Difference between revisions of "AI Results"

From IHE Wiki
Jump to: navigation, search
(See Also)
(Details)
 
(11 intermediate revisions by the same user not shown)
Line 5: Line 5:
  
 
==Summary==
 
==Summary==
This profile establishes baseline data handling and presentation capabilities for an image display product to be “AI-Ready”. Result generation products can leverage these data formats to be compatible with a variety of displays and site workflows.
+
This profile standardizes storage & display of AI Results for image interpretation.  It establishes baseline data handling and presentation capabilities for an image display product to be “AI-Ready”. Result generation products can leverage these data formats to be compatible with a variety of displays and site workflows.
  
 
AIR defines:
 
AIR defines:
* content formats for encoding data
+
* a set of '''result “primitives”'''
 +
* '''Encoding requirements''' for each primitive (DICOM-based storage)
 
* transactions for moving that content around
 
* transactions for moving that content around
* behaviors for basic handling and display of the content
+
* '''Baseline display requirements''' for each primitive
  
 
==Benefits==
 
==Benefits==
* '''Interoperable Results''':
+
'''Interoperable Results''':
:* Results need to be presented to radiologists in their reading environment. This profile provides interoperability between result generation products and radiology reading workstation systems/software.
+
* Results need to be presented to radiologists in their reading environment. This profile provides interoperability between result generation products and radiology reading workstation systems/software.
* '''Study Integrated Results''':
+
:* Radiologists expect AI-generated results to be presented in the context of the study to which they apply and expect them to supplement (rather than replace) traditional image analysis results and thus a given study will be composed of acquired images, AI results, and traditional clinical data. The result objects defined in this profile are defined for the existing imaging data storage infrastructure.
+
* '''Effective Presentation''':
+
:* Effective use of results hinges on presenting them in conjunction with the associated images during the busy process of reading the study.
+
* '''Convergence of Result Encoding''':
+
:* Many AI results are results that a human could otherwise have produced, and those human results may be used as training data for the AI. Analysis results might also be used by other AIs, e.g., in a GAN (generative adversarial network). AI and non-AI results need to be handled together. Convergent encoding of results facilitates this, as well as data pooling and sharing between sites.
+
* '''Display Primitives''':
+
:* It is unrealistic to expect radiology displays to implement specific display capabilities for each of the myriad of algorithms being developed. To minimize implementation complexity for displays, and avoid needing different software for each new analysis result, compose analysis results from a reasonable set of primitives.
+
  
==Details==
+
'''Study Integrated Results''':
 +
* Radiologists expect AI-generated results to be presented in the context of the study to which they apply and expect them to supplement (rather than replace) traditional image analysis results and thus a given study will be composed of acquired images, AI results, and traditional clinical data. The result objects defined in this profile are defined for the existing imaging data storage infrastructure.
  
''<A few paragraphs, if appropriate, providing more details (mostly in user-speak, not tech-speak) on what the profile does and how it works.>''
+
'''Effective Presentation''':
 +
* Effective use of results hinges on presenting them in conjunction with the associated images during the busy process of reading the study.
 +
 
 +
'''Convergence of Result Encoding''':
 +
* Many AI results are results that a human could otherwise have produced, and those human results may be used as training data for the AI. Analysis results might also be used by other AIs, e.g., in a GAN (generative adversarial network). AI and non-AI results need to be handled together. Convergent encoding of results facilitates this, as well as data pooling and sharing between sites.
 +
 
 +
'''Display Primitives''':
 +
* It is unrealistic to expect radiology displays to implement specific display capabilities for each of the myriad of algorithms being developed. To minimize implementation complexity for displays, and avoid needing different software for each new analysis result, compose analysis results from a reasonable set of primitives.
 +
 
 +
==Details==
  
''<If the user might be familiar with the mechanisms used by the profile, you can mention them here. E.g. Evidence Documents is based on DICOM Structured Report (SR) Templates.>''
+
AI Result Primitives:
 +
* '''Qualitative Findings''': E.g. Breast composition is Extremely Dense, Pneumothorax is Present
 +
* '''Measurements''': E.g. Adrenal Gland Nodule Length = 52mm
 +
* '''Locations''': E.g. Blood clot at (123.9, 57.2, 752.3)
 +
* '''Segmentations''': E.g. Lung Nodule surface
 +
* '''Parametric Maps''': E.g. Saliency Map for pneumonia
 +
* '''Tracking IDs''': E.g. Primary tumor for patient 32457 (longitudinal)
 +
* '''Image References''': E.g. Best view of brain hemorrhage
  
''<If the user might have an appreciation for the problems addressed in the profile, you can mention them here, but keep it short.  E.g. Mapping HL7 Order fields to DICOM Modality Worklist attributes can be inconsistent in the marketplace, so Scheduled Workflow provides vendors with more detailed instructions.>''
 
 
 
==Systems Affected==
 
==Systems Affected==
* ''PACS systems may store, manage, and/or display Evidence Documents.''
+
* '''Evidence Creator''' - encodes results as specified
* ''Display systems may query, retrieve and display Evidence Documents.''
+
:* May an analysis or AI software package itself, or it may be a proxy, gateway, or partner system that encodes and transmits results on behalf of such software. The profile does not distinguish between analysis executed locally, in the cloud, hosted in processing servers, or running in standalone workstations.
* ''Reporting workstations may retrieve, process and include details from Evidence Documents in reports
+
* '''Image Manager (PACS)''' - stores and manages AI result objects as part of study
 +
* '''Image Display''' - retrieves results and displays as specified
 +
* '''Imaging Document Consumer''' - uses results in ways other than displaying them
 +
:* May include decision support systems, clinical databases, and report creators. 
  
