Overview​

This release introduces a redesigned self-scheduling workflow in the Blume Patient Portal, focused on streamlining the appointment booking process and reducing patient drop-off.

Overview​

Report deletion has been replaced with Entered in Error (EIE), a standard, universally accepted report status. Final and Preliminary reports marked EIE stay in the system with their full audit trail, author, timestamps, and linked observations.

What's New​

1. User Access Control (UAC)​

What changed:​

Two new UAC permissions under Document Viewer, both enabled by default for Admins only:

• "Mark Entered in Error" --- allows marking a report as EIE

• "View All Entered in Error Reports" --- allows viewing EIE reports and undoing the EIE status to return the report to its previous state

Benefit:​

Gives organizations granular control over who can mark reports as entered in error and who can view/undo them, supporting role-based workflows.

2. Reversible Report Status​

What changed:​

Delete button renamed to "Entered in Error" with cross icon replacing trash icon. EIE reports are clearly labelled in Study Explorer as no longer clinically active. Reports can be undone at any time and return to their previous state.

Benefit:​

Protecting patient record integrity and supporting clinical compliance by retaining a full audit trail while clearly marking reports as no longer active. Reports are never truly deleted.

3. Signed Indicator and Status Tracking​

What changed:​

Final reports now display a signed indicator with a green tick. Reports in Entered in Error status are shown in the Patient Chart.

Benefit:​

Provides clear visual feedback about report status and signature state, making it easy to identify which reports are clinically active.

4. Compatibility with Critical Findings​

What changed:​

Entered in Error is compatible with Critical Findings. A CF-flagged report marked EIE keeps its flag and history (inactive) and is removed from the CF worklist; undoing EIE restores the CF in full.

Benefit:​

Ensures critical findings history is preserved while maintaining accurate worklist status. No data is lost.

Important Notes​

• Entered in Error applies to both Final and Preliminary reports.

• Only the Entered in Error status changes; all linked data (audit trail, author, timestamps, observations) is retained.

• NOT supported in MD.ai and RadPair editors --- EIE and Delete actions are hidden on reports authored through those systems.

• The restriction applies only to reports authored in MD.ai or RadPair. Reports authored natively in the platform support Entered in Error as normal, even for the same user.

Where to Find It​

• Document Viewer → Report with (Entered in Error) cross icon (visible when the UAC permission is enabled)

• Patient Chart → Report showing Entered in Error status

Overview​

Readers can flag the findings in a diagnostic report by severity from the Document Viewer toolbar and record when the concerned party has been notified/communicated (e.g., Referring Physician). Flags track the finding notification workflow with visual indicators showing Critical, Urgent, or Important priority levels.

What's New​

1. Flag Drawer with Three Severity Levels​

What changed:​

The flag icon in the top toolbar opens a drawer showing three severity levels: Critical (Red), Urgent (Orange), and Important (Yellow).

Benefit:​

Provides a standardized way to categorize and prioritize findings based on urgency and clinical significance.

2. Confirmation Message and Undo​

What changed:​

When a flag is selected, a 5-second confirmation message appears in the bottom-left corner with an Undo option. Click Undo to reverse the action with no history recorded.

Benefit:​

Provides immediate feedback and prevents accidental flag actions, allowing quick correction if needed.

3. Mark Notified Tracking​

What changed:​

Below each flag level is a " Mark Notified" option to record that the referring physician has been contacted. A green checkmark (✓) appears on the toolbar flag when marked as notified.

Availability --- Native DV: Available on Preliminary, Final, and Final Amended reports; NOT during amendments.

Availability --- MD.ai: Available on Final and Final Amended reports. Flagging an MD.ai report while a draft (Preliminary) is being added.

Benefit:​

Tracks which findings have been communicated to referring physicians, ensuring critical notifications are not missed.

4. Finding History and Action Tracking​

What changed:​

"View History" opens a side panel showing Finding Details, Action History (all completed flag actions), and Notification Status. Undone actions do not appear in history.

Benefit:​

Provides a complete audit trail of all flag actions and notification status for compliance and tracking purposes.

