Simple DICOM Viewer Build 4 can import medical-image volumes, scientific volumes, ordinary image stacks, polygon meshes, and supported scene files. A user can select one file, several related files, or an entire containing folder.

All successfully imported content is converted into the app’s shared, memory-protected Int16 volume representation. It can then use the same GPU rendering controls, including direct volume rendering, maximum intensity projection, isosurface rendering, presets, window and level, opacity, transfer functions, clipping planes, crop controls, and Apple Vision Pro immersive presentation.

Import compatibility depends on the exact file structure, scalar type, compression method, companion files, available memory, operating-system support, and the limits documented below.

QUICK LIST OF ACCEPTED INPUTS

Medical and scientific volumes:
• DICOM: .dcm, .dicom, extensionless DICOM image instances, multiple files, folders, and nested folders
• NRRD: .nrrd
• Detached NRRD: .nhdr with a companion payload such as .raw or .raw.gz
• NIfTI-1 and NIfTI-2: .nii
• Compressed NIfTI: .nii.gz
• MetaImage: .mhd and .mha
• RAW volume: .raw or .raw.gz with a matching .json sidecar
• Legacy VTK image data: .vtk
• VTK XML ImageData: .vti

Image files and image stacks:
• TIFF: .tif and .tiff
• PNG: .png
• JPEG: .jpg and .jpeg
• BMP: .bmp
• HEIC: .heic
• HEIF: .heif
• Multi-page image files supported by Apple ImageIO, including multi-page TIFF where the source contains multiple frames

Polygon meshes and scenes:
• STL: .stl
• Wavefront OBJ: .obj
• Polygon File Format: .ply
• glTF 2.0: .gltf
• Binary glTF 2.0: .glb
• Universal Scene Description: .usd, .usda, .usdc, and .usdz

HOW TO IMPORT

1. Open Simple DICOM Viewer.
2. Select Import or Load.
3. Choose one file, several files, or a containing folder.
4. For a format with companion files, select the containing folder instead of only the header file.
5. Wait while the app identifies the format, validates the source, applies the active memory budget, and prepares the render volume.
6. Review the displayed source name, dimensions, spacing, value range, downsampling status, and visible orientation.
7. Reset crop and clipping controls when checking a newly imported dataset.
8. Select Generic Grayscale first when an unfamiliar volume appears dark or transparent.
9. On Apple Vision Pro, confirm the volume in the window before opening immersive space.

1. DICOM INPUT

Accepted selections:
• A single .dcm file
• A single .dicom file
• Multiple DICOM image instances
• Extensionless DICOM image instances
• A folder containing one series
• A folder containing several series
• A nested folder tree from a PACS, workstation, disc, drive, or Files provider

Native DICOM transfer syntaxes:
• 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

Native DICOM pixel handling:
• Single-sample monochrome images
• MONOCHROME1 and MONOCHROME2
• 8-bit allocated pixels
• 16-bit allocated pixels
• 32-bit allocated pixels
• Signed or unsigned stored values
• Common rescale slope and intercept
• Single-frame and compatible multiframe data
• Pixel spacing, slice thickness, image position, and image orientation where available
• Series grouping using Series Instance UID
• Geometry-based slice ordering where sufficient metadata is available

Primary intended DICOM content:
• Uncompressed CT image series
• Uncompressed MR image series
• Other compatible uncompressed scalar monochrome DICOM image series may load when they use the supported encoding and consistent dimensions

DICOM limitations:
• The built-in native parser does not decode compressed or encapsulated DICOM pixel data.
• JPEG Baseline, JPEG Lossless, JPEG-LS, JPEG 2000, RLE, deflated, proprietary, and vendor-specific transfer syntaxes require an actually linked and appropriately licensed codec engine.
• Color and multi-sample DICOM images are not handled by the native volume parser.
• Structured reports, presentation states, radiotherapy objects, encapsulated documents, waveforms, videos, and non-image DICOM objects are not volume inputs.
• DICOMDIR itself is not rendered. Select the folder that contains the referenced image instances.
• The app automatically chooses the largest coherent readable series in the current native workflow. Confirm that the selected series is the one you intended to view.
• Always verify patient and study identity, orientation, spacing, rescale values, and visible anatomy against an appropriate reference system.

Recommended DICOM preparation:
• Use a de-identified copy for education, demonstrations, support, or App Review.
• Keep one coherent image series in a clearly named folder when possible.
• For maximum native compatibility, export as uncompressed Explicit VR Little Endian or Implicit VR Little Endian.
• Do not modify or overwrite the original clinical archive.