 
'''Actors & Transactions:'''
 
'''Actors & Transactions:'''
  
''<Insert an actor-transaction diagram, and or list of Content Definitions>''
+
[[Image:AIR Actor Diagram.png|500px]]
  
 
==Specification==
 
==Specification==
Line 61: Line 72:
 
* '''AI Workflow for Imaging''' [AIW-I] can be used to coordinate and manage the execution of AI algorithms whose results can be encoded and displayed using AIR.
 
* '''AI Workflow for Imaging''' [AIW-I] can be used to coordinate and manage the execution of AI algorithms whose results can be encoded and displayed using AIR.
  
 
[[Category:Profiles]]
 
  
 
This page is based on the [[Profile Overview Template]]
 
This page is based on the [[Profile Overview Template]]
  
 +
[[Category:Profiles]]
 
[[Category:RAD Profile]]
 
[[Category:RAD Profile]]
 
[[Category:DICOM]]
 
[[Category:DICOM]]
 
[[Category:DICOMweb]]
 
[[Category:DICOMweb]]

Latest revision as of 18:20, 9 December 2020

AI Results specifies how medical imaging analysis results are reliably stored, retrieved, and displayed. The central use case involves results generated by artificial intelligence (AI Model) algorithms.

Summary

This profile standardizes storage & display of AI Results for image interpretation. It establishes baseline data handling and presentation capabilities for an image display product to be “AI-Ready”. Result generation products can leverage these data formats to be compatible with a variety of displays and site workflows.

AIR defines:

  • a set of result “primitives”
  • Encoding requirements for each primitive (DICOM-based storage)
  • transactions for moving that content around
  • Baseline display requirements for each primitive

Benefits

Interoperable Results:

  • Results need to be presented to radiologists in their reading environment. This profile provides interoperability between result generation products and radiology reading workstation systems/software.

Study Integrated Results:

  • Radiologists expect AI-generated results to be presented in the context of the study to which they apply and expect them to supplement (rather than replace) traditional image analysis results and thus a given study will be composed of acquired images, AI results, and traditional clinical data. The result objects defined in this profile are defined for the existing imaging data storage infrastructure.

Effective Presentation:

  • Effective use of results hinges on presenting them in conjunction with the associated images during the busy process of reading the study.

Convergence of Result Encoding:

  • Many AI results are results that a human could otherwise have produced, and those human results may be used as training data for the AI. Analysis results might also be used by other AIs, e.g., in a GAN (generative adversarial network). AI and non-AI results need to be handled together. Convergent encoding of results facilitates this, as well as data pooling and sharing between sites.

Display Primitives:

  • It is unrealistic to expect radiology displays to implement specific display capabilities for each of the myriad of algorithms being developed. To minimize implementation complexity for displays, and avoid needing different software for each new analysis result, compose analysis results from a reasonable set of primitives.

Details

AI Result Primitives:

  • Qualitative Findings: E.g. Breast composition is Extremely Dense, Pneumothorax is Present
  • Measurements: E.g. Adrenal Gland Nodule Length = 52mm
  • Locations: E.g. Blood clot at (123.9, 57.2, 752.3)
  • Segmentations: E.g. Lung Nodule surface
  • Parametric Maps: E.g. Saliency Map for pneumonia
  • Tracking IDs: E.g. Primary tumor for patient 32457 (longitudinal)
  • Image References: E.g. Best view of brain hemorrhage

Systems Affected

  • Evidence Creator - encodes results as specified
  • May an analysis or AI software package itself, or it may be a proxy, gateway, or partner system that encodes and transmits results on behalf of such software. The profile does not distinguish between analysis executed locally, in the cloud, hosted in processing servers, or running in standalone workstations.
  • Image Manager (PACS) - stores and manages AI result objects as part of study
  • Image Display - retrieves results and displays as specified
  • Imaging Document Consumer - uses results in ways other than displaying them
  • May include decision support systems, clinical databases, and report creators.

Actors & Transactions:

AIR Actor Diagram.png

Specification

Profile Status: Trial Implementation

Documents: AI Results Profile - TI

  • For an overview of the scope and goals, see "Introduction to this Supplement", and Volume 3 Section 6.5.1.
  • For an overview of the participating actors and the transactions between them, see Volume 1 Section 49.1 and Section 49.4.2.
  • For key concepts that affect implementation and deployment and may be helpful in understanding the rest of the profile, see Volume 1 Section 49.4.1.
  • For the key technical requirements on encoding and displaying analysis results, see Volume 3 Section 6.5.3 and Volume 2 Section 4.136.4.1.3 respectively.


Underlying Standards:

See Also

Related Profiles

  • AI Workflow for Imaging [AIW-I] can be used to coordinate and manage the execution of AI algorithms whose results can be encoded and displayed using AIR.


This page is based on the Profile Overview Template