SHIFT and ALT

Client

Ramin Zabih, Professor, Computer Science Department (rdz@cs.cornell.edu)
George Shih, Weill Cornell Medical College (ges9006@med.cornell.edu)

The clients propose two project applying imaging techniques to health care.

SHIFT: Health Care Micro-integration

Healthcare institutions use many different information systems, most of which are proprietary and have few or no API’s available for integration. Most integrations occur using older healthcare standards such as HL7 and/or DICOM, requiring commercial vendor permission and involvement which takes a lot of time and resources, if permission is ever granted.

With the advent of always connected mobile devices (smartphones, tablets, etc) that have cameras, a new ad-hoc form of micro integration is possible using small pieces of information (e.g., patient name, medical record number, and/or exam number).

The SHIFT project will:

1. Take a photo from a hospital computer screen which cones down on a small piece of patient information (e.g., medical record number).
2. Use computer vision techniques (e.g., OCR) to process the photo.
3. Send that piece of data via a URI to another health information system (which would either be web-based or have a RESTful API which accepts URI’s).
4. Lasso would then launch that new application which would provide a micro integration obviating proprietary vendor integration solutions.

SHIFT would be able to store different types of URI’s for different micro integrations.

ALT - Alternative Open Source Imaging Mobile Viewer

There is an ongoing explosion of mobile tablets (mostly iPad, some Android Honeycomb) in medicine, which promises to replace the ubiquitous clinician clipboard as the viewing platform of patient information at the bedside and on clinical rounds.

Currently, there is no open source native application for the iPad for viewing imaging exams (eg, x-ray, CT, MRI, ultrasound). Certain commercial vendors are implementing their own native mobile app for the tablet, but requires purchase of their product. This project would implement a hybrid native app (native app shell with buttons, with web elements) which interfaces with an open source imaging backend called DCM4CHEE (http://dcm4che.org). The purpose of this open source project would be to allow easy deployment of a free solution that allows viewing of imaging exams on tablets.

We propose project ALT to include an iPad native app (using hybrid web approach) which has the following short list of features:

• Query imaging exam using patient name, medical record number, or exam number
• Buttons for scrolling through images
• Ten window level presets (eg, lung, abdomen, head, etc)
• Some local caching of images to improve image scrolling performance
• Hooks into DCM4CHEE WADO (Web Access to DICOM Objects) -- providing a free backend server
• RESTful API to allow for integration into other applications (eg, including SHIFT)