DICOM Basics using .NET and C# - Understanding DICOM Structured Reports

Nov 29, 2025

Introduction

This is part of my series of articles on the DICOM standard. In this tutorial, we'll explore DICOM Structured Reports (SR), which provide a standardized way to encode clinical findings, measurements, and observations in a structured, machine-readable format.

Unlike traditional text reports, Structured Reports encode information using coded concepts and a hierarchical tree structure. This enables automated processing, data mining, and integration with clinical decision support systems.

Prerequisites

Before you begin, ensure you have the following:

A .NET development environment (Visual Studio or Visual Studio Code)
The Fellow Oak DICOM library (fo-dicom) installed via NuGet
Basic understanding of DICOM concepts from previous tutorials
You can find all the code demonstrated in this tutorial on GitHub here

“Time is the substance from which I am made. Time is a river which carries me along, but I am the river.” ~ Jorge Luis Borges

The Theory Behind Structured Reports

Traditional radiology reports are narrative text: prose written by radiologists describing their findings. While human-readable, narrative reports are semantically opaque to computers. You cannot easily extract that a lesion measures 2.3cm or that a finding is in the right upper lobe without natural language processing. Structured Reports solve this by encoding clinical content as discrete, machine-readable data elements.

The SR content tree model represents clinical observations as a hierarchy of typed nodes. Each node has a relationship to its parent (CONTAINS, HAS PROPERTIES, etc.), a value type (TEXT, CODE, NUM, IMAGE), and a concept name indicating what the node represents. This creates a semantic graph where meaning is explicit rather than implied. A finding doesn't just contain the text "2.3cm" - it explicitly encodes a numeric measurement with units, linked to the anatomical location it measures.

The use of coded concepts from terminologies like SNOMED CT, LOINC, and RadLex enables semantic interoperability. When a radiologist reports "hepatocellular carcinoma," the code 25370001 from SNOMED CT conveys the same meaning regardless of language, local terminology variations, or abbreviations. This codification enables decision support systems to recognize findings, aggregate data for research, and support quality metrics.

The IMAGE reference type creates direct links between observations and evidence. When a measurement refers to a specific image, the SR can encode the exact SOP Instance UID and spatial coordinates. This traceability means you can navigate from a finding directly to the image that supports it, essential for verification and follow-up comparison.

SR templates (defined in DICOM Part 16) provide document patterns for specific use cases. A CAD SR for mammography follows a different structure than a dose report. Templates constrain what content items are required, optional, or prohibited, ensuring that SR documents from different systems are structurally comparable. This is the foundation for meaningful data exchange and analytics.

Understanding Structured Reports

DICOM Structured Reports encode clinical content as a tree of content items. Each content item has:

Value Type: The type of content (CODE, TEXT, NUM, IMAGE, etc.)
Relationship Type: How this item relates to its parent (CONTAINS, HAS PROPERTIES, etc.)
Concept Name: What this item represents (coded concept)
Value: The actual content based on the value type

SR Content Types

Structured Reports support various content types:

Value Type	Contains	Example
TEXT	Free-text description	”No acute findings”
CODE	Coded concept	Diagnostic code from SNOMED
NUM	Numeric measurement with units	”5.2 cm”
IMAGE	Reference to DICOM image	Image SOP Instance UID
UIDREF	Reference by UID	Related SR or other object
CONTAINER	Group of related items	”Findings” section

Common SR SOP Classes

SOP Class	UID	Use Case
Basic Text SR	1.2.840.10008.5.1.4.1.1.88.11	Simple text reports
Enhanced SR	1.2.840.10008.5.1.4.1.1.88.22	More structure and codes
Comprehensive SR	1.2.840.10008.5.1.4.1.1.88.33	Full SR capabilities
X-Ray Radiation Dose SR	1.2.840.10008.5.1.4.1.1.88.67	Dose reporting
Mammography CAD SR	1.2.840.10008.5.1.4.1.1.88.50	CAD results

Step 1 of 3: Creating a Simple Structured Report

Let's create a Basic Text Structured Report:

using System;
using System.Diagnostics;
using System.IO;
using FellowOakDicom;

namespace StructuredReportsExample
{
    public class Program
    {
        private static readonly string OutputPath =
            Path.Combine(AppDomain.CurrentDomain.BaseDirectory, "Output");
        private static readonly string OutputSrFile =
            Path.Combine(OutputPath, "sample_structured_report.dcm");

        public static void Main(string[] args)
        {
            try
            {
                LogToDebugConsole("=== DICOM Structured Reports Tutorial ===");

