The LA-iMageS program is a graphical application that automates the data processing and elemental distribution visualization of LA–ICP–MS bioimaging. The overall architecture of LA-iMageS can be seen in Fig. 1.
Input data
LA-iMageS uses datasets in PerkinElmer Elan XL format (*.xl) as input, which is commonly generated by ICP–MS instrument control software from PerkinElmer, such as Elan 6 × 00 or Elan DRC-e. Each input dataset should be placed in a folder containing the XL files corresponding to each data line taken by the ICP–MS instrument. Each line file must contain a number that indicates the order in which it has been acquired by the ICP–MS instrument. For instance, in a dataset with ten lines, a valid set of names can be: line 1.xl, line 2.xl, line 3.xl, line 4.xl, line 5.xl, line 6.xl, line 7.xl, line 8.xl, line 9.xl, and line 10.xl. Additionally, the dataset folder may include two optional files: parameters.conf, containing the ICP–MS data acquisition parameters, and positions.txt, containing the physical position of each line.
Data acquisition parameters
The optional parameters.conf file is used by the LA-iMageS software to automatically load acquisition parameters. If this file is not present in the dataset’s directory, the user must manually introduce them. However, experience demonstrates that it is a good practice to save the acquisition parameters along with data line files.
The data acquisition parameters that can be specified in this file are the following:
-
Standard the standard element in the dataset. Internal standard is an element used to normalize the results and to overcome instrumental oscillation. The other elements will be normalized using this element as standard. Users must choose a specific element to be monitored, for example, a known matrix element in the sample or those intentionally added by the user. The chosen element should minimize the variability of the ablation process, which can be caused by local differences in tissue thickness and/or different interaction between the laser and the sample surface, allowing the observed signal to correspond to an elemental concentration in a specific location.
-
Ablation speed speed set by the user in the laser operational mode as continuous firing. This parameter relates to the spot size, which is determined by the laser beam diameter, and by the frequency, which corresponds to the repetition rate of the laser. Usually, the value of ablation speed used is lower than spot size.
-
Acquisition time refers to the time needed for the acquisition of one point considering all the elements monitored by the ICP–MS. This parameter is intrinsically correlated to ICP–MS parameters, such as the number of isotopes monitored, sweeps, number of replicates and dwell (or residence) time. The acquisition time should not be higher than 1.0 s, since the elemental distribution information would be lost.
-
Space interval represents the distance among the center of two lines. The lowest space interval results in the highest image resolution.
Data lines positions
The positions.txt file can be used to specify the physical position of each line in the instrument during the ablation process. This information is important for the elemental data extraction process explained below, since data acquisition can be made with laser position in the horizontal or vertical profile.
This file is optional and is used by the LA-iMageS software to read the position of each line and determine whether they are horizontal or vertical: if each line has the same X position while the initial and final Y-positions are different, it means that it is vertical; otherwise, if each line has the same Y-position while the initial and final X-positions are different, it means that it is horizontal. If this file is not present in the dataset’s directory, then LA-iMageS will consider the lines to be horizontal and will automatically generate their positions based on the acquisition parameters.
Since this file can be easily generated by ICP–MS instrument control software, we strongly encourage keeping a positions file along with the data lines files.
Elemental data extraction
Elemental data is extracted from the input dataset in XL format, using the two optional configuration files if necessary. LA-iMageS parses input data in order to obtain one two-dimensional matrix per element in the dataset, which stores the analyte distribution in the sample. Acquisition parameters are used along each positions file for axis definition.
For instance, when data is acquired using the laser in the horizontal position, acquisition time and ablation speed parameters are used for x-axis definition: line measurements are separated by intervals of acquisition time multiplied by ablation speed. Conversely, the y-axis is simply defined by the spacing among the lines (i.e., space interval parameter). When data is acquired using the laser in the vertical position, the x-axis and y-axis are opposite to those in the horizontal mode, as Fig. 2 illustrates.
After the elemental data extraction process, each element is normalized by the specified standard element, dividing its intensity matrix by that of the standard.
Data visualization
The main user interface of LA-iMageS (Fig. 3) is organized into three main sections: the Toolbar, the Clipboard, and the Analysis viewer. Through the toolbar, users can access the main functions of LA-iMageS, where ‘data analysis’ is the most important operation. On the clipboard tree, users can find a list of loaded datasets. Finally, users can explore elemental data through the Analysis viewer panel.
The most important section is the Analysis viewer panel, which consists of a 2D/3D representation of the current element distribution with a menu bar and a right sidebar providing access to several configuration options. It is important to stress out that 3D visualization reflects the signal intensity of each analyte only, and not the real topology of the sample.
The right sidebar allows users to select the element distribution that is being currently displayed, enable or disable the 3D view, and control the camera position.
The menu bar contains three submenus: (1) File, which allows saving the current analysis; (2) Graphical Settings, which enables the customization of the elemental distribution image; and (3) Export, which provides exporting facilities.
Tuning up the elemental distribution image
Elemental distribution images can be tuned up throughout the Graphical Settings submenu. A very useful feature of this submenu is the interpolation level, since it allows creating new data points by interpolation within the range of the original set of data points, thus increasing image quality. Figure 4 illustrates how interpolation can help to improve image resolution.
Another important aspect to obtain good images is the color map adjustment. LA-iMageS allows customizing the color palette used to represent the image, as well as the range of values of the color map. Since each element has its own intensity range, this latter option is especially useful to obtain comparable images of different elemental distributions by setting a color map within the same range of values for each element (Fig. 5).
Data export
After exploring data in the LA-iMageS, it is expected that users can employ their results in other complementary applications such as a word processor or another analysis package. To fulfill these needs, both elemental distributions (i.e., the data matrices) and images can be easily saved throughout the Export submenu of the Analysis viewer panel. While elemental distributions are exported using comma-separated values (CSV) files, 2D/3D bioimages are exported into portable network graphics (PNG) files.
When exporting data, users can choose to save only the element shown in the Analysis viewer panel, or all the elements of the dataset in a row with the same export configuration.
Implementation
The LA-iMageS software is implemented in Java using the AIBench framework [20]. It is provided as a self-contained, multiplatform Java standalone application. LA-iMageS is an open source project hosted on Github. The code architecture is interface driven, so developers can easily integrate new data formats and/or functions. The jzy3D and the Apache Commons Math Java libraries are also integrated within the project to render 2D/3D images and to perform mathematical operations (e.g., bilinear interpolation).