As radiology becomes more and more digital, teaching files can be created faster and more easily if appropriately integrated with other applications, according to Dr. Luciano Prevedello at Brigham and Women’s Hospital.
As radiology becomes more and more digital, teaching files can be created faster and more easily if appropriately integrated with other applications, according to Dr. Luciano Prevedello at Brigham and Women's Hospital.
For his efforts in customizing the RSNA's Medical Imaging Resource Center (MIRC) into a user-friendly and intuitive teaching file, Prevedello took home the top prize for best student paper at the Society for Imaging Informatics in Medicine meeting.
The ideal digital teaching file system must be versatile, reliable, access-controlled to prevent inappropriate use, intuitive, compatible with a variety of different images, and flexible. It should also not be time-consuming to create.
MIRC software is an open source program that can be customized and set up as a teaching file system for public or private use. It is also used to share clinical trial data. The program has been in use worldwide for years and as such has been continuously tested, evaluated, and improved.
While many institutions use this program to create digital teaching files, creating the files can be time-consuming if the tools are not optimally integrated with other radiological applications.
At the Brigham, researchers opted to install MIRC in a server to have more flexible and reliable transmission of data. This move also gave them the capability to expand the data space, adding relatively low-cost hard drives as needed in the future.
Prevedello and colleagues achieved integration between PACS and the MIRC server using built-in features of both applications. Images are viewed and selected in PACS and sent to the MIRC server as DICOM files through a "DICOM send" function available in many PACS today. All information is de-identified to be compliant with the Health Insurance Portability and Accountability Act.
Information obtained from DICOM headers populates the Abstract and History fields with the patient's age, sex, body part studied, imaging modality, and a brief description of the reason for the study. Other fields in the Index Content Section that are not visible to the user, such as anatomy and modality, are automatically populated.
"Adding information to the Index Content Section is important so that cases can be indexed and searched by specific criteria," Prevedello said.
MIRC software uses a template in order to incorporate images and information into a document. Researchers at the Brigham created a template to allow users to choose between a format that is more suitable for lecture presentation and one used for studying purposes.
The current version of MIRC makes it difficult and time-consuming to insert videos, calling for the need to write HTML code. Researchers solved this problem by embedding the code in a section of the template, so authors need type only the name of the file to be inserted. Animated ultrasound videos also work using this functionality.
Researchers created a button that simulates cine visualization to accommodate multiple continuous slices of images from multislice CT. They did this by incorporating JAVA and Extensible Stylesheet Language (XSL) codes in two files of the program using a regular text editor.
"With this function, images can be presented automatically with the click of one button only," Prevedello said.
The Brigham program has restricted access to staff and residents in the radiology department. After creating and reviewing their cases, authors can show selected MIRC cases in any conference room of the hospital using password-protected access. A published paper coming out soon from the group fully details how they implemented the program.
Those interested can find information about MIRC here and here.
Deep Learning Detection of Mammography Abnormalities: What a New Study Reveals
June 19th 2023In multiple mammography datasets with the original radiologist-detected abnormality removed, deep learning detection of breast cancer had an average area under the curve (AUC) of 87 percent and an accuracy rate of 83 percent, according to research presented at the recent Society for Imaging Informatics in Medicine (SIIM) conference.
Detecting Intracranial Hemorrhages: Can an Emerging AI Advance Have an Impact?
June 9th 2022A 3D whole brain convolutional neural network could provide enhanced sensitivity and specificity for diagnosing intracranial hemorrhages on computed tomography, according to new research presented at the Society for Imaging Informatics in Medicine (SIIM) conference in Kissimmee, Fla.
Deep Learning Detection of Mammography Abnormalities: What a New Study Reveals
June 19th 2023In multiple mammography datasets with the original radiologist-detected abnormality removed, deep learning detection of breast cancer had an average area under the curve (AUC) of 87 percent and an accuracy rate of 83 percent, according to research presented at the recent Society for Imaging Informatics in Medicine (SIIM) conference.
Detecting Intracranial Hemorrhages: Can an Emerging AI Advance Have an Impact?
June 9th 2022A 3D whole brain convolutional neural network could provide enhanced sensitivity and specificity for diagnosing intracranial hemorrhages on computed tomography, according to new research presented at the Society for Imaging Informatics in Medicine (SIIM) conference in Kissimmee, Fla.
2 Commerce Drive
Cranbury, NJ 08512