                // Ensure output directory exists
                if (!Directory.Exists(OutputPath))
                {
                    Directory.CreateDirectory(OutputPath);
                }

                // Create and save a sample SR
                CreateSampleStructuredReport();

                // Read and display the created SR
                ReadStructuredReport(OutputSrFile);

                LogToDebugConsole("Structured Report tutorial completed.");
            }
            catch (Exception e)
            {
                LogToDebugConsole($"Error: {e.Message}");
            }
        }

        private static void LogToDebugConsole(string message)
        {
            Debug.WriteLine(message);
        }
    }
}

Step 2 of 3: Building the SR Dataset

Now let's build the complete Structured Report with all required modules:

private static void CreateSampleStructuredReport()
{
    LogToDebugConsole("--- Creating Sample Structured Report ---");

    var dataset = new DicomDataset();

    // Patient Module
    dataset.Add(DicomTag.PatientName, "Doe^John");
    dataset.Add(DicomTag.PatientID, "SR123456");
    dataset.Add(DicomTag.PatientBirthDate, "19800101");
    dataset.Add(DicomTag.PatientSex, "M");

    // Study Module
    var studyUid = DicomUID.Generate();
    dataset.Add(DicomTag.StudyInstanceUID, studyUid);
    dataset.Add(DicomTag.StudyDate, DateTime.Now.ToString("yyyyMMdd"));
    dataset.Add(DicomTag.StudyTime, DateTime.Now.ToString("HHmmss"));
    dataset.Add(DicomTag.AccessionNumber, "SR-ACC-001");
    dataset.Add(DicomTag.StudyDescription, "Chest CT Findings Report");

    // Series Module
    var seriesUid = DicomUID.Generate();
    dataset.Add(DicomTag.SeriesInstanceUID, seriesUid);
    dataset.Add(DicomTag.SeriesNumber, "1");
    dataset.Add(DicomTag.Modality, "SR");  // Structured Report modality
    dataset.Add(DicomTag.SeriesDescription, "CT Findings SR");

    // SOP Common Module - Basic Text SR
    var sopUid = DicomUID.Generate();
    dataset.Add(DicomTag.SOPClassUID, DicomUID.BasicTextSRStorage);
    dataset.Add(DicomTag.SOPInstanceUID, sopUid);
    dataset.Add(DicomTag.InstanceNumber, "1");

    // SR Document General Module
    dataset.Add(DicomTag.InstanceCreationDate, DateTime.Now.ToString("yyyyMMdd"));
    dataset.Add(DicomTag.InstanceCreationTime, DateTime.Now.ToString("HHmmss"));
    dataset.Add(DicomTag.ContentDate, DateTime.Now.ToString("yyyyMMdd"));
    dataset.Add(DicomTag.ContentTime, DateTime.Now.ToString("HHmmss"));

    // Completion Flag: COMPLETE or PARTIAL
    dataset.Add(DicomTag.CompletionFlag, "COMPLETE");
    // Verification Flag: VERIFIED or UNVERIFIED
    dataset.Add(DicomTag.VerificationFlag, "UNVERIFIED");

    // SR Document Content Module - Content Sequence
    var contentSequence = new DicomSequence(DicomTag.ContentSequence);

    // Add a TEXT content item with findings
    var findingsItem = new DicomDataset();
    findingsItem.Add(DicomTag.RelationshipType, "CONTAINS");
    findingsItem.Add(DicomTag.ValueType, "TEXT");

    // Concept Name Code Sequence - What this item represents
    var conceptNameSeq = new DicomSequence(DicomTag.ConceptNameCodeSequence);
    var conceptName = new DicomDataset();
    conceptName.Add(DicomTag.CodeValue, "121071");
    conceptName.Add(DicomTag.CodingSchemeDesignator, "DCM");
    conceptName.Add(DicomTag.CodeMeaning, "Finding");
    conceptNameSeq.Items.Add(conceptName);
    findingsItem.Add(conceptNameSeq);

    // The actual text value
    findingsItem.Add(DicomTag.TextValue,
        "No acute cardiopulmonary abnormality. " +
        "Clear lung fields bilaterally. Heart size within normal limits.");

    contentSequence.Items.Add(findingsItem);
    dataset.Add(contentSequence);

    // Save the Structured Report
    var dicomFile = new DicomFile(dataset);
    dicomFile.Save(OutputSrFile);

