一個基於 Python PyQt6 與 pynetdicom 打造的輕量級 DICOM 伺服器與影像管理工具。
此專案旨在提供一個完整的圖形化介面，讓醫療資訊工程人員或研究者能輕鬆地接收、傳送、預覽以及編輯 DICOM 檔案。
- DICOM 接收端 (SCP)：支援在背景啟動 C-STORE 服務，監聽特定 Port 並接收外部節點傳送的 DICOM 影像。支援多種 Transfer Syntax，涵蓋 JPEG Lossless/Lossy、JPEG 2000、RLE 等壓縮格式。
- DICOM 傳送端 (SCU)：支援將本地選取的 DICOM 檔案或直接從本地資料庫挑選檔案，推播 (C-STORE) 至遠端伺服器 (如 PACS)。傳送前會自動解壓縮並動態偵測所需的 SOP Class，確保相容性最大化。
- 連線測試 (C-ECHO)：提供一鍵 Ping 遠端 DICOM 節點，測試網路連線狀態 (Verification SOP Class)。
- 連線節點管理 (Peers)：內建連線對象儲存功能，自動將常用 AE Title、IP 與 Port 設定永久保存在本地 SQLite 資料庫中，提供下拉選單快速切換，並支援一鍵刪除已儲存的配置。
- 基礎影像預覽：支援從 SQLite 資料庫或匯入本地檔案進行 2D 灰階/彩色影像顯示，並自動套用 Rescale Slope/Intercept (HU 值轉換)。
- 動態多幀支援：針對多幀影像 (Multi-frame) 具備滑桿導航系統，支援跨檔案的統一幀索引 — 同一檢查 (Study) 底下的多個 DICOM 檔案會被自動展開為連續幀序列。
- ECG 波形圖檢視：支援 12 導程 (12-Lead) 的 DICOM 心電圖 (ECG) 繪製與捲動預覽，具備 Channel Sensitivity 校正與深色主題風格。
- 文件解析支援：自動偵測 Encapsulated PDF 類型的 DICOM 並提供「一鍵匯出以外部程式開啟」功能。
- 響應式影像縮放：影像會隨視窗大小自動等比縮放，ECG 波形圖則保持原始解析度並透過 ScrollArea 提供捲動瀏覽。
- 標籤檢視 (DICOM Tags)：一鍵展開完整的 DICOM Tag 階層式文字內容，使用等寬字型對齊顯示。
- 標籤編輯器：內建防呆機制的編輯介面，支援文字/數值/Multi-value (反斜線分隔) 型別的安全覆寫。自動排除不安全的 VR 類型 (OB, OW, OF, OD, SQ, UN) 與 PixelData 等標籤。新增 SOP UID 亂數產生器，產生後顯示於獨立文字框供快速複製。
- 一鍵匿名化：提供專用的「去識別化/匿名化」工具，勾選機敏屬性 (如 Patient Name, Patient ID, Birth Date, Sex, Institution Name 等) 即可自動抹除或替換為安全字串。支援全部勾選/全部取消的快速操作。
- 以 Study 為單位的檢查管理：左側清單以 Study UID 為群組，聚合同一檢查的所有影像，方便整體瀏覽與刪除操作。
- 安全刪除機制：執行整筆檢查或單張影像刪除時，系統會自動判斷檔案來源。若是透過 SCP 接收至內部
dicom_files目錄的檔案會連同實體檔案一併刪除；若是外部手動匯入的檔案，則僅移除資料庫紀錄而保留原始實體檔案，防止誤刪使用者珍貴資料。
- SQLite 紀錄追蹤：所有接收與開啟的患者 Metadata 會自動同步存入
dicom_local.db進行索引管理，搭配database_manager.py封裝完整的 CRUD 操作。 - 記憶體快取 (LRU Cache)：透過
functools.lru_cache為 DICOM Dataset 提供記憶體級快取，避免重複磁碟 I/O，並在編輯/匿名化後自動清除快取以確保一致性。 - 每日日誌記錄：整合
logging模組,同時輸出至 Console 與每日 Rotating Log 檔案 (logs/dicom_app_YYYY-MM-DD.log)，並於主視窗的狀態列 (Status Bar) 顯示重要活動變化。
請確保您的系統已安裝 Python 3.9+，並具備以下依賴套件：
pip install PyQt6 pydicom pynetdicom numpy matplotlib📌 補充說明
若需處理壓縮格式的影像 (如 JPEG-LS、JPEG 2000)，可能額外需要安裝解碼引擎：pip install pylibjpeg pylibjpeg-libjpeg pylibjpeg-openjpeg或使用 GDCM：
pip install python-gdcm
直接執行 main_window.py 啟動主程式：
python main_window.py如需打包為獨立的 Windows 執行檔，可使用 PyInstaller：
pip install pyinstaller
pyinstaller --noconsole --name DICOM_Server main_window.py打包完成後，執行檔位於 dist/DICOM_Server/ 目錄下。
┌──────────────────────────────────────────────────────────────────┐
