During this year, the MRIPS file, web, and ftp servers have been maintained at their current level. Enhancements continued to be made to the IDL and PV-Wave servers. Over the past fiscal year MEDx has been upgraded to Version 3.44 and has continued to be used widely at NIH. There have been approximately 650 NIH customer requests from 12 ICDs. This indicates that MEDx has widespread usage across the campus. The following additions have been made to MEDx over the past year: MEDx has been upgraded to version 3.44. The new functionality includes the following: 1. A set of applications added to the volumetric menu in the MEDx toolbox. These applications were developed primarily as scripts for NIH users. For the most part, these applications have been published in peer review journals. Including them as part of MEDx shall increase the availability of these well tested applications. A set of tutorials has been written and appears on the NIH MEDx web page. a. MS Brain Lesion Analysis. This application is used to track the progression of brain lesions over time which are visibile in multiple MRI sequences. A novel method is used to determine if a lesion is the same in each temporal acquisition. b. Automated DSC Perfusion Analysis. This application implements a novel technique to automatically define the arterial input function. This is the main manual step in the MR perfusion analysis module already existing in MEDx. With this new application, the MR perfusion analysis is far more automated and much less time consuming. c. CT Brain Extraction. MEDx has two techniques to automatically or semi-automatically segment out the brain in MR images. However, methods do not exist to perform this function on CT images. We have implemented a method that is fully automatic and works well on CT images. The output of these de-skulled CT brains can be used in the Analysis of Brain Lesions (ABLe)module in MEDx. d. Automated Body Lesion Detection. We previously developed an application for NCI researchers to investigate the volumetric growth rate of plexiform neurofibroma lesions in NF1 patients as seen in MRI scans. This method is published and has been used for several years. To make the initiation of the program quicker, as well as to make it useable by a wider group of people, we have added it as an option to the MEDx toolbox. e. Subcut/Intra-abdominal Fat Measurement. This application was developed to measure the subcutaneous as well as intra-abdominal fat in CT scans of the abdomen. The measurements are burned into the images and sent back to the PACS system. 2. Extension of the Analysis of Brain Lesions (ABLe) module. The following functions have been added to ABLe in the 3.44 version of MEDx (ABLe 2.2). a. Option to load a lesion mask created from another application such as MRIcro. b. New output report formatted for SPSS database and including addition lesion/structure interactions. c. Option in Two Group Analysis to specify threshold for number of subjects in each group. d. New behavioral analysis which automatically performs a voxel based lesion-symptom mapping, creating a table representing Talairach locations where subjects with lesions at a voxel perform statistically significantly worse on behavioral tests than subjects without lesions at that voxel. e. Unlimited number of subjects are now allowed for group analysis (only dependent on the amount of RAM on the machine). Previous versions of ABLe had a maximum of 32 subjects. f. New html reporting structure allowing for user to navigate through hierarchy of anatomic brain structures looking for lesion involvement. 3. Availability of MEDx on the MacOS X operating system. All modules with the exception of SPM 99 are available.
| Solomon, Jeffrey; Mavinkurve, Sara; Cox, Derrick et al. (2004) Computer-assisted detection of subcutaneous melanomas: feasibility assessment. Acad Radiol 11:678-85 |
| Solomon, Jeffrey; Warren, Katherine; Dombi, Eva et al. (2004) Automated detection and volume measurement of plexiform neurofibromas in neurofibromatosis 1 using magnetic resonance imaging. Comput Med Imaging Graph 28:257-65 |