5. Worklist Priority and Ordering​

What changed:​

Flags appear in the worklist with color indicators. Critical flags, then Urgent, then Important studies sort first. A flag set during an amendment takes precedence for worklist ordering.

Benefit:​

Ensures highest-priority findings are always visible first, making it easy to focus on the most urgent cases.

6. Reading Physicians Can Flag​

What changed:​

Reading Physicians can now flag reports, and flags are viewable by all users.

Benefit:​

Extends finding prioritization to all reporting users, not just radiologists, ensuring comprehensive coverage.

Important Notes​

• Flags can be added while creating a report or after signing.

• Only the user who signed the report can edit it during amendment; new flags take precedence for worklist ordering.

• When a flagged report is marked Entered in Error, the flag remains but becomes inactive; the CF worklist entry is removed.

• Flags on other reports in a multi-report study still count and keep the worklist entry.

• Undoing Entered in Error restores the flag with its complete history.

• Mark Notified availability differs between Native DV and MD.ai-based on report state.

Where to Find It​

• Document Viewer → Flag icon in top toolbar

• Worklist → Flags displayed on prioritized reports

• Pending Signatures panel → View History option for audit trail

Overview​

Blume has been rebuilt to deliver a faster, more reliable experience for viewing studies, scheduling appointments, and sharing results.

The updated platform improves overall speed and reliability. The login and onboarding screens have been refreshed, so they may look different when you first open the updated app.

All existing features—including viewing studies, scheduling appointments, and sharing results—remain available, with improved performance and a smoother overall experience.

What's New​

1. One-Time Sign-In Required​

What changed:​

After installing the update, existing users will be prompted to sign in even if you were signed in before updating the app. Use your current account details; you do not need to create a new account.

Benefit:​

This one-time sign-in enables a smooth move to the rebuilt platform while keeping your account secure and unchanged.

2. Password-Based Login​

What changed:​

Blume will continue to use password-based authentication, and the existing sign-in methods remain unchanged.

Benefit:​

You can continue to sign in with your existing credentials using the same process as before. Onboarding Guide

Important Notes for Existing Users​

• Sign In vs. Sign Up: After updating Blume, sign in with your existing account. You don't need to create a new account.
• New Visual Design: The login and onboarding screens now have an updated appearance. Your existing data and settings are not affected and will remain available.
• No Feature Changes: All core functionality remains the same. This release is focused on performance and reliability improvements.

Where to Find It​

Blume 2.0 is available for download from the App Store and Google Play. Update the mobile app, to use the latest version.

Overview​

Sign Delay gives radiologists and reading physicians a brief, configurable window to undo a signed report before it is permanently finalized. Once a report is signed, it enters a Pending Signature state and a countdown begins. When the timer reaches zero, the signature is locked in.

What's New​

1. Pending Signature State After Signing​

What changed:​

Every report now enters a Pending Signature state immediately after signing— this now includes MD.ai reports, not just Native DV. A countdown timer begins using the organization's configured delay duration (default: 2 minutes), and the report remains editable by the signing user until the timer ends. The signature is not finalized until the countdown reaches zero or the user selects Sign Now.

Benefit:​

Provides a brief review window before the signature is permanently applied, allowing users to address last-minute changes without disrupting the signing workflow.

2. Pending Signatures Panel​

What changed:​

A new Pending Signatures panel is accessible via the bell (notification) icon. After signing, the report is pinned at the top of this panel with a live countdown timer and available actions for any report awaiting finalization — including MD.ai reports. Undo cancels the signature on both MD.ai and OmegaAI simultaneously. The panel is accessible from both the Worklist and the Document Viewer.

Benefit:​

Allows users to monitor and manage pending signatures from a central location across both systems, without reopening individual reports, while providing full visibility into either side.

3. Undo and Sign Now Actions​

What changed:​

Users can Undo during the sign delay period to return the report to an editable (unsigned) state. Alternatively, selecting Sign Now bypasses the remaining delay and immediately finalizes the signature. For MD.ai reports, finalization happens on MD.ai first, then OmegaAI; if the report is edited in MD.ai during the delay window, it is not signed, so the two systems never end up holding different signed content. Once Sign Now is used, Undo is no longer available. When a report is unsigned and then signed again, the countdown restarts from the full 2-minute delay period. Sign Now is also available on failed reports and immediately finalizes without queue rejoin or additional delay.

