How an AI result image gets back into the PACS

A common first question about imaging AI is simple: where does the result show up? The best answer is usually "in the study the radiologist already opens."

Step 1: the study is sent twice

When a scan finishes, the scanner or a routing system sends the images to the PACS. The same images are also sent to the AI system. These happen in parallel, so storing the study is never waiting on the AI.

Step 2: the AI produces an image

The AI analyzes the images and generates an output, such as an image with a marked region of interest. The output is packaged as standard DICOM so other systems can handle it like any other image.

Step 3: the result returns to the same study

The DICOM result is sent to the PACS and linked to the original study by its identifiers. If the identifiers do not match, the result lands in the wrong place or not at all, which is why matching and testing matter.

Step 4: the radiologist reads both

The radiologist opens one study and sees the original images and the AI result together. The AI is a support tool. The radiologist decides.

What can go wrong

• The result arrives after the radiologist has already started reading.
• The study identifiers do not match, so the result is orphaned.
• The PACS displays the AI series in a way that is easy to miss.

Each of these is a configuration and testing question, which is the core of implementation work.

Learn more→