August 2023
Features
New MD.ai Chat Features
Custom Report Prompts
Help users explore and interact with reports by leveraging the power of MD.ai Chat. The custom report prompt feature allows report owners to share their report with specific customizable prompts attached which will be available for one-click chat interactions. These example prompts can help inspire users to understand the report more deeply, as well as generate their own prompts.
Interactive Prompt Templates
Make prompts interactive for easy fill in the blank queries.
Edit, Rate, Copy, Delete, Export, Email
New chat features allow for easy editing or deleting of prompts, and interactive response actions including edit, rate, copy, delete, export to report, and send as email.
Faster De-ID
Remove unwanted pixel data faster with improved processing speeds, and the option to use auto inference to automatically remove unwanted data, like burned in PHI, or require a manual review step before removal.
Compress Exported Annotations
New export options allow annotations, images, model outputs, and DICOM metadata to be exported compressed to facilitate the management of large projects. Exported file size is also displayed in the UI.
Timestamps on Video Annotations
For simplified annotation, timestamps are now automatically applied to image-scoped annotations applied to videos, and timestamp information is included in annotation exports.
DICOM SR and DICOM SEG support
New functions support annotation export to DICOM SR/DICOM SEG, and DICOM SR import.
Dataset Sorting
Locate and navigate between datasets faster with new sorting options including alphabetical sort and date created sort. These options can be found in projects in the navigation column, in the Dataset
menu.
Bug Fixes
- Fixed bug where metadata label annotations were not visible to users in crowdsourcing mode.
- Fixed bug in MPR display mode where image would fail to display if PixelSpacing value was null.
- Fixed bug causing blinded users with label restrictions to be unable to see annotations they applied.
- Fixed bug delaying update of dataset number when multiple dataset deletions or additions were occuring.
- Implemented additional annotation import validations to prevent importing of annotations with labels from alternative projects and subsequent invalid label errors.
- Resolved bug in which network errors during conversion of JPG, PNG, or BMP files to DICOM could result in duplicates.
- Fixed bug that could cause out of memory error when cloning model outputs as annotations.
- Resolved bug causing slowed performance of study ordering in projects with multiple datasets.
- Fixed invalid session login error which could occur on attempted login if multiple pages open for extended time.
- Fixed bug that was preventing autosegmentation from working on color images.
- Fixed bug that was leading to slow thumbnail display loading in exam lists.
- Resolved bug where activating an annotation would display annotation measurements, even when the
Annotation Measurements
layer was toggled off. - Fixed bug that could cause out of memory error when processing large numbers of >500MB DICOM files.