Benefit:​

Provides flexibility to make corrections before finalization, complete signing immediately when no additional review is needed, or recover quickly from signing failures.

4. Handling Signing Failures​

What changed:​

If a report fails to sign, users are immediately notified with a failure toast ("Diagnostic report sign failed. Retry signing.") and can access the failed report from the Pending Signatures panel. The failed report displays an error message: "Report could not be signed. Sign now to try again, or cancel signing." Users can select Sign Now to immediately retry the signature or Cancel Sign to revert the report to an editable state.

Benefit:​

Ensures users are always aware of signing failures and can quickly recover by retrying the signature or making corrections before signing again.

5. Amendment Handling​

What changed:​

Amendments follow the same Sign Delay flow and finalize as Amended on both MD.ai and OmegaAI.

Benefit:​

Keeps the amendment workflow consistent with standard signing, with the same review window protection.

6. Administrator Configuration​

What changed:​

Sign Delay is on by default for all organizations — existing and new. Administrators can configure the delay duration or disable the feature via Organization Settings. MD.ai reports use this same setting, so no separate configuration is needed. The valid range is 0.5 to 15 minutes in 0.5-minute increments.

Benefit:​

Gives organizations full control over the delay duration or the ability to opt out of the feature entirely.

To configure Sign Delay:

1. From the OmegaAI home screen, select your organization.
2. Go to Details — a panel opens on the left.
3. Scroll down to Workflow Steps and click to open it.
4. On the toolbar at the top right corner, click the Settings button.
5. Locate Sign Delay and toggle it on or off to enable or disable the feature.
6. Adjust the delay duration as needed. Valid range: 0.5 – 15 minutes in 0.5-minute increments.
7. Save your changes before navigating away.

Important Notes​

• Sign Delay is enabled by default for all organizations, including existing and new users. No opt-in or configuration required. Every report enters the sign delay queue when signed — there are no exceptions.

• Only the user who signed the report can undo or edit it while in a Pending Signature state.Other users cannot modify the report during the delay period.

• Once the countdown reaches zero, the signature is permanently finalized.

• The default sign delay duration is 2 minutes. Delay values can be configured between 0.5 and 15 minutes in 0.5-minute increments.

• A value of 0 is not permitted as a delay duration. To disable Sign Delay, administrators must turn the feature off via the toggle.

• Re-signing after an Undo restarts the full 2-minute countdown from scratch. Selecting Sign Now permanently finalizes the report and removes the Undo option.

• If a report fails to sign, users are notified immediately and can retry with Sign Now or revert with Cancel Sign from the Pending Signatures panel.

• Administrators control feature enablement and duration through Organization Settings.

  MD.ai Compatibility:

• Existing Users: Sign Delay will be active immediately — no action required. After signing any report, users will see the Pending Signatures panel via the bell (notification) icon along with a live countdown banner at the top of the report viewport.

• New Users: Sign Delay will be active as part of the default organization setup.

• No new admin setting required for MD.ai — uses existing Organization Settings → Sign Delay duration.

• Radpair reports still skip Sign Delay entirely.

Where to Find It​

Administrators: OmegaAI Home → Select Organization → Details → Workflow Steps → Settings → Sign Delay toggle

End users: Bell (notification) icon → Pending Signatures panel

Overview​

Hanging Protocols in the OmegaAI Image Viewer now go well beyond flat 2D stacks. This release adds MPR, rotating PET MIP, and PET-CT fusion viewports; smarter, metadata-driven series matching; and four new default protocols built around them.

What's New​

1. MPR, Rotating PET MIP & PET-CT Fusion Viewports​

What changed:​

Hanging Protocol viewports are no longer limited to a flat 2D stack. Seven new view codes let a viewport open directly into an advanced view as soon as a study loads, with no manual setup:

• CT: MPR Axial / Coronal / Sagittal — builds a multiplanar reconstruction from the thinnest axial CT series in the chosen plane. Even an axial-only CT can open as coronal or sagittal.
• PT: PET 3D MIP — a rotating whole-body PET projection, built automatically from the attenuation-corrected PET series.
• PET-CT: Fusion Axial / Sagittal / Coronal — PET overlaid on CT in the chosen plane, using the PET display preset's colormap.

