. Engstrom, J. G. Smith, E. Ostenfeld, Assessment of ventricular dimensions from computed tomography in a prospective cohort study: the SCAPIS Malmo-cohort. EuroEcho Imaging Summary of LV segmentation validation The result in [1] showed a bias between automatic LV segmentation by Segment CT and manual delineation for LVM 10.9 ± 8.5 g and EDV -9.3 ± 5.7 ml in a set of 200 patients. No uncertainty/error information is showed in the software together with the LV measurements. 35

42 CHAPTER 7. CT CARDIAC ANALYSIS - STEP BY STEP 7.4 RV analysis Manual RV segmentation For manual delineation of the RV, the following tools can be used:,,, (i). If the checkbox Single Frame Mode (m) is selected, then the operation is only applied to the current time frame Erase RV segmentation To erase the RV segmentation, select Clear All RV Segmentation from RV menu (a) Validation of RV segmentation 1. E. Heiberg, J. Sjogren, M. Ugander, M. Carlsson, H. Engblom, and H. Arheden, Design and validation of Segment freely available software for cardiovascular image analysis, BMC Med Imaging 10 p 1, Summary of RV segmentation validation Manual segmentation and measurements for right ventricular volumes are performed withing the scope of [1]. No uncertainty/error information is showed in the software together with the RV measurements. 36

43 8 MR Cardiac analysis - Step by step 8.1 LV analysis Automatic LV segmentation [1] 1. Start the LV analysis by select mode (h) and select (i). A new interface is open, as shown in Figure 20. Figure 20: LV analysis GUI. 2. Select the slices covering the left ventricle by left-click to select most basal slice and right-click to select most epicardial apical slice. 3. Review the slice selection in the long-axis views. 4. Ensure that the center cross is in the middle of the LV. 5. Start the automatic LV segmentation. 37

44 CHAPTER 8. MR CARDIAC ANALYSIS - STEP BY STEP 6. The segmentation result is presented in the main window where volumes curve can be reviewed (k) and the measured LV volumes are presented, according to Figure 21 (j). Figure 21: LV analysis result. 7. If needed, correction of the LV segmentation can be performed by using the following tools:,,,,,,,, and (i) Manual LV segmentation For manual delineation of the LV, the following tools can be used:,,, (i) Erase LV segmentation To erase the LV segmentation, select (i) from mode (h) Copying LV segmentation To copying the LV segmentation to another image stack, select Import Segmentation From Another Image Stack from LV menu (a). 38

45 8.2. RV ANALYSIS Validation of LV segmentation 1. J. Tufvesson, E. Hedstrom, K. Steding-Ehrenborg, M. Carlsson, H. Arheden, E. Heiberg, Validation and development of a new automatic algorithm for time resolved segmentation of the left ventricle in magnetic resonance. Biomed Res Int, 2015: RV analysis Automatic RV segmentation [1] 1. Start the RV analysis by select mode (h) and select (i). A new interface is open, as shown in Figure 22. Figure 22: RV analysis GUI. 39

46 CHAPTER 8. MR CARDIAC ANALYSIS - STEP BY STEP 2. Set RV center and select the slices covering the right ventricle by leftclick to select most basal slice and most apical slice. 3. Review the slice selection and center line in the long-axis views. 4. Start the automatic RV segmentation. 5. The segmentation result is presented in the main window where volumes curve can be reviewed (k) and the measured RV volumes are presented, according to Figure 23 (j). Figure 23: RV analysis result. 6. If needed, correction of the RV segmentation can be performed by using the following tools:,,,, (i) Manual RV segmentation For manual delineation of the RV, the following tools can be used:,,, (i) Erase RV segmentation To erase the RV segmentation, select (i) from mode (h). 40

47 8.2. RV ANALYSIS Validation of RV segmentation 1. M. A. Aneq, E. Nylander, T. Ebbers, and J. Engvall, Determination of right ventricular volume and function using multiple axially rotated MRI slices, Clin Physiol Funct Imaging 31(3) pp ,

48

49 9 ROI analysis - Step by step 9.1 Manual ROI analysis 1. Click on mode (h). 2. If the checkbox Single Frame Mode (m) is selected, then the operation is only applied to the current time frame. 3. To place a ROI, use one of the tools (i) or. 4. Scale the ROI with the tool (i). 5. Translate the ROI with the tool (i). 6. Change the ROI segmentation by using the correction tool (i). 7. Select ROIs with Shift-click with the tool (i). 8. Set the label of selected ROIs with the tool (i). 9. Set the color of selected ROIs with the tool (i). 10. Select ROI Analysis from ROI menu to perform ROI analysis according to Figure 24 (a). 9.2 Erase ROI segmentation Delete selected ROIs by selecting Delete Selected ROIs, and all ROIs by selecting Clear All ROIs in the ROI menu (a). 9.3 Validation of the ROI analysis 1. E. Heiberg, J. Sjogren, M. Ugander, M. Carlsson, H. Engblom, and H. Arheden, Design and validation of Segment freely available software for cardiovascular image analysis, BMC Med Imaging 10 p 1,

50 CHAPTER 9. ROI ANALYSIS - STEP BY STEP Figure 24: ROI analysis GUI Summary of ROI validation Manual segmentation and measurements for ROI volumes are performed withing the scope of [1]. No uncertainty/error information is showed in the software together with the ROI measurements. 44