│ AE Title: [DICOM_SCP] Port: [11112] [▶ 啟動] [■ 停止] │
│ [📂 開啟 DICOM] [🗑 刪除紀錄] [📤 傳送 DICOM] │
├───────────────────┬──────────────────────────────────────────────┤
│ 檢查清單 │ 影像預覽 [📋 標籤] [✏️ 編輯] [👤 匿名化] [🗑 刪除]│
│ │ │
│ Patient ID │ ... │ ┌──────────────────────┐ │
│ ────────── │ ... │ │ │ │
│ P12345 │ ... │ │ DICOM Image │ │
│ P98765 │ ... │ │ / ECG / PDF │ │
│ │ │ │ │ │
│ │ │ └──────────────────────┘ │
│ │ │ [◀] ════════════════ [▶] 3 / 10 │
├─────────────┴─────┴──────────────────────────────────────────────┤
│ 狀態列: 已接收: John Doe (P12345) │
└──────────────────────────────────────────────────────────────────┘
- 頂部控制列：輸入 AE Title 與 Port 後，點擊「▶ 啟動」即可將本機作為 DICOM Server 運行。啟動後輸入欄位會自動鎖定，直到停止為止。
- 左側清單：顯示曾接收或開啟過的「檢查紀錄」(以 Study 為單位分組)，點擊即可展開預覽。
- 右側操作區：DICOM 專屬影像/波形畫布，上方提供「📋 DICOM 標籤」、「✏️ 編輯 DICOM」、「👤 匿名化」、「🗑 刪除單張」的操作按鈕。底部為多幀導航控制列 (僅在多幀/多檔案時顯示)。
- 狀態列：即時顯示最新的操作訊息與檔案活動紀錄。
dicom_server_gui/
├── main_window.py # 主程式進入點與整體 GUI 視窗框架
├── dicom_scp_worker.py # DICOM SCP (接收端) — QThread 背景監聽 C-STORE/C-ECHO
├── dicom_scu_worker.py # DICOM SCU (傳送端) — QThread 背景執行 C-STORE 與 C-ECHO 測試
├── send_dicom_dialog.py # 「傳送 DICOM」功能的操作視窗介面 (含 Peer 管理)
├── dicom_editor_dialog.py # DICOM Tag 編輯器表單 (含 VR 型別安全轉換)
├── dicom_anonymize_dialog.py # 去識別化/匿名化工具的對話框介面
├── dicom_image_utils.py # 像素萃取、HU Windowing、多幀處理、PDF 提取 (含 LRU Cache)
├── ecg_utils.py # 12 導程 ECG 波形解析與 Matplotlib 繪製
├── database_manager.py # SQLite 資料庫封裝 — 病患/檢查紀錄的 CRUD 操作
├── logger_config.py # 集中控制 Console 與檔案日誌輸出格式
└── README.md # 專案說明文件
📌 執行時自動產生的目錄與檔案 (不含在版本控制中)
dicom_local.db— SQLite 資料庫，首次啟動時自動建立dicom_files/— SCP 接收的 DICOM 檔案存放目錄logs/— 每日日誌檔輸出目錄
main_window.py
├── database_manager.py
├── dicom_scp_worker.py
├── dicom_scu_worker.py (via send_dicom_dialog.py)
├── send_dicom_dialog.py
├── dicom_editor_dialog.py
├── dicom_anonymize_dialog.py
├── dicom_image_utils.py
├── ecg_utils.py
└── logger_config.py (所有模組共用)
SCP / SCU 皆已配置支援以下 Transfer Syntax：
| Transfer Syntax | UID |
|---|---|
| Implicit VR Little Endian | 1.2.840.10008.1.2 |
| Explicit VR Little Endian | 1.2.840.10008.1.2.1 |
| Explicit VR Big Endian | 1.2.840.10008.1.2.2 |
| Deflated Explicit VR Little Endian | 1.2.840.10008.1.2.1.99 |
| JPEG Baseline (8-bit) | 1.2.840.10008.1.2.4.50 |