Slab thickness and projection (Maximum, Average, or Minimum Intensity) can be saved for each MPR and fusion viewport, so the viewport opens exactly as configured. If a study can't support one of these views, the viewport shows the best available 2D series instead of going blank — for fusion, it shows CT alone with a "Fusion unavailable" indicator.

Benefit:​

Radiologists land on the right multiplanar or fused view automatically — no manual reconstruction or series hunting required.

2. Display Presets for Every Modality​

What changed:​

Each viewport's rules now include a single Display Preset picker, listing the presets available for that viewport's modality — CT, MR, US, PT, NM, and more, not just PET as before. Selecting a preset applies its window level and, where configured, its colormap and auto-windowing behavior together.

Benefit:​

Consistent, clinically correct windowing and colormaps on open, authored once and reused across every protocol.

3. Smarter, Metadata-Driven Series Matching​

What changed:​

Viewport matching rules can now weigh study details beyond modality and body part — slice thickness, study description, and dozens of other fields — with numeric and multi-value comparisons. Each rule can be a strict requirement or just a preference, so OmegaAI still hangs the best available series rather than leaving a viewport empty.

Benefit:​

Protocols can express real clinical preferences — like favoring a primary-read series over a thin reconstruction — without ever risking a blank viewport.

4. Four New Default Hanging Protocols for PET, PET-CT, and MPR​

What changed:​

OmegaAI's Default Hanging Protocol catalog adds four new protocols: PET Default, PET-CT Default, PET-CT Quick Read, and MPR Default, each built around the new viewport types above. PET and PET-CT cells apply a clinically standard SUV-scaled colormap, and fall back safely with a clear indicator when a study's data can't support SUV scaling.

Benefit:​

PET, PET-CT, and MPR-heavy workflows get a clinically reviewed, ready-to-use starting protocol out of the box.

Important Notes​

• Backward compatible: Existing protocols behave exactly as before, with no impact on studies that don't use the new viewport types.
• Series matching: A condition on a metadata field that doesn't apply to the bound series simply never matches — it doesn't cause an error.
• SUV safety fallback: A "SUV unavailable" indicator with hidden SUV stats is expected behavior when a study's DICOM metadata can't support SUV calculation — it isn't an error.

Where to Find It​

• MPR, PET 3D MIP & Fusion view codes: Hanging Protocol Configuration → View Codes panel → drag CT: MPR Axial/Coronal/Sagittal, PT: PET 3D MIP, or PET-CT: Fusion Axial/Sagittal/Coronal onto a viewport
• Display Presets & series matching: Hanging Protocol Configuration → select a viewport → HP Rules tab
• New Default protocols: Image Viewer → Quick Access Strip or Layout Panel → PET Default / PET-CT Default / PET-CT Quick Read / MPR Default

Overview​

PET/CT SUV measurement gets a dedicated Volumetric SUV tool with a full-view region overlay and adjustable thresholds, plus a fix so every ROI tool reports SUV correctly on fused viewports.

What's New​

1. Volumetric SUV Becomes Its Own Tool​

What changed:​

The Probe tool is back to reporting a plain point SUV value on PET data, just as it does on any other modality. Lesion segmentation now lives in its own Volumetric SUV tool, available from the toolbar, the Image Control Wheel, or a custom hotkey. Clicking a lesion automatically segments the region, reports SUV max/mean/peak and volume, overlays the boundary across every plane and the whole-body MIP, and lets you adjust the segmentation threshold or SUV normalization on the fly.

Benefit:​

A quick point SUV check with Probe, and a dedicated tool with visual confirmation and adjustable thresholds when you need volumetric measurement.

2. PET/CT Fusion ROI Tools Now Report SUV, Not HU​

What changed:​

Rectangle, Elliptical, Circle, and Freehand ROI tools drawn on a fused PET/CT viewport now report SUV statistics instead of Hounsfield Units, matching what the same tool reports on a PET-only viewport.

Benefit:​

SUV measurement is now consistent across every ROI tool in Fusion Mode — no more switching viewports just to get a valid SUV reading.