2. NRRD AND NHDR INPUT

Accepted files:
• Embedded NRRD: .nrrd
• Detached NRRD header: .nhdr
• Detached payload referenced by the header, including .raw or .raw.gz when correctly named

Accepted encodings:
• raw
• ascii
• text
• txt
• gzip
• gz

Accepted data characteristics:
• Three-dimensional scalar volumes
• Four-dimensional files when the first and fastest axis is an interleaved component, vector, RGB, RGBA, color, list, or similar component axis
• Common signed and unsigned integer scalar types
• 32-bit and 64-bit floating-point scalar types
• Little-endian and big-endian payloads
• Byte skip and line skip
• Spacing derived from space directions or spacings when present
• Multiple components reduced to one scalar using luminance, magnitude, or the first component as appropriate

Companion-file rule:
For .nhdr, select the containing folder so the referenced data file remains within the file-access scope granted by the system picker.

NRRD limitations:
• NRRD LIST payloads are not supported.
• printf-style multi-file data patterns are not supported.
• A detached dataset should use one .nhdr plus one raw or gzip payload.
• The current render model preserves voxel spacing but does not apply the NRRD world origin or complete direction matrix to the scene transform.
• Confirm orientation independently before using measurements or spatial comparisons.

3. NIFTI INPUT

Accepted files:
• NIfTI-1 single-file volume: .nii
• NIfTI-2 single-file volume: .nii
• Gzip-compressed NIfTI: .nii.gz

Accepted data characteristics:
• NIfTI-1 headers
• NIfTI-2 headers
• Little-endian and big-endian data
• Common 8-bit, 16-bit, 32-bit, and 64-bit signed and unsigned integer types
• 32-bit and 64-bit floating-point types
• RGB and RGBA data reduced to grayscale luminance
• vox_offset
• Scaling slope and intercept
• Spatial units converted to millimetres where specified
• Voxel spacing
• The first 3D frame of a higher-dimensional NIfTI dataset

Compressed NIfTI:
The .nii.gz adapter uses the included bounded zlib/gzip bridge. Expanded output is limited according to the active device memory budget.

NIfTI limitations:
• Complex, binary, and 128-bit scalar datatypes must be converted before import.
• Only the first 3D frame of a 4D or higher-dimensional dataset is rendered.
• The current adapter reads dimensions and voxel spacing but does not apply the NIfTI qform or sform matrix to reorient the rendered object.
• Confirm left-right orientation and world alignment against a trusted reference.

4. METAIMAGE INPUT

Accepted files:
• MetaImage header: .mhd
• Embedded MetaImage: .mha
• A detached payload referenced by ElementDataFile

Accepted data characteristics:
• Two-dimensional and three-dimensional data
• LOCAL payloads embedded in .mha
• One detached data file referenced by .mhd
• Binary raw payloads
• ASCII payloads
• zlib-compressed payloads
• gzip-compressed payloads
• Little-endian and big-endian values
• Common signed and unsigned integer types
• 32-bit and 64-bit floating-point types
• Multiple channels reduced to luminance, magnitude, or the first component
• ElementSpacing or ElementSize
• HeaderSize, including -1 for compatible uncompressed binary payloads
• Rescale slope and intercept fields when present

Companion-file rule:
For .mhd, select the containing folder with the referenced ElementDataFile.

MetaImage limitations:
• MetaImage LIST slice collections are not supported.
• Higher-dimensional MetaImage data must be exported as a 3D volume.
• Use one .mha file or one .mhd plus one payload.
• The current renderer does not apply full MetaImage TransformMatrix, Offset, or anatomical-orientation metadata to the scene transform.
• Confirm orientation independently.

5. RAW VOLUME INPUT

Accepted files:
• Uncompressed raw payload: .raw
• Gzip-compressed raw payload: .raw.gz
• Required same-base JSON sidecar: .json

Naming example:
• research_volume.raw
• research_volume.json

or:
• research_volume.raw.gz
• research_volume.json

The JSON sidecar can describe:
• width
• height
• depth
• dimensions or size array
• scalar type
• little-endian or big-endian byte order
• voxel spacing
• byte offset or header size
• number of components or channels
• component reduction
• rescale slope
• rescale intercept
• modality
• description or series name
• compressed or gzip status

Supported scalar names include common forms of:
• char / int8
• unsigned char / uint8
• short / int16
• unsigned short / uint16
• int / int32
• unsigned int / uint32
• long long / int64
• unsigned long long / uint64
• float / float32
• double / float64

RAW JSON example:

{
  "width": 512,
  "height": 512,
  "depth": 300,
  "type": "short",
  "endian": "little",
  "spacing": [0.7, 0.7, 1.0],
  "offset": 0,
  "components": 1,
  "slope": 1.0,
  "intercept": 0.0,
  "modality": "CT",
  "description": "Research volume"
}

RAW limitations:
• RAW has no self-describing header; incorrect dimensions, type, byte order, or offset will produce an incorrect volume.
• The app cannot infer orientation from a raw byte stream.
• Validate the JSON sidecar carefully before interpreting the result.

