import ij.*; import ij.io.*; import ij.util.Tools; import ij.process.*; import ij.plugin.*; import java.io.*; import java.util.*; import ij.measure.*; import java.awt.image.*; /** Imports a Z series(image stack) from a Biorad MRC 600 confocal microscope. The width, height and number of images are extracted from the first 3 16-bit word in the 76 byte header. Use Image/Show Info to display the header information. */ public class Biorad_Reader extends ImagePlus implements PlugIn { private final int NOTE_SIZE = 96; private BufferedInputStream f; private String directory; private String fileName; private String notes = ""; private int lutOffset = -1; public void run(String arg) { OpenDialog od = new OpenDialog("Open Biorad...", arg); directory = od.getDirectory(); fileName = od.getFileName(); if (fileName==null) return; IJ.showStatus("Opening: " + directory + fileName); FileInfo fi = null; try {fi = getHeaderInfo();} catch (Exception e) { IJ.showStatus(""); IJ.showMessage("BioradReader", ""+e); return; } if (fi!=null) { FileOpener fo = new FileOpener(fi); ImagePlus imp = fo.open(false); if (imp==null) return; setStack(fileName, imp.getStack()); try { Calibration BioRadCal = getBioRadCalibration(fi.width, fi.height, fi.nImages); setCalibration(BioRadCal); } catch (Exception e) { IJ.showStatus(""); IJ.showMessage("BioradReader", ""+e); return; } boolean hasLut = false; if (lutOffset!=-1) { try {hasLut = getLut(fi);} catch (Exception e) { IJ.showStatus(""); IJ.showMessage("BioradReader", "Can't read LUT from file because "+e); hasLut = false; } } if (hasLut) { ColorModel cm = new IndexColorModel(8, 256, fi.reds, fi.greens, fi.blues); getProcessor().setColorModel(cm); getStack().setColorModel(cm); } if (!notes.equals("")) setProperty("Info", notes); if (arg.equals("")) show(); } } int getByte() throws IOException { int b = f.read(); if (b ==-1) throw new IOException("unexpected EOF"); return b; } int getShort() throws IOException { int b0 = getByte(); int b1 = getByte(); return ((b1 << 8) + b0); } int getInt() throws IOException { int b0 = getShort(); int b1 = getShort(); return ((b1<<16) + b0); } void openFile() throws IOException { f = new BufferedInputStream(new FileInputStream(directory+fileName)); } FileInfo getHeaderInfo() throws IOException { openFile(); int width = getShort(); // 0-1 int height = getShort(); // 2-3 int nImages = getShort(); // 4-5 f.skip(8); int byte_format = getShort(); //14-15 f.skip(38); int magicNumber = getShort(); // 54-55 f.close(); // A Biorad .PIC file should have 12345 in bytes 54-55 String notBioradPicFile = "This does not seem to be a Biorad Pic File"; if (magicNumber!=12345) throw new IOException(notBioradPicFile); FileInfo fi = new FileInfo(); fi.fileFormat = fi.RAW; fi.fileName = fileName; fi.directory = directory; fi.width = width; fi.height = height; fi.nImages = nImages; fi.offset = 76; switch (byte_format) { case 1: fi.fileType = fi.GRAY8; break; case 0: fi.fileType = fi.GRAY16_UNSIGNED; break; } return fi; } /** Extracts the calibration info from the ASCII "notes" at the end of Biorad pic files. */ Calibration getBioRadCalibration(int width, int height, int nImages) throws IOException { Calibration BioRadCal = new Calibration(); int NoteFlag, NoteType, Offset; String NoteContent = new String(); byte[] TempByte = new byte[80]; byte[] RawNote = new byte[80]; double ScaleX, ScaleY; double ScaleZ=0, framesPerSecond = 0, frameInterval = 0; boolean xyt = false; Offset = 76 + height * width * nImages; openFile(); f.skip(Offset); /** Do ... While : cycle through notes until you reach the last note (indicated by bytes 2-5) For each note, different from 'live note', see if it contains axis calibration data. of so, extract the necessary values. */ do { f.skip(2); NoteFlag = getInt(); // 2-5 f.skip(4); NoteType = getShort(); // 10-11 f.skip(4); // store bytes 16-95 in a byte array f.read(RawNote); // replace illegal characters with 32 byte ch; for (int i=0; i<80; i++) { ch = RawNote[i]; TempByte[i] = (ch>=32 && ch<=126)?ch:32; } String Note = new String(TempByte); notes += Note + "\n"; // save note for Image/Show Info command /** only analyze notes != 1, so skip all the notes that say 'Live ...', they don't contain calibration info */ if (NoteType!