| JPEG Extended (12-bit) | 1.2.840.10008.1.2.4.51 |
| JPEG Lossless SV1 | 1.2.840.10008.1.2.4.70 |
| JPEG Lossless | 1.2.840.10008.1.2.4.57 |
| JPEG 2000 Lossless | 1.2.840.10008.1.2.4.90 |
| JPEG 2000 | 1.2.840.10008.1.2.4.91 |
| RLE Lossless | 1.2.840.10008.1.2.5 |
- Patient ID 一致性：當使用「編輯」或「匿名化」更改了
Patient ID時，基於 DICOM 架構中以 Patient ID 作為病患唯一識別的設計，系統會將該檔案作為一名全新病患重新登錄，左側清單中也會多出一筆紀錄。操作前皆會跳出二次確認視窗。 - 防火牆設定：若要作為 DICOM Server (SCP) 運作，請確保選用的 Port (如
11112) 在您的作業系統防火牆規則中處於放行狀態。 - DICOM 快取機制：系統使用 LRU Cache 快取最近讀取的 2 筆 DICOM 資料集以提升效能。在編輯或匿名化操作後，快取會自動被清除以避免顯示過期資料。
- 檔案命名規則：SCP 接收到的檔案會以
SOPInstanceUID.dcm作為檔名自動儲存至dicom_files/目錄，確保檔案唯一性。
此工具僅供醫療資訊研究與測試用途，不應用於臨床診斷。使用者須自行確保符合所有適用的醫療資訊安全與隱私法規 (如 HIPAA, GDPR 等)。
A lightweight DICOM server and image management tool built with Python PyQt6 and pynetdicom.
This project aims to provide a complete graphical interface for medical IT engineers or researchers to easily receive, send, preview, and edit DICOM files.
- DICOM Receiver (SCP): Supports running a C-STORE service in the background, listening on a specific port to receive DICOM images from external nodes. Supports various Transfer Syntaxes, including JPEG Lossless/Lossy, JPEG 2000, RLE, etc.
- DICOM Sender (SCU): Supports pushing locally selected DICOM files or selecting files directly from the local database to remote servers (e.g., PACS) via C-STORE. Automatically decompresses and dynamically detects required SOP Classes before transmission to maximize compatibility.
- Connection Test (C-ECHO): Provides a one-click Ping to remote DICOM nodes to test network connectivity (Verification SOP Class).
- Peer Management: Built-in connection saving feature persistent via the local SQLite database, automatically recording and quickly switching commonly used AE Title, IP, and Port settings. Also supports one-click deletion of saved configurations.
- Basic Image Preview: Supports displaying 2D grayscale/color images from SQLite database or imported local files, automatically applying Rescale Slope/Intercept (HU value conversion).
- Dynamic Multi-frame Support: Features a slider navigation system for multi-frame images. Unifies frame indexing across files - multiple DICOM files under the same Study are automatically expanded into a continuous frame sequence.