6. LEGACY VTK INPUT

Accepted file:
• .vtk

Accepted dataset declarations:
• STRUCTURED_POINTS
• IMAGE_DATA

Accepted content:
• ASCII scalar arrays
• Binary scalar arrays
• DIMENSIONS
• SPACING or ASPECT_RATIO
• POINT_DATA
• CELL_DATA
• SCALARS
• One or more scalar components
• Common VTK signed and unsigned integer and floating-point numeric types
• Standard big-endian legacy binary data

Legacy VTK limitations:
• The adapter is a focused image-volume reader, not the complete VTK framework.
• POLYDATA, UNSTRUCTURED_GRID, RECTILINEAR_GRID, FIELD-only data, multiblock datasets, and other non-image datasets are not accepted by this volume adapter.
• Origin is not currently used to reposition the render volume.
• Unsupported attributes or structures return an error instead of being approximated silently.

7. VTK XML IMAGEDATA INPUT

Accepted file:
• .vti

Accepted content:
• VTK XML ImageData
• WholeExtent and Piece Extent
• PointData or CellData scalar arrays
• ASCII DataArray values
• Inline base64 binary
• Appended base64 data
• Appended raw data
• UInt32 block headers
• UInt64 block headers
• Little-endian and big-endian values
• vtkZLibDataCompressor compressed blocks
• Multiple scalar components reduced to a renderable scalar volume

VTI limitations:
• The adapter is a focused native ImageData reader, not the complete VTK framework.
• Unsupported compressors are rejected.
• Parallel .pvti files are not imported.
• Multiblock and multi-piece structures are not reconstructed.
• Non-image VTK XML datasets are not imported through the VTI adapter.
• Origin and direction metadata are not currently applied to the rendered scene transform.

8. ORDINARY IMAGES AND IMAGE STACKS

Accepted extensions:
• .tif
• .tiff
• .png
• .jpg
• .jpeg
• .bmp
• .heic
• .heif

Accepted selections:
• One image
• Multiple selected images
• A folder of ordered image slices
• Multi-page files that Apple ImageIO exposes as multiple frames, including multi-page TIFF

Processing:
• Images are decoded with Apple image frameworks.
• Images are converted to 8-bit grayscale and then stored in the shared Int16 volume.
• Large dimensions and long stacks are sampled according to the active device budget.
• Image files are ordered using localized filename order.

Recommended stack naming:
• slice_0001.tif
• slice_0002.tif
• slice_0003.tif

Image-stack limitations:
• Ordinary image formats usually do not contain reliable 3D medical orientation or slice spacing.
• A neutral spacing is used when calibrated spacing is unavailable.
• JPEG is lossy and may contain compression artifacts.
• Color is reduced to grayscale.
• Do not treat an ordinary image stack as metrically calibrated without independent metadata.

9. STL, OBJ, AND PLY INPUT

Accepted files:
• .stl — ASCII and binary STL
• .obj — Wavefront OBJ vertex and face geometry
• .ply — ASCII and binary PLY

Processing:
• Polygon geometry is read and converted to a compact surface voxel volume.
• The resulting volume uses the same Composite, MIP, Isosurface, transfer-function, clipping, crop, iPhone, iPad, and immersive rendering paths.

Mesh limitations:
• The conversion is a visualization-oriented surface voxelization.
• It is not an exact CAD, finite-element, or watertight solid conversion.
• Materials, textures, per-face colors, animation, rigging, manufacturing tolerances, and semantic metadata are not preserved by the shared volume renderer.
• Very large meshes may be triangle-limited or voxelized at a reduced grid resolution.

10. GLTF AND GLB INPUT

Accepted files:
• .gltf
• .glb

Accepted glTF 2.0 content:
• JSON .gltf containers
• Binary GLB 2.0 containers
• Embedded base64 buffers
• External .bin buffers
• Buffer views
• Accessors
• Node hierarchy
• Node transforms
• Indexed triangle primitives
• Non-indexed triangle primitives
• Triangle strips
• Triangle fans
• Common component and index types

Companion-file rule:
For .gltf with external .bin files, select the containing folder.

glTF and GLB limitations:
• Draco-compressed geometry is not supported.
• meshopt-compressed geometry is not supported.
• Sparse accessors are not supported.
• Animation and skinning are not imported.
• Non-triangle geometry is not fabricated.
• Materials, textures, lights, cameras, and scene appearance are not reproduced in the volume renderer.
• Geometry is voxelized into the shared integer-volume representation.

