Guides

Results Format

Description of the default output results (.xlsx) format

Introduction

Numeric results generated by MIA workflows are exported to Excel spreadsheets. By default, each time a workflow is run, the results for all images processed will be collected into a single spreadsheet, however, this behaviour can be changed such that each input file generates its own spreadsheet.

Output results files can contain multiple, highly configurable sheets. Typically, a results file will include a summary sheet with one row per input image, individual object sheets with one row per detected object and a couple of record-keeping sheets.

The output file format is configured in the Output control module.

Summary sheet

  • The summary sheet is intended to provide a quick overview of the measurements contained within the individual object sheets. Information is broken down into four classes of column.
    • By default, each row of the summary sheet will summarise the results from a single input image.
    • By changing the "Summary mode" parameter of the "Output control" module to "Per timepoint per input file", a new row can be generated for each timepoint in the image, allowing for quick analysis of time-dependent processes.
    • Alternatively, by setting this to "Group by metadata", results can be grouped by a specific metadata item, with one row per unique metadata value.
  • Metadata values:
    • Record input file-specified properties, such as filenames and series numbers.
    • These are important for result filtering and sorting, since results from multiple files may be included in a single results file.
    • Column headings simply state the name of the metadata item.
    • Metadata columns can be enabled/disabled using the "Ind" controls in the "Output control" module.
  • Image measurements:
    • These columns record the individual measurements assigned to each object.
    • Columns headings are in the form:[Image name]_(IM) // [Measurement name]
    • For example, an image called "Binary" with the measurement "THRESHOLD // GLOBAL Huang" would be called "Binary_(IM) // THRESHOLD // GLOBAL Huang".
    • Image measurements columns can be enabled/disabled using the "Ind" controls in the "Output control" module.
  • Object counts:
    • These simply state the number of objects of the specified class.
    • These have headers in the form:[Objects name]_(OBJ) // NUMBER
    • For example, for the count of detected "Nuclei" objects, this would be "Nuclei_(OBJ) // NUMBER".
    • Object count columns can be enabled/disabled using the "Show object counts" control in the "Output control" module. Object counts are either on or off for all objects.
  • Object measurement statistics:
    • Since multiple objects may be detected for an input file, each with their own measurements, these columns provide a single summary measurement statistic.
    • Headers for these values are in the form:[Objects name]_(OBJ_[Statistic]) // [Measurement name]
    • For example, the header corresponding to the mean of the "SHAPE // N_VOXELS" measurement for "Nuclei" objects would be "Nuclei_(OBJ_MEAN) // SHAPE // N_VOXELS".
    • Supported statistics are the mean ("MEAN"), minimum ("MIN"), maximum ("MAX"), standard deviation ("STD") and sum ("SUM").
    • Object measurement statistic columns can be enabled/disabled using the "Mean", "Min", "Max", "Sum" and "Std" controls in the "Output control" module. The "Ind" control determines if the measurement appears in the object-specific results sheet.

Object sheets

  • The individual object sheets include the individual measurements for each detected object. Each row of an object results sheet corresponds to a single detected object.
  • The first column is always ("OBJECT_ID"), which is the unique ID number for that object in that analysis run.
    • Note: IDs are only unique to that analysis run, so if multiple input files were processed, the same ID numbers will likely appear.
  • As with the summary sheet, each row has columns for file-specific metadata items as well as a column for each exported measurement.
  • Metadata and measurement columns can be disabled with the "Ind" controls in the "Output control" module.

Record-keeping sheets

  • The final two sheets are always record-keeping sheets. These are useful for retaining knowledge of how a workflow was run (e.g. plugin version) as well as seeing if any errors occurred during execution.
  • The "Configuration" sheet includes some device-specific information such as the computer operating system, plugin version and date/time of execution. It also includes a full copy of the .mia workflow file used. In the event of the workflow file being lost, these rows can be copied into a text file and saved with the .mia extension.
  • The "Log" sheet contains a list of any error messages encountered during the analysis run. All being well, this sheet should be empty.