The Current State of Image Exchange
* denotes a link to an article requiring a subscription
It’s been one year since our previous article on interoperability and the state of image exchange has continued to evolve. The need to share information between healthcare organizations was exemplified and magnified by the COVID crisis. As one would expect this increased demand has expanded the market for information sharing technology, including technology for image exchange. This market is expected to reach $7.97 billion by 2032. In this current state of heightened awareness, many challenges remain.
First, many different capabilities are being labelled image exchange. When discussing image exchange, one needs to understand the exact capability that is being discussed. These capabilities include:
- Electronically responding to a request for image exchange
- Providing links to a given study at the point of care
- Sharing thumbnails of key images or reference images
Image Exchange Defined
Expectations
In the spirit of Enterprise Imaging, with the goal of creating a comprehensive longitudinal medical record, image exchange should be defined as the capability to provide access to the full fidelity image data set and supporting metadata. Images should be accessed with an appropriate viewer with tools that allow interaction with the images to support a complete evaluation of the information being presented. As referential viewers become more sophisticated these capabilities are easily delivered.
Electronic Sharing
The current state of image exchange is based on the electronic sharing of imaging studies. CDs and other physical media are dwindling as these capabilities become more widespread. The opportunities in this space lie with how various image exchange vendors interact with each other and what triggers are used to initiate the image exchange.
Study Link
Some diagnostic and referential viewers offer providers and patients, at the point of imaging viewing, the ability to share a link to a specific study. This link may be shared with providers, patients, or other relevant parties. A welcome capability, this link sharing advances information exchange, and provides access to the entire contents of the imaging study and its metadata. It also creates a mechanism for patient directed image sharing. Patient directed image sharing should be considered a highly desirable feature of an image exchange program.
Thumbnails and Reference Images
The ability to share key images or thumbnails of imaging studies is now being offered through the EMR. This process has several limitations. First, not all imaging professionals create key images, even when studies have significant findings. In situations where key images are not created, a random reference image will be shared. Thus, the shared image may not contain any relevant pathology; the diagnostic report will provide more valuable information. Imaging professionals may not desire to select key images fearing that clinical decisions will be made without a complete understanding of the pathology demonstrated in the imaging study. This capability serves more as a study preview than image exchange.
Current State Centers on Cloud-Based Image Exchange
As image exchange has transitioned from physical media such as CDs to a digital format, cloud-based models have emerged as the current standard. Cloud-based services create a copy of the imaging study from the acquiring institution and that copy is maintained in the image exchange vendor’s cloud. Studies are maintained in the cloud for a variable period. Caregivers at the receiving institution may view the images in the cloud and/or elect to import the imaging study into their local image archive. These services also offer the ability to auto-send imaging studies from the cloud to the receiving institution’s image archive. This architecture underlies that leading image exchange vendors as identified by KLAS. These vendors include Intelerad (Ambra), Nuance PowerShare, and lifeIMAGE.
The image exchange event may be initiated by the acquiring institution when a specific event occurs; for example, the patient is transferred from the Emergency Department to a Trauma Center. Alternatively, when the receiving institution is seeking imaging studies on a patient they must have knowledge of the existence of the image study at a particular institution A request to share must be sent to the organization that created the study.
Currently many healthcare systems may have more than one image exchange vendor operating in their environment. This situation places a large burden on healthcare organizations. They must maintain several different exchange vehicles. For tertiary and quaternary organizations, the numbers can be quite staggering and the environment unnecessarily complex. Many of the image exchange vendors have been unwilling or unable to establish connections amongst themselves.
Architecting Connectivity
Linked Multihub Model Would Connect Image Exchange Vendors
Larson and colleagues in a recent article theorized multiple different models for connectivity for image exchange. These models were created to illustrate ways in which image exchange can be amplified to allow universal image sharing. The cloud model previously described exemplifies their single central hub model and best describes the current state.
These authors acknowledge that the most achievable architecture to establish widespread image exchange interconnectivity throughout the United States is the linked multihub model. In this model each organization can select its own image exchange vendor and that vendor acts as a central hub. The central hubs (image exchange vendors) then connect with each other. A universal network of central hubs would then provide the needed interconnectivity amongst all the individual healthcare clients. It would not necessarily eliminate the vendors that exist today. Resistance for the various image exchange vendors to connect with each other is the main hurdle which must be overcome to create this structure for image exchange. As recognized by the authors, the market needs to demand that these connections occur. We applaud Larson and team for their advocacy.
Larson et al identified the peer-to-peer network as the “most desirable in the long run” and acknowledges that building such a network is an enormous undertaking. In the peer-to-peer network imaging studies are directly shared from one organization to another. Such a network eliminates intermediaries such as the central hub. We previously referenced the similar global image sharing infrastructure proposed by Motta et al. As the image exchange evolution occurs, we can anticipate seeing emerging peer-to-peer networks seamless interact with central hubs (cloud-based systems) in a hybrid of the linked multihub system and peer-to-peer network. Again, healthcare organizations should set these types of expectations for their image exchange vendor partners.
Today’s Federated Networks Taking Us Closer to the Future
Companies such as Medicom Technologies are developing a federated network which is a peer-to-peer network. The Medicom network enables image sharing between all members of the network. This network introduces the concept of automated search. A receiving institution does not need to have prior knowledge about the existence of a study. Additional functionality supports automated search and retrieval streamlining the process and freeing manpower to support other workflows. This model most closely meets the functionality required for the free flow of information among healthcare providers, reducing information gaps and supporting information flow across the care continuum. In countries other than the United States, country wide image exchange systems have already been established. The UK has built a country wide medical imaging sharing system that spans more than 500 institutions utilizing Sectra technology.