=1) { NoteContent = getField(Note,1); /** See if first field contains keyword "AXIS_2" --> X-axis calibration (don't know why Biorad calls it '2'. */ if (NoteContent.indexOf("AXIS_2") >= 0 ) { //IJ.showMessage(Note); //IJ.showMessage(getField(Note, 4)); ScaleX = s2d(getField(Note, 4)); BioRadCal.pixelWidth = ScaleX; // fifth field contains units (mostly microns) BioRadCal.setUnit(getField(Note, 5)); } else if ( NoteContent.indexOf("AXIS_3") >= 0 ) { // contains Y-axis calibration //IJ.showMessage(Note); //IJ.showMessage(getField(Note, 4)); ScaleY = s2d( getField(Note, 4)); BioRadCal.pixelHeight = ScaleY; BioRadCal.setUnit(getField(Note, 5)); } else if ( NoteContent.indexOf("AXIS_4") >= 0 ) { // contains Z-axis calibration //IJ.showMessage(Note); //IJ.showMessage(getField(Note, 4)); ScaleZ = s2d( getField(Note, 4)); BioRadCal.pixelDepth = ScaleZ; if (getField(Note,5).indexOf("Seconds")>=0) { xyt = true; } /**ImageJ does not contain seperate units for the Z-direction. If the AXIS_4 units are "Seconds" we are dealing with an xyt scan (no z movement) otherwise we are dealing with a z-series. see below "if (xyt)" */ } else if ( NoteContent.indexOf("INFO_FRAME_RATE")>=0 ) { // Contains frame rate //IJ.showMessage(Note); //IJ.showMessage(getField(Note, 3)); framesPerSecond = s2d(getField(Note,3)); frameInterval = 1/framesPerSecond; BioRadCal.frameInterval = frameInterval; } } Offset += NOTE_SIZE; // Jump to next note // stop if this was the last note } while ( NoteFlag != 0 & f.available()>=NOTE_SIZE); /** A LUT can optionally follow the notes as a raw 768 byte LUT, save the current offset and latter we will try to read the LUT */ lutOffset = Offset; if (xyt) { /**If the file is an xyt scan (no z movement) the frame interval (1/frame rate from the INFO_FRAME_RATE note) should be used as the pixel depth, otherwise (z-series) the pixel depth from the AXIS_4 note should be used */ BioRadCal.pixelDepth = frameInterval; } f.close(); return BioRadCal; // return the filled biorad calibration } /** tries to read a 768 byte raw LUT at the end of the pic file. lutOffset is the byte after the last note, it was saved when the calibration information was read **/ boolean getLut(FileInfo fi) throws IOException { openFile(); f.skip(lutOffset); fi.reds = new byte[256]; fi.greens = new byte[256]; fi.blues = new byte[256]; boolean hasLut = false; if (f.available()>=768) { hasLut = true; f.read(fi.reds); f.read(fi.greens); f.read(fi.blues); } f.close(); return hasLut; } /* Extracts a certain field from a string. A space is considered as the field delimiter. */ String getField(String str, int fieldIndex) { char delimiter = ' '; int startIndex=0, endIndex; for (int i=1; i=0 && endIndex>=0) return str.substring(startIndex, endIndex); else return ""; } /** Converts a string to a double. Returns 1.0 if the string does not contain a valid number. Updated to use ImageJ's Tool class from the utility package */ double s2d(String s) { return Tools.parseDouble(s); } } /* Bio-Rad(TM) .PIC Image File Information (taken from: "Introductory Edited Version 1.0", issue 1/12/93.) (Location of Image Calibration Parameters in Comos 6.03 and MPL .PIC files) The general structure of Bio-Rad .PIC files is as follows: HEADER (76 bytes) Image data (#1) . . Image data (#npic) NOTE (#1) . ; NOTES are optional. . NOTE (#notes) RGB LUT (color Look Up Table) Header Information: The header of Bio-Rad .PIC files is fixed in size, and is 76 bytes. ------------------------------------------------------------------------------ 'C' Definition byte size Information (bytes) ------------------------------------------------------------------------------ int nx, ny; 0 2*2 image width and height in pixels int npic; 4 2 number of images in file int ramp1_min, ramp1_max; 6 2*2 LUT1 ramp min. and max. NOTE *notes; 10 4 no notes=0; has notes=non zero BOOL byte_format; 14 2 bytes=TRUE(1); words=FALSE(0) int n; 16 2 image number within file char name[32]; 18 32 file name int merged; 50 2 merged format unsigned color1; 52 2 LUT1 color status unsigned file_id; 54 2 valid .PIC file=12345 int ramp2_min, ramp2_max; 56 2*2 LUT2 ramp min. and max. unsigned color2; 60 2 LUT2 color status BOOL edited; 62 2 image has been edited=TRUE(1) int _lens; 64 2 Integer part of lens magnification float mag_factor; 66 4 4 byte real mag. factor (old ver.) unsigned dummy[3]; 70 6 NOT USED (old ver.