- ECG Waveform Viewer: Supports DICOM Electrocardiogram (ECG) 12-lead drawing and scrolling preview, featuring Channel Sensitivity correction and a dark theme.
- PDF Document Parsing: Automatically detects Encapsulated PDF DICOMs and provides a "one-click export and open externally" feature.
- Responsive Image Scaling: Images automatically scale proportionally with window size, while ECG waveforms keep their original resolution and provide scrolling via ScrollArea.
- DICOM Tags Viewer: One-click expansion of the complete DICOM Tag hierarchical text content, displayed with monospace fonts.
- Tag Editor: Built-in foolproof editing interface supporting text/numeric/multi-value (backslash separated) types for safe overwriting. Automatically excludes unsafe VR types (OB, OW, OF, OD, SQ, UN) and tags like PixelData. Includes a Random SOP UID Generator that displays generated UIDs in a read-only textbox for easy copying.
- One-click Anonymizer: Dedicated de-identification/anonymization tool. Select sensitive attributes (e.g., Patient Name, Patient ID, Birth Date, Sex, Institution Name) to automatically erase or replace them with safe strings. Supports select-all/deselect-all for quick operations.
- Study-based Inspection Management: The left list groups records by Study UID, aggregating all images of the same study for an easy overview and deletion operations.
- Safe Deletion Mechanism: When deleting an entire study or a single image, the system automatically checks the file's origin. It permanently deletes physical files only if they are located inside the internal
dicom_filesdirectory. External files imported manually are preserved on the hard drive (only database records are removed) to prevent accidental data loss.
- SQLite Tracking: Patient metadata for all received and opened files is automatically synchronized to
dicom_local.dbfor index management, withdatabase_manager.pyencapsulating complete CRUD operations. - Memory Cache (LRU Cache): Provides memory-level caching for DICOM Datasets via
functools.lru_cacheto avoid repeated disk I/O, and automatically clears the cache after editing/anonymization to ensure consistency. - Daily Logging: Integrates the
loggingmodule to output to both console and daily Rotating Log files (logs/dicom_app_YYYY-MM-DD.log), and displays important activity changes on the main window's Status Bar.
Please make sure your system has Python 3.9+ installed, with the following dependencies:
pip install PyQt6 pydicom pynetdicom numpy matplotlib📌 Note
If you need to handle compressed images (e.g., JPEG-LS, JPEG 2000), additional decode engines might be required:pip install pylibjpeg pylibjpeg-libjpeg pylibjpeg-openjpegOr using GDCM:
pip install python-gdcm
Run the main program directly:
python main_window.pyIf you need to package it as a standalone Windows executable, use PyInstaller:
pip install pyinstaller
pyinstaller --noconsole --name DICOM_Server main_window.pyAfter packaging, the executable file is located in the dist/DICOM_Server/ directory.
┌──────────────────────────────────────────────────────────────────┐
│ AE Title: [DICOM_SCP] Port: [11112] [▶ Start] [■ Stop] │
│ [📂 Open DICOM] [🗑 Delete Record] [📤 Send] │
├───────────────────┬──────────────────────────────────────────────┤
│ Study List │ Image Preview [📋 Tags] [✏️ Edit] [👤 Anon] │
│ │ │
│ Patient ID │ ... │ ┌──────────────────────┐ │
│ ────────── │ ... │ │ │ │
│ P12345 │ ... │ │ DICOM Image │ │
│ P98765 │ ... │ │ / ECG / PDF │ │
│ │ │ │ │ │
│ │ │ └──────────────────────┘ │
│ │ │ [◀] ════════════════ [▶] 3 / 10 │
├─────────────┴─────┴──────────────────────────────────────────────┤
│ Status Bar: Received: John Doe (P12345) │
└──────────────────────────────────────────────────────────────────┘
- Top Control Bar: Enter AE Title and Port, then click "▶ Start" to run your local machine as a DICOM Server. Input fields are locked while running until stopped.
- Left List: Displays received or opened "Study Records" (grouped by Study), click to preview.