    LogToDebugConsole($"Created SR file: {OutputSrFile}");
}

Step 3 of 3: Reading Structured Reports

Now let's read and navigate the SR content tree:

private static void ReadStructuredReport(string filePath)
{
    LogToDebugConsole("--- Reading Structured Report ---");

    var file = DicomFile.Open(filePath);
    var dataset = file.Dataset;

    LogToDebugConsole($"  SOP Class: {dataset.GetSingleValueOrDefault(DicomTag.SOPClassUID, "")}");
    LogToDebugConsole($"  Modality:  {dataset.GetSingleValueOrDefault(DicomTag.Modality, "")}");
    LogToDebugConsole($"  Patient:   {dataset.GetSingleValueOrDefault(DicomTag.PatientName, "")}");
    LogToDebugConsole($"  Study:     {dataset.GetSingleValueOrDefault(DicomTag.StudyDescription, "")}");
    LogToDebugConsole($"  Status:    {dataset.GetSingleValueOrDefault(DicomTag.CompletionFlag, "")}");

    // Navigate Content Sequence
    var contentSequence = dataset.GetSequence(DicomTag.ContentSequence);
    if (contentSequence != null)
    {
        LogToDebugConsole("  Content Items:");
        foreach (var item in contentSequence.Items)
        {
            var valueType = item.GetSingleValueOrDefault(DicomTag.ValueType, "");
            var relationshipType = item.GetSingleValueOrDefault(DicomTag.RelationshipType, "");

            LogToDebugConsole($"    - Type: {valueType}, Relationship: {relationshipType}");

            // Get concept name if present
            var conceptNameSeq = item.GetSequence(DicomTag.ConceptNameCodeSequence);
            if (conceptNameSeq != null && conceptNameSeq.Items.Count > 0)
            {
                var codeMeaning = conceptNameSeq.Items[0].GetSingleValueOrDefault(
                    DicomTag.CodeMeaning, "");
                LogToDebugConsole($"      Concept: {codeMeaning}");
            }

            // Get value based on type
            if (valueType == "TEXT")
            {
                var textValue = item.GetSingleValueOrDefault(DicomTag.TextValue, "");
                LogToDebugConsole($"      Text: {textValue}");
            }
            else if (valueType == "NUM")
            {
                var numValue = item.GetSingleValueOrDefault(DicomTag.NumericValue, "");
                LogToDebugConsole($"      Numeric Value: {numValue}");
            }
            else if (valueType == "CODE")
            {
                var codeSeq = item.GetSequence(DicomTag.ConceptCodeSequence);
                if (codeSeq != null && codeSeq.Items.Count > 0)
                {
                    var code = codeSeq.Items[0].GetSingleValueOrDefault(DicomTag.CodeMeaning, "");
                    LogToDebugConsole($"      Code: {code}");
                }
            }
        }
    }
}

Sample output:

=== DICOM Structured Reports Tutorial ===

--- Creating Sample Structured Report ---
Created SR file: C:\...\Output\sample_structured_report.dcm

--- Reading Structured Report ---
  SOP Class: 1.2.840.10008.5.1.4.1.1.88.11
  Modality:  SR
  Patient:   Doe^John
  Study:     Chest CT Findings Report
  Status:    COMPLETE
  Content Items:
    - Type: TEXT, Relationship: CONTAINS
      Concept: Finding
      Text: No acute cardiopulmonary abnormality. Clear lung fields bilaterally. Heart size within normal limits.

Structured Report tutorial completed.

Common Use Cases for Structured Reports

Radiation Dose Reporting: Automatic capture of CT dose information
CAD Results: Computer-aided detection findings for mammography
Key Image Notes: Annotations and significant findings
Measurements: Tumor measurements over time
Clinical Findings: Structured radiology reports

Best Practices

Use standard coded concepts from DCM, SNOMED-CT, or LOINC
Include image references when findings relate to specific images
Follow IHE profiles for specific use cases (e.g., dose reporting)
Use Comprehensive SR for complex reports with multiple content types
Validate SR structure against the DICOM standard

Conclusion

DICOM Structured Reports provide a powerful way to encode clinical findings in a standardized, machine-readable format. Unlike free-text reports, SR content can be processed automatically, enabling analytics, clinical decision support, and quality reporting.

While creating full-featured Structured Reports can be complex, understanding the basic structure and content types provides a foundation for working with SR in medical imaging applications. Many modern use cases, such as radiation dose tracking and CAD integration, rely heavily on Structured Reports.

Please check out the next tutorial in this series where we cover DICOM Print operations.