11. USD-FAMILY INPUT

Accepted files:
• .usd
• .usda
• .usdc
• .usdz

Processing:
• The app uses Apple Model I/O as the scene-import bridge.
• The installed operating system is asked whether it supports the selected extension.
• Exposed polygon meshes and object transforms are read.
• Visible geometry is converted into the shared surface-voxel volume.

USD limitations:
• Coverage depends on the installed operating system and the contents of the asset.
• Not every USD schema, material network, shader, animation, skeleton, procedural, point instancer, or referenced asset is guaranteed to be exposed by Model I/O.
• Materials and textures are not preserved in the volume renderer.
• USD geometry is voxelized for volume visualization rather than reproduced as a complete scene graph.

COMPANION FILES — IMPORTANT

Select the containing folder for:
• .nhdr with .raw, .raw.gz, or another referenced payload
• .mhd with its referenced ElementDataFile
• .raw or .raw.gz with the matching .json sidecar
• .gltf with external .bin buffers
• A DICOM series containing many image instances
• A folder of TIFF, PNG, JPEG, BMP, HEIC, or HEIF slices

The system document picker grants access only to the selected items. Selecting the containing folder is the most reliable way to make companion files available to the importer.

MEMORY AND PERFORMANCE

To reduce memory pressure, the app may:
• Limit the number of scanned files
• Limit source bytes
• Limit maximum dimensions
• Limit total voxel count
• Sample large volumes directly into a smaller final volume
• Crop empty outer borders
• Limit mesh triangles
• Reduce mesh voxelization resolution
• Reduce drawable resolution and frame rate on iPhone
• Temporarily reduce ray samples during interaction
• Cancel an earlier import when a new import starts

A successfully imported volume may therefore be downsampled. The app should display whether adaptive reduction occurred. For research or measurement workflows, review source dimensions and spacing before relying on the result.

PRIVACY AND DATA HANDLING

The current app workflow uses the Apple system file picker and processes selected files locally whenever possible. It does not require an Orthopractis account and does not automatically upload imported datasets to Orthopractis servers.

Import only data you are authorized to access. Remove patient names and identifiers before demonstrations, App Review, screenshots, support requests, or publication. Do not email identifiable DICOM studies to the general support address.

INTENDED USE AND LIMITATION

Simple DICOM Viewer is intended for visualization, education, research, demonstration, engineering evaluation, and software testing. It is not a substitute for a qualified healthcare professional, a validated diagnostic workstation, or a regulated medical device.

Rendering presets, window and level, opacity, transfer functions, clipping, cropping, adaptive sampling, scalar reduction, mesh voxelization, and missing orientation transforms can reveal, hide, or alter the appearance of structures. Independently verify all important results.

FREQUENTLY ASKED QUESTIONS

Can I import a complete DICOM folder?
Yes. Select the containing folder. The app scans nested files, groups readable instances into series, and loads the largest coherent native-compatible series.

Can I import compressed DICOM?
Not through the built-in native parser. Compressed DICOM requires an actually linked and appropriately licensed codec engine.

Can I import NRRD from 3D Slicer?
Yes, when it is a supported embedded .nrrd or one .nhdr plus one raw/gzip payload. LIST and printf-pattern multi-file payloads are not supported.

Can I import .nii.gz?
Yes. NIfTI-1 and NIfTI-2 .nii.gz files are decompressed using the included bounded gzip bridge.

Can I import .vti?
Yes. The focused VTI adapter supports ASCII, binary, appended, raw/base64, UInt32/UInt64 headers, and vtkZLibDataCompressor for scalar ImageData.

Can I import STL, OBJ, PLY, glTF, GLB, or USDZ?
Yes, subject to the limitations above. Polygon geometry is converted into a surface voxel volume so it can use the app’s volume-rendering controls.

Are mesh colors and textures preserved?
No. Build 4 voxelizes visible geometry into a scalar volume; it does not reproduce complete materials or scene shading.

Can I import a folder of PNG or TIFF slices?
Yes. Use zero-padded filenames to maintain predictable ordering. Physical spacing and medical orientation may be unavailable.

Why is my volume rotated or mirrored?
Some non-DICOM formats include orientation matrices that the current shared volume model does not apply. Verify orientation against a reference and do not assume patient-space alignment.

Why was my volume reduced?
The app applies device-specific memory limits and adaptive sampling to prevent excessive allocations and operating-system termination.