Important Notes​

• Volumetric SUV Tool: A newer capability that may not yet be enabled in every environment — reach out to your OmegaAI representative if it doesn't appear in your toolbar. Measurements created by the Probe tool before this change remain labeled as Probe entries.
• SUV safety fallback: A "SUV unavailable" indicator with hidden SUV stats is expected behavior when a study's DICOM metadata can't support SUV calculation — it isn't an error.

Where to Find It​

• Volumetric SUV Tool: Image Viewer → Probe toolbar group (Fusion Mode) → Volumetric SUV; also available via the Image Control Wheel and Hotkey Customization
• Fusion ROI SUV: PET/CT Fusion Mode → draw any ROI tool on the fused viewport

Overview​

Window Level presets in the OmegaAI Image Viewer are now tailored to the modality you're actually viewing, replacing the old CT-centric preset list that used to appear on every study type regardless of modality.

What's New​

1. Presets Filtered by Modality​

What changed:​

The Window Level preset list is now filtered to the modality being viewed — CT, MR, US, PT, NM, and more each get their own relevant list, instead of a single generic CT-centric list showing up everywhere.

Benefit:​

Every modality gets presets that are actually clinically relevant to it, so you're never scrolling past irrelevant CT presets on an MR or ultrasound study.

2. Corrected CT Presets & Auto-Windowing for Other Modalities​

What changed:​

CT presets like Brain, Stroke, Mediastinum, Soft Tissue, and Abdomen now use corrected clinical values, with a new Subdural preset added. Modalities where a fixed window doesn't make sense — MR, US, CR/DX, and NM — now default to auto-windowing strategies instead, calculated from the image data itself.

Benefit:​

Presets reflect current clinical standards, and modalities that never suited fixed numeric values now window correctly by default.

3. Compound Presets for PET & Nuclear Medicine, Mammography Support​

What changed:​

PET and Nuclear Medicine presets now apply window level and color map together in a single selection (Hot Iron, Hot Metal Blue, PET Rainbow, Cool, Inverted Gray), with correct SUV-unit windowing. Mammography gains sigmoid VOI LUT support, with correct handling of "For Presentation" vs. "For Processing" images.

Benefit:​

One click gets you the correct look for PET/NM studies, and mammography windowing now matches how those images are meant to be displayed.

4. Three-Tier Preset Configuration​

What changed:​

Presets now resolve through three tiers — System defaults, Organization overrides, and User overrides — so administrators and individual radiologists can customize presets without losing the underlying system defaults.

Benefit:​

Organizations and individual readers can fine-tune presets to their own preferences without breaking the baseline for everyone else.

Important Notes​

• Backward compatible: All original CT preset names are preserved, so existing Hanging Protocols referencing them by name continue to work unchanged.

Where to Find It​

• Window Level Presets: Image Viewer → Window Level tool → preset dropdown (filtered to the active viewport's modality)

Overview​

The Indicators column on the worklist now shows whether a study's patient has a Blume Patient Portal account, so you no longer need to open the patient record to check.

Overview​

OmegaAI now enables DICOM Stations to be marked as authoritative sources of study identity. When enabled, a study created from an HL7 order that receives images from an authoritative station will use the station's Study Instance UID (0020,000D) as the canonical identifier, replacing OmegaAI's placeholder UID. This resolves bidirectional workflow issues where annotations and reports sent back to source systems could not be matched due to UID mismatches.

What's New​

1. Authoritative DICOM Station Setting​

What changed:​

• New toggle added to DICOM Station device configuration (Organization > Device > DICOM Station Features > Authoritative). Defaults to OFF for all new stations.

Benefit:​

• Administrators can explicitly trust specific PACS and gateways as authoritative sources of study identity, enabling automatic reconciliation of annotations and reports sent back to the source system.

2. Study Instance UID Preservation and Coercion Warnings​

What changed:​

• When Authoritative is enabled and a source UID differs from OmegaAI's order-generated placeholder, the source UID becomes canonical. The placeholder UID is retained in a merged-UID reference table. A C-STORE Warning (0xB000) is returned to the sending system, and the object records an Original Attributes Sequence (0400,0561) with the original UID, modification datetime, modifying system, and COERCE reason code.

