Free online DICOM viewer: open your scan in the browser

1

Gather the study files

Copy the whole folder from the CD or the USB stick to your computer. The images usually live inside a folder called DICOM or IMAGES, and often have no extension at all: files named IM000001 or similar are still DICOM. If the imaging center handed you a ZIP, use it as is.

2

Open them in the viewer on this page

Click “Open my study” and select every file, or drag the entire folder onto the dotted area. On a computer there is also an “Open a folder” button that loads the whole study in one go; on iPhone and iPad the system offers no folder picker, so select the files or open the ZIP.

3

Let the browser build the volume

The viewer splits the series, sorts the slices by their real position and reconstructs the volume on your own machine. A progress screen shows how many files it has read. Nothing is uploaded: open the Network tab of your browser’s developer tools and you will not see your study leaving.

4

Adjust window and level and scroll the slices

Pick the preset for the tissue you care about — lung, mediastinum, bone, brain, abdomen — and move through the slices with the mouse wheel, the arrow keys or cine playback. Without that adjustment a CT looks grey and flat; the exact values behind each preset are listed further down.

5

Measure, reconstruct and compare

Trace lengths and angles: they come out in real millimeters because the study carries its physical calibration. Open the multiplanar reconstruction to navigate all three planes, the 3D view to explain a finding, and the side-by-side mode to put two series next to each other.

6

Export before closing the tab

Everything lives in the memory of the tab and disappears when you close it. If you wrote a report, export or print it before leaving; and if it should stay archived next to a visit note, that already calls for a clinical record behind it.

Nothing leaves your computer in order to view the study. The browser reads the files you selected, decodes the slices and assembles the volume on your own machine: there is no upload to our servers, we keep no copy, and nothing remains once the tab is closed. You can verify it yourself in the Network tab of your browser’s developer tools — while you explore the study, no request carrying your images ever goes out. That is also why no account is needed: there is nothing to protect on our side, because nothing of yours is on our side.

The honest limit sits on the other side of that same design. This is a review viewer, not a reading workstation: it is made for reviewing, measuring, comparing series and walking a patient through their images, but a formal diagnostic read is done on displays calibrated for radiology, with their own quality controls. And because everything lives in the tab, your measurements, key images and report vanish when you close it unless you export them first.

There is exactly one situation where the study leaves your device, and it only happens if you ask for it: the button that sends the study to your doctor. Instead of burning a CD or attaching loose files to a chat, the viewer packs the study and encrypts it on your own device before it leaves. What we store are bytes we cannot read.

The key that opens the study travels inside the link itself, in the part after the “#” — the one fragment of a URL that browsers never send to any server. So the full link, and only the full link, opens the study: not us, and not anyone who intercepts the stored file. The link expires on its own after 72 hours, can be opened a limited number of times, can be revoked whenever you want, and the file is deleted afterwards. The flip side is worth saying out loud: if you lose the link, the study is unrecoverable — there is no key to reset because we never had it.

For a patient, that means handing the radiologist’s images to a second opinion without mailing a disc. For a doctor, it means receiving a study that opens in any browser, full resolution, without asking the patient to install anything.

The viewer does not distinguish between “CT scans” and “everything else”: it reads DICOM, and DICOM is what comes out of any imaging machine. What changes is what you can do with each kind of study, because a volume of slices and a single projection image are not the same thing.

On the file side it accepts what actually shows up in real life: .dcm files and DICOM files with no extension, entire folders, the ZIP from the imaging center — unpacked inside the browser — NIfTI volumes (.nii and .nii.gz), the usual format in research, and the report PDF. RAR and 7z archives cannot be opened inside a browser; if you were given one, extract it on your computer and drop the resulting folder here. If what you have is a specific modality, each one has its own guide: CT scans, MRI, X-rays and ultrasound.

This is the difference between seeing a study and actually reading it. A CT slice measures density in Hounsfield units (HU), from about −1000 for air to more than 1000 for cortical bone: thousands of possible values, while a screen only draws 256 shades of grey. The window (window width) defines how wide a range of densities gets spread across those 256 shades, and the level (window center) sets where that range is centered. Everything outside it is crushed to pure white or pure black.

That is why the same slice looks completely different depending on the setting, and why a chest is reviewed twice: once with a lung window and once with a mediastinal window. These are the presets built into the viewer on this page, with their values:

In MRI and ultrasound the pixel values are not Hounsfield units — those modalities do not measure density — so there the window is adjusted by eye, dragging with the mouse until the tissue you are after separates from the rest. The viewer starts from whatever setting is stored in the study itself, which is usually the one the technologist chose.

This is the confusion that wastes the most time. Converting the study to JPG and opening it in a photo app “works” in the sense that an image appears on screen, but almost everything that makes a study useful is lost. A real DICOM viewer has to solve four problems that an image app does not.

Bit depth: a CT stores thousands of density levels and a JPG only 256. Converting freezes the range: you can no longer switch from a lung window to a bone window, because the data that made it possible is gone.

Physical scale: every study declares how many millimeters a pixel measures and how far apart the slices are. With that, a measurement on the image is worth real millimeters. Without it, you are counting pixels.

Series order and geometry: the files on a CD do not arrive in order. The viewer groups them by series and sorts them by each slice’s real position in the patient — the only order that reconstructs the volume without distorting it.

Reconstruction: once the volume is in order it can be cut along any plane (MPR) or rendered in 3D. That is what turns a pile of images into a study you can navigate.

The JPG is still useful for messaging a single image to a colleague. For reviewing, measuring or comparing, always keep the original DICOM files — and open them with a viewer that understands them.

Because there is nothing to install, there is nothing operating-system-specific either. The same viewer runs on Windows, macOS, Linux and ChromeOS, in any up-to-date Chrome, Edge, Firefox or Safari with WebGL2 — which in practice means any computer bought in the last several years, including a Chromebook. No admin rights, no IT ticket, no “this CD only has a Windows viewer” dead end on a Mac.

Phones and tablets work too, with one practical note for Apple devices: the iOS file picker does not offer a whole-folder option, so on iPhone and iPad you should either select all the files of the study at once or — much easier — open the ZIP directly, which the viewer unpacks in the browser. If your imaging center offers a download, ask for the ZIP; if you have the folder on a computer, compress it once and open that on your phone.

The free viewer forgets everything when the tab closes, and for a one-off look that is exactly right. It stops being right when the study belongs to your patient and you will need it again: the comparison against the scan from six months ago, the signed report filed with that visit’s note, the measurements you made last time, access from your other office. That no longer fits in a browser tab — it needs a record behind it.

It is the same viewer, mounted inside Healthy Record: you upload the CD or USB into the chart, the study stays in the patient’s history with access control and an audit log of who opened it, and the report is archived with the encounter. The trial is 14 days, in your own practice, with your own studies.