=real lens mag.) ------------------------------------------------------------------------------ Additional information about the HEADER structure: Bytes Description Details ------------------------------------------------------------------------------ 0-9 nx, ny, npic, ramp1_min, ramp1_max; (all are 2-byte integers) 10-13 notes NOTES are present in the file, otherwise there are none. NOTES follow immediately after image data at the end of the file. Each note os 96 bytes long. 14-15 byte_format Read as a 2 byte integer. If this is set to 1, then each pixel is 8-bits; otherwise pixels are 16-bits. 16-17 n Only used in COMOS/SOM when the file is loaded into memory. 18-49 name The name of the file (without path); zero terminated. 50-51 merged see Note 1. 52-53 colour1 54-55 file_id Read as a 2 byte integer. Aways set to 12345. Just a check that the file is in Bio-Rad .PIC format. 56-59 ramp2_min/max Read as 2 byte integers. 60-61 color2 Read as a 2 byte integer. 62-63 edited Not used in disk files. 64-65 int_lens Read as a 2 byte integer. Integer part of the objective lens used. 66-69 mag_factor Read as a 4-byte real. mag. factor=(float)(dispbox.dy*2)/(float)(512.0*scandata.ly) where: dispbox.dy = the width of the image. scandata.ly = the width of the scan region. the pixel size in microns can be calculated as follows: pixel size = scale_factor/lens/mag_factor where: lens = the objective lens used as a floating pt. number scale_factor = the scaling number setup for the system on which the image was collected. 70-75 dummy[3] Last 6 bytes not used in current version of disk file format. (older versions stored a 4 byte real lens mag here.) ------------------------------------------------------------------------------ Note 1 : Values stored in bytes 50-51 : 0 : Merge off 1 : 4-bit merge 2 : Alternate 8-bit merge 3 : Alternate columns merge 4 : Alternate rows merge 5 : Maximum pixel intensity merge 6 : 256 colour optimised merge with RGB LUT saved at the end of each merge. 7 : As 6 except that RGB LUT saved after all the notes. Information about NOTE structure and the RGB LUT are not included in this file. Please see the Bio-Rad manual for more information. ============================================================================== Info added by Geert Meesen from MRC-600 and MRC-1024 Manuals. ------------------------------------------------------------- Note Structure : Bytes Description Details ------------------------------------------------------------------------------ 0-1 Display level of this note 2-5 =0 if this is the last note, else there is another note (32 bit integer) 10-11 Note type := 1 for live collection note, := 2 for note including file name, := 3 if note for multiplier file, := 4, 5, etc.,; additional descriptive notes 16-95 Text of note (80 bytes) ============================================================================= Info added by Geert Meesen from personal experiments. ------------------------------------------------------------ - Until now I only have experience with 8-bit images from the MRC-1024 confocal microscope. The newer microscopes (Radiance 2000, for example) are capable of generating 16 bit images, I think. I have access to such a microscope and will try to find out later. For now it should be possible to look at the byte-word flag in the header. - I have experience with two types of images : --- One slice in the Z-direction, 3 channels of recording. This type is stored as a three-slice image with the 3 channels in consecutive layers. (Single-Slice) --- Different Z slices with only one channel. (Z-stack) - The header should contain some info about the pixel-size, but until now I was not really able to interpret this info. It's easier to extract the info from the notes at the end. You can find 3 notes saying something like (from AUPCE.NOT, a Z-stack file) AXIS_2 001 0.000000e+00 2.999667e-01 microns AXIS_3 001 0.000000e+00 2.999667e-01 microns AXIS_4 001 0.000000e+00 1.000000e+00 microns AXIS_9 011 0.000000e+00 1.000000e+00 RGB channel These lines give the pixelsize for the X (axis_2), Y (axis_3) and Z (axis_4) axis in the units mentioned. I don't know if this unit is always 'microns'. For a Single-Slice images you get ( from AB003A.NOT, a Single-Slice image) : AXIS_2 001 0.000000e+00 1.799800e+00 microns AXIS_3 001 0.000000e+00 1.799800e+00 microns AXIS_4 011 0.000000e+00 1.000000e+00 RGB channel It seems that AXIS_4 is used for indicating an RGB channel image. */