Benefit:​

• UID changes are now transparent and auditable. All previous UIDs remain discoverable through QIDO-RS, WADO-RS, C-FIND, and C-MOVE. Existing external references (worklist IDs, prior queries, links in reports) that stored the old UID continue to work.

3. Per-Station Trust Configuration​

What changed:​

• Authoritative setting is configured per DICOM Station device, not globally. Each station is independently marked as trusted or untrusted.

Benefit:​

• A single well-behaved gateway can be marked as authoritative without changing how any other station is treated. Gives administrators fine-grained control over which sources are trusted.

Important Notes​

• Default Behavior: Authoritative is OFF by default for all DICOM Station devices, including newly created ones. No existing configurations change behavior on upgrade.

• When This Setting Takes Effect: The setting only takes effect when a study was created by an HL7 order (placeholder UID) and the DICOM object then matches it by accession. Where studies are created by the DICOM object itself, it never comes into play.

• Bidirectional Workflows Only: Enable Authoritative only for sources where you expect to send annotations, reports, or other derived content back to the source system and require automatic UID-based matching.

• DICOM Station vs Connection Setting: This is a DICOM Station setting, not a connection setting. Ingest reads the receiving station by AE Title, so each station is ticked separately.

• UI Is Not a Safety Boundary: A FHIR PUT /Device sets the value whether or not the toggle switch is visible in the UI.

• Supported Workflows: Re-sending the same study under a different UID is supported. Every superseded UID is kept as an alias, allowing external systems holding an older UID to resolve via the alias.

• Not Recommended For:

  • Sources with UIDs that aren't globally unique (reusing one UID across different studies mis-identifies studies)
  • Unidirectional workflows where there's no risk, just no benefit

• Testing Required: Before enabling Authoritative for a production PACS or gateway, test the bidirectional workflow (send and return annotations) in a non-production environment to verify correct reconciliation at the source.

• Configuration Examples:

  • CSH Compass PACS (Bidirectional Workflow): Navigate to Organization > Device > CSH Compass PACS > DICOM Station Features > Toggle Authoritative ON. Compass UIDs are now canonical; returned annotations carry the original UID and match automatically at CSH.

  • Ultrasound Modality (Unidirectional Workflow): Leave Authoritative OFF. OmegaAI uses the order-generated UID. If the ultrasound re-sends the same exam with a different UID, both sets of objects are stored under the order UID with no UID conflicts.

• DICOM Compliance:

  • C-STORE Warning (0xB000): Returned when Study Instance UID is coerced (per DICOM PS3.4 §B.4)

  • Original Attributes Sequence (0400,0561): Object records original UID, modification datetime, modifying system, and COERCE reason (per DICOM PS3.3 §C.12.1.1.9)

  • Benefits

    1. Bidirectional Reconciliation — Annotations and reports match at the source system without manual intervention or worklist reconciliation.

    2. No Data Loss — All superseded UIDs are retained in the merged-UID reference table and remain queryable and accessible through QIDO-RS, WADO-RS, C-FIND, and C-MOVE.

    3. Transparency and Auditability — C-STORE Warning (0xB000) is returned to the sending system and Original Attributes Sequence (0400,0561) records the original UID and reason code.

    4. Per-Station Configuration — Each DICOM Station is independently marked as trusted or untrusted. One gateway can be authoritative without affecting others.

    5. Safe by Default — OFF by default for all stations. No existing configurations change behavior on upgrade.

Where to Find It​

Location in OmegaAI:

Organization
  └─ Device
      └─ [Select DICOM Station Device]
          └─ Device Details Form
              └─ DICOM Station Features
                  └─ Authoritative ← Toggle is here (last item)

Available On: DICOM Station devices only (not on OmegaAI Link or API devices)

Permission: Read-only when the Device form is in view mode; editable in edit mode, the same as every other field in the DICOM Station Features group. No separate permission applies to this field.

Programmatic access: Can be queried and set through the FHIR Device API (device.extension[where(url='authoritative')].valueBoolean)

Updated Help Articles​

Preserve Study Instance UIDs from Trusted DICOM Sources