请求协助：基于Fo-Dicom实现DICOM头增删改查及像素数据存储（C#桌面端）

Hey there! Let's tackle your Fo-Dicom needs step by step—whether you're modifying DICOM headers or embedding pixel data, I've got practical code snippets and explanations to help you out in your C# desktop app.

Prerequisites #

First, make sure you've added Fo-Dicom to your project via NuGet. You can do this using the Package Manager Console:

Install-Package fo-dicom

Or via .NET CLI:

dotnet add package fo-dicom

1. DICOM Header Operations (Add/Update/Delete/Query) #

Fo-Dicom uses the DicomDataset class as the core container for all DICOM tag data. Here's how to handle each operation:

Add/Update DICOM Tags #

Adding and updating tags use similar logic—Fo-Dicom's AddOrUpdate method handles both cases (creates the tag if it doesn't exist, overwrites it if it does). For custom private tags, you'll need to define the tag explicitly first:

// Load an existing DICOM file, or create a new empty one
var dicomFile = DicomFile.Open("path/to/your/dicom.dcm");
var dataset = dicomFile.Dataset;

// Add/update a standard tag (Patient Name)
dataset.AddOrUpdate(DicomTag.PatientName, "Smith^Alice");

// Add a custom private tag
var privateTag = new DicomTag(0x7000, 0x0010, DicomVR.LO, "MyAppPrivateCreator");
dataset.AddOrUpdate(privateTag, "CustomPatientNote");

// Save the modified file
dicomFile.Save("path/to/updated_dicom.dcm");

Delete DICOM Tags #

Always check if a tag exists before removing it to avoid exceptions, or use the safe TryRemove method:

// Remove a single tag safely
if (dataset.TryRemove(DicomTag.PatientAddress, out _))
{
    Console.WriteLine("Patient Address tag removed successfully");
}

// Remove multiple tags at once
var tagsToCleanup = new[] { DicomTag.PatientPhone, DicomTag.PatientEmail };
foreach (var tag in tagsToCleanup)
{
    dataset.TryRemove(tag, out _);
}

// Save changes
dicomFile.Save("path/to/cleaned_dicom.dcm");

Query DICOM Tags #

Use type-specific TryGet methods to safely retrieve tag values, or iterate through all tags to inspect the full dataset:

// Get a string tag value (Patient Name)
if (dataset.TryGetString(DicomTag.PatientName, out var patientName))
{
    Console.WriteLine($"Patient Name: {patientName}");
}

// Get a numeric tag value (Study ID)
if (dataset.TryGetInt32(DicomTag.StudyID, out var studyId))
{
    Console.WriteLine($"Study ID: {studyId}");
}

// Iterate through all tags in the dataset
foreach (var element in dataset.Elements)
{
    Console.WriteLine($"Tag: {element.Tag}, VR: {element.ValueRepresentation}, Value: {element.ToString()}");
}

2. Embedding Image Pixel Data into DICOM File #

To add pixel data, you'll need to set mandatory DICOM metadata (like image dimensions, bit depth, and UIDs) before converting your image data into a DicomPixelData object.

From a Bitmap Image #

If you're working with standard image files (PNG, JPG), you can directly convert a Bitmap to DICOM pixel data:

// Load your source image
using var sourceImage = new Bitmap("path/to/your/image.png");

// Create a new DICOM file with required metadata
var dicomFile = new DicomFile();
var dataset = dicomFile.Dataset;

// Set mandatory DICOM tags for image storage
dataset.AddOrUpdate(DicomTag.PatientName, "Smith^Alice");
dataset.AddOrUpdate(DicomTag.StudyInstanceUID, DicomUID.Generate());
dataset.AddOrUpdate(DicomTag.SeriesInstanceUID, DicomUID.Generate());
dataset.AddOrUpdate(DicomTag.SOPInstanceUID, DicomUID.Generate());
dataset.AddOrUpdate(DicomTag.SOPClassUID, DicomUID.CTImageStorage); // Adjust based on your image type
dataset.AddOrUpdate(DicomTag.Rows, sourceImage.Height);
dataset.AddOrUpdate(DicomTag.Columns, sourceImage.Width);
dataset.AddOrUpdate(DicomTag.BitsAllocated, 8);
dataset.AddOrUpdate(DicomTag.BitsStored, 8);
dataset.AddOrUpdate(DicomTag.HighBit, 7);
dataset.AddOrUpdate(DicomTag.PixelRepresentation, 0); // 0 = unsigned, 1 = signed
// Use "MONOCHROME2" for grayscale images instead
dataset.AddOrUpdate(DicomTag.PhotometricInterpretation, PhotometricInterpretation.Rgb.Value);

// Convert bitmap to DICOM pixel data and add the frame
using var pixelData = DicomPixelData.Create(dataset, true);
pixelData.AddFrame(sourceImage);

// Save the final DICOM file
dicomFile.Save("path/to/image_with_pixels.dcm");

From a Raw Pixel Array #

If you have a raw byte array of pixel data (e.g., from a sensor), you can directly inject it into the DICOM dataset:

// Example: 512x512 8-bit grayscale pixel array
int width = 512;
int height = 512;
byte[] rawPixels = new byte[width * height]; // Populate this array with your pixel data

// Create dataset with mandatory metadata
var dataset = new DicomDataset();
dataset.AddOrUpdate(DicomTag.PatientName, "Smith^Alice");
dataset.AddOrUpdate(DicomTag.StudyInstanceUID, DicomUID.Generate());
dataset.AddOrUpdate(DicomTag.SeriesInstanceUID, DicomUID.Generate());
dataset.AddOrUpdate(DicomTag.SOPInstanceUID, DicomUID.Generate());
dataset.AddOrUpdate(DicomTag.SOPClassUID, DicomUID.CTImageStorage);
dataset.AddOrUpdate(DicomTag.Rows, height);
dataset.AddOrUpdate(DicomTag.Columns, width);
dataset.AddOrUpdate(DicomTag.BitsAllocated, 8);
dataset.AddOrUpdate(DicomTag.BitsStored, 8);
dataset.AddOrUpdate(DicomTag.HighBit, 7);
dataset.AddOrUpdate(DicomTag.PixelRepresentation, 0);
dataset.AddOrUpdate(DicomTag.PhotometricInterpretation, PhotometricInterpretation.Monochrome2.Value);

// Inject raw pixel data
using var pixelData = DicomPixelData.Create(dataset, true);
pixelData.AddFrame(rawPixels);

// Save to file
var dicomFile = new DicomFile(dataset);
dicomFile.Save("path/to/raw_pixel_dicom.dcm");

Quick Tips #

• Always validate your modified DICOM files with a DICOM viewer to ensure compliance with DICOM standards.
• For private tags, make sure to register the private creator tag first if your use case requires it.
• When working with large pixel datasets, consider using streaming to avoid excessive memory usage.

内容的提问来源于stack exchange，提问作者Ahsan Iqbal