Health Information Exchanges Support Image Sharing
Health Information Exchanges are recognizing the importance of imaging studies as an essential component of a patient’s medical record. To provide the most comprehensive view of the patient, image-enabled HIEs are becoming more common. Examples include HealtheConnections in the Hudson Valley, the Kentucky HIE (KHIE) partnered with Philips Healthcare , Big Sky Care Connect in Montana using Medicom, and CRISP in the Maryland and DC area. The incorporation of image sharing into the HIE brings with it an appreciation for the clinical provider's needs and expectations, distinct from the imaging professional's needs and expectations. This distinction is discussed below.
Image Exchange versus Remote Image Viewing
At the beginning of this article current capabilities were described as “electronically responding to a request for image exchange “. The form of that response was not defined. In this article two different mechanisms have been discussed:
- Creation of a copy of the imaging study which is stored or transmitted through a cloud intermediary
- Direct send of a copy of the imaging study from one institution to another
A third mechanism is being utilized, especially as part of the HIE integrations
- Remote viewing of a study that is maintained in the acquiring institution’s image archive
In this remote viewing mechanism, there is no copy of the study and no movement of imaging data into the recipient’s image archive. This functionality is highly desired by clinical providers as they interact with the HIE. These clinical providers do not need the study to be locally. Contrast this need with the needs of the imaging professional. One of the primary drivers behind image exchange for the imaging professional has been the collection of relevant comparison examinations created at external organizations. These studies needed to be imported into the image archive for display in the diagnostic viewer. As image exchange serves clinical care across the enterprise different mechanisms for image exchange/sharing will be implemented to meet different end-user needs. Perhaps the mechanism of remote image viewing should be differentiated from actual image exchange/sharing.
Triggers for Image Exchange
Lastly, let us explore what triggers an image exchange event. Currently, an image exchange is initiated when a provider, with advanced knowledge of a study’s existence, makes a request to the acquiring organization . This request may be verbal, written, or electronic. There are many issues surrounding this request which are beyond this scope of this article. These issues include concerns about patient privacy, HIPAA and release of information policies. When these concerns are satisfied the acquiring institution fulfills the request for access. In many institutions image exchange is seen primarily as a service supporting radiology or cardiology providing relevant comparison examinations. Image exchange needs to expand beyond this current state in two ways. First, the trigger for exchange needs to evolve from radiology- or cardiology- focused to supporting the entire care continuum. Triggers should include any care event, such as a clinical appointment, where access to the patient’s imaging studies will support timely, high quality care. Secondly, these triggers need to move from manual to automated. To support this automation searching functionality is required. The need to know about the existence of a study must be eliminated. The care event should trigger an automated search of the network. As rules, regulations and attitudes evolve, these automated searches should lead to the automated retrieval of studies. Only then will we have timely access to the comprehensive medical record.
Final Note
By continuing to explore and refine the vision of the future of image exchange and by setting high expectations for our vendor partners, image exchange will continue to evolve along a path that best supports the needs of patients, clinical providers and imaging professionals.
Cheryl Petersilge is paid consultant with Medicom Health Technologies, LLC and Philips Healthcare.
References
1. Image sharing is critical to robust healthcare information exchange. http://vidagos.com/image-sharing-is-critical-to-robust-health-information-exchange-hie/. Accessed October 10, 2022
2. Medical image exchange systems market further expand at a CAGR of 7.8% to reach $7.97 BN by the year 2032. https://www.globenewswire.com/news-release/2022/03/15/2403732/0/en/Medical-Image-Exchange-Systems-Market-further-expand-at-a-CAGR-of-7-8-to-reach-US-7-97-Bn-by-the-year-2032.html. Accessed October 10, 2022
3. 2022 Best in KLAS Image Exchange. https://klasresearch.com/best-in-klas-ranking/image-exchange/2022/236. Accessed October 10, 2022
4. Larson DB, Krishnaraj A, Mendelson DS, Langlotz CP, Wald C. Moving toward seamless interinstitutional electronic image transfer. JACR. 19;460-468, 2022. https://www.sciencedirect.com/science/article/abs/pii/S1546144022000096 *
5. Motta GHMB, Araújo DAB, Lucena-Neto JR, Azevedo-Marques PM, Cordeiro SS, Araújo-Neto SA. Towards an Information Infrastructure for Medical Image Sharing. J Digit Imaging. 33:88-98, 2020. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064679/pdf/10278_2019_Article_243.pdf
6. UK-wide medical imaging sharing system expands beyond hospitals. https://www.buildingbetterhealthcare.com/news/article_page/UK-wide_medical_imaging_sharing_system_expands_beyond_hospitals/198437. Accessed October 10, 2022
7. HealtheConnections and Nuvance health expand HIE services through radiology image exchange connection. https://www.prnewswire.com/news-releases/healtheconnections-and-nuvance-health-expand-hie-services-through-radiology-image-exchange-connection-301314060.html. Accessed October 10, 2022
8. KHIE Image Exchange. https://khie.ky.gov/Resources/Pages/Whats-Happening-at-KHIE.aspx. Accessed October 10, 2022
9. Big Sky Care Connect, Montana’s premier healthcare information exchange (HIE) platform finalizes partnership with Medicom Technologies, Inc. (Medicom) to image enable their provider health information network. https://www.prnewswire.com/news-releases/big-sky-care-connect-montanas-premier-healthcare-information-exchange-hie-platform-finalizes-partnership-with-medicom-technologies-inc-medicom-to-image-enable-their-provider-health-information-network-301472467.html . Accessed October 10, 2022
10. CRISP Clinical data. https://www.crisphealth.org/applications/clinical-data/#imaging-worklist. Accessed October 10, 2022