Table of Contents

Advanced Data Analysis: Calculating Image Stacks and ROIs

Script
Function Overview

Key Parameters:

Step-by-Step Explanation

Step 1: Caching for Efficiency
Step 2: Initializing the Scan
Step 3: Processing Data Points
Step 4: Finalizing ROI Data

Customization Tips
Conclusion

In the next part of your tutorial, we'll focus on a core data analysis function: calculate_stack_and_rois. This function is essential for processing experimental data, calculating image stacks, and extracting regions of interest (ROIs) based on specific criteria. This section will guide users through understanding, utilizing, and customizing this function within the broader context of your data analysis workflow.

Advanced Data Analysis: Calculating Image Stacks and ROIs

In this section, we delve into a vital aspect of our data processing pipeline: calculating image stacks and extracting regions of interest (ROIs) from our experimental data. The function calculate_stack_and_rois plays a great role in this process, leveraging the SFData library to interact with experimental data and the Cristallina library for specific analysis tasks.

Script

@memory.cache
def calculate_stack_and_rois(run_number, rois, lower_cutoff_threshold=7.6, max_step=None, calc_stacks=True):

    scan = SFScanInfo(f"/sf/cristallina/data/p21640/raw/run{run_number:0>4}/meta/scan.json")
    
    angles = []
    
    for roi in rois:
        roi.values = deque()

    stacks = []
    
    for i, step in tqdm(enumerate(scan)):
        if max_step is not None:
            if i > max_step:
                break
                
        subset = step["JF16T03V01",]
        
        files = [step.files[0].fname, step.files[1].fname, step.files[2].fname]

        if calc_stacks:
            stack = cr.analysis.perform_image_stack_sum(files, channel='JF16T03V01', lower_cutoff_threshold=lower_cutoff_threshold)
            stacks.append(stack)
    
        for roi in rois:
            roi.values.append(cr.analysis.perform_image_roi_crop(files, roi=roi))

        angles.append(scan.readbacks[i])
    
    for roi in rois:
        roi.values = np.array(roi.values)
        roi.values = np.where(roi.values < lower_cutoff_threshold, 0, roi.values)
        roi.intensities = np.sum(roi.values[:], axis=(1, 2, 3))

    return angles, stacks, rois

Function Overview

calculate_stack_and_rois takes a run number and a list of ROI definitions as inputs, along with optional parameters to fine-tune the analysis. It processes data from a specified experiment run, computes image stacks if required, and extracts intensity values for each ROI. The function returns a list of angles associated with each data point, the computed image stacks (if calculated), and the updated ROIs with their calculated intensities.

Key Parameters:

run_number: Identifier for the experiment run to process.
rois: List of ROI objects defining areas of interest.
lower_cutoff_threshold: Intensity threshold for considering pixel values in the analysis.
max_step: Maximum number of data points to process (useful for debugging or processing subsets of data).
calc_stacks: Boolean indicating whether to calculate image stacks.

Step-by-Step Explanation

Step 1: Caching for Efficiency

The @memory.cache decorator is used to cache the results of this function. This means that repeated calls with the same arguments will return instantly, significantly speeding up re-analysis or iterative development processes.

Step 2: Initializing the Scan

The function starts by creating an SFScanInfo object, which provides access to metadata for the specified experiment run:

scan = SFScanInfo(f"/sf/cristallina/data/p21640/raw/run{run_number:0>4}/meta/scan.json")

*(make sure to change data folder to the right pgroup)

Step 3: Processing Data Points

For each step in the scan, the function checks if it should continue (based on max_step), computes image stacks (if calc_stacks is true), and extracts ROI data:

Image Stack Calculation: Uses cr.analysis.perform_image_stack_sum to sum images, applying a lower cutoff threshold.
ROI Extraction: For each ROI, cr.analysis.perform_image_roi_crop is called to extract and process the relevant area from the images.

Step 4: Finalizing ROI Data

After processing all steps, the function finalizes the ROI data by converting values to arrays, applying the cutoff threshold, and calculating total intensities for each ROI.

Customization Tips

ROI Definitions: Customize the rois parameter to focus on specific areas of interest in your data.
Threshold Adjustments: The lower_cutoff_threshold can be adjusted based on the signal-to-noise ratio or other experimental considerations.
Partial Data Processing: Use max_step to limit analysis to a subset of the data, useful for testing or when working with large datasets.

Conclusion

The calculate_stack_and_rois function is a versatile tool in our data analysis process, providing detailed insights into specific regions of interest across experimental runs. By understanding and leveraging this function, users can extract valuable quantitative information from the data.