- Right Operation Area: DICOM exclusive image/waveform canvas. Top buttons provide "📋 DICOM Tags", "✏️ Edit DICOM", "👤 Anonymize", "🗑 Delete Single Frame" features. Bottom has a multi-frame navigation bar (only shows for multi-frame/multi-file records).
- Status Bar: Real-time display of the latest operation messages and file activity records.
dicom_server_gui/
├── main_window.py # Main entry point and main GUI frame
├── dicom_scp_worker.py # DICOM SCP (Receiver) — QThread for C-STORE/C-ECHO
├── dicom_scu_worker.py # DICOM SCU (Sender) — QThread for C-STORE & C-ECHO testing
├── send_dicom_dialog.py # "Send DICOM" dialog UI (includes Peer Management)
├── dicom_editor_dialog.py # DICOM Tag Editor dialog (includes safeguard for VR types)
├── dicom_anonymize_dialog.py # De-identification/anonymization tool dialog UI
├── dicom_image_utils.py # Pixel extraction, HU Windowing, Multi-frame processing, PDF extraction (with LRU Cache)
├── ecg_utils.py # 12-lead ECG waveform parsing & Matplotlib drawing
├── database_manager.py # SQLite database encapsulation — CRUD operations for Study records
├── logger_config.py # Centralized logger config for Console and file outputs
└── README.md # Project documentation
📌 Auto-generated Directories and Files (Not in Version Control)
dicom_local.db— SQLite database, created on first launchdicom_files/— Directory for SCP received DICOM fileslogs/— Directory for daily log file outputs
main_window.py
├── database_manager.py
├── dicom_scp_worker.py
├── dicom_scu_worker.py (via send_dicom_dialog.py)
├── send_dicom_dialog.py
├── dicom_editor_dialog.py
├── dicom_anonymize_dialog.py
├── dicom_image_utils.py
├── ecg_utils.py
└── logger_config.py (Shared by all modules)
Both SCP / SCU are configured to support the following Transfer Syntaxes:
| Transfer Syntax | UID |
|---|---|
| Implicit VR Little Endian | 1.2.840.10008.1.2 |
| Explicit VR Little Endian | 1.2.840.10008.1.2.1 |
| Explicit VR Big Endian | 1.2.840.10008.1.2.2 |
| Deflated Explicit VR Little Endian | 1.2.840.10008.1.2.1.99 |
| JPEG Baseline (8-bit) | 1.2.840.10008.1.2.4.50 |
| JPEG Extended (12-bit) | 1.2.840.10008.1.2.4.51 |
| JPEG Lossless SV1 | 1.2.840.10008.1.2.4.70 |
| JPEG Lossless | 1.2.840.10008.1.2.4.57 |
| JPEG 2000 Lossless | 1.2.840.10008.1.2.4.90 |
| JPEG 2000 | 1.2.840.10008.1.2.4.91 |
| RLE Lossless | 1.2.840.10008.1.2.5 |
- Patient ID Consistency: When changing
Patient IDvia "Edit" or "Anonymize", based on the DICOM architecture where Patient ID is the unique identifier, the system will re-register the file as an entirely new patient, adding a new record in the left list. A confirmation dialog will always appear before doing this. - Firewall Settings: To run as a DICOM Server (SCP), make sure the selected Port (e.g.,
11112) is allowed in your Operating System's firewall rules. - DICOM Cache Mechanism: The system uses an LRU Cache to cache the recently read 2 DICOM datasets for performance optimization. After edit or anonymization operations, the cache is automatically cleared to avoid displaying outdated data.
- File Naming Convention: Files received by the SCP will be automatically saved into the
dicom_files/directory, usingSOPInstanceUID.dcmas the filename to ensure uniqueness.
This tool is strictly intended for Medical IT Research and Testing purposes only, and should NOT be used for clinical diagnosis. Users must ensure compliance with all applicable medical information security and privacy regulations (e.g., HIPAA, GDPR, etc.).