Abstract:Spatial cluster analysis (SCA) offers valuable insights into biological images; a common SCA technique is sliding window analysis (SWA). Unfortunately, SWA's computational cost hinders its application to larger images, limiting its use to small-scale images. With advancements in high-resolution microscopy, images now exceed the capabilities of previous SWA approaches, reaching sizes up to 70,000 by 85,000 pixels. To overcome these limitations, this paper introduces the parallel integral image approach to SWA, surpassing previous methods. We achieve a remarkable speedup of 131,806x on small-scale images and consistent speedups of over 10,000x on a variety of large-scale microscopy images. We analyze the computational complexity advantages of the parallel integral image approach and present experimental results that validate the superior performance of integral-image-based methods. Our approach is made available as an open-source Python PIP package available at <a class="link-external link-https" href="https://github.com/OckermanSethGVSU/BioPII" rel="external noopener nofollow">this https URL</a>.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: **The computational cost of Sliding Window Analysis (SWA) in processing large - scale biological images is too high, resulting in limited applications**. Specifically, Sliding Window Analysis (SWA) is a commonly used Spatial Clustering Analysis (SCA) technique and is widely used in biological image analysis. However, with the development of high - resolution microscopy technology, the size of biological images has increased dramatically, for example, reaching 70,000 x 85,000 pixels. The computational cost of the traditional SWA method on such large - scale images is very high, and it is difficult to complete the analysis within a reasonable time, thus limiting its use in practical applications. To solve this problem, this paper introduces an SWA method based on the parallel integral image technique, aiming to significantly accelerate the computational process of SWA so that it can efficiently process large - scale biological images. Through this method, the author has achieved the following improvements: 1. **Reduction of algorithm complexity**: The computational complexity of traditional SWA is \(O((r - w + 1)\times(c - w + 1)\times w^{2})\), where \(r\) and \(c\) are the number of rows and columns of the image respectively, and \(w\) is the window size. The complexity of the integral image method is \(O(r\times c)+ 4(r\times c)\) and is not affected by the window size. Through parallelization, the complexity is further reduced to \(O(r\times c)/p+4(r\times c)/p\), where \(p\) is the number of parallel operations. 2. **Significant speed improvement**: Experimental results show that for small images, the new method has achieved a 131,806 - fold speed improvement; for various large - scale microscope images, the speed improvement is more than 10,000 times. 3. **Open - source code tool**: The author has implemented this method as a Python PIP package named BioPII and released it as open - source, so that other researchers can easily integrate it into existing biological image analysis software. In summary, the main objective of this paper is to overcome the computational bottleneck of traditional SWA in processing large - scale biological images by introducing the parallel integral image technique, thereby improving the efficiency and practicality of spatial clustering analysis.

Accelerating Biological Spatial Cluster Analysis with the Parallel Integral Image Technique

A Scalable Parallel Software Volume Rendering Algorithm for Large-Scale Unstructured Data

Parallel Scan Hyperspectral Fluorescence Imaging System and Biomedical Application for Microarrays

Scalable analysis of Big pathology image data cohorts using efficient methods and high-performance computing strategies

ParSymG: a Parallel Clustering Approach for Unsupervised Classification of Remotely Sensed Imagery

Virtual alignment of pathology image series for multi-gigapixel whole slide images

Parallel Processing of Image Databases for Accelerated Morphological Analysis

Rapid Image Deconvolution and Multiview Fusion for Optical Microscopy.

High-throughput Analysis of Large Microscopy Image Datasets on CPU-GPU Cluster Platforms

Parallel Multiple Instance Learning for Extremely Large Histopathology Image Analysis.

Parallelized analysis of spatial gene expression patterns by database integration

SCIMAP: A Python Toolkit for Integrated Spatial Analysis of Multiplexed Imaging Data

Accelerating iterative deconvolution and multiview fusion by orders of magnitude

Parallel Versus Distributed Data Access for Gigapixel-Resolution Histology Images: Challenges and Opportunities

Pixel-based statistical analysis by a 3D clustering approach: Application to autoradiographic images

Benchmarking Embedding Aggregation Methods in Computational Pathology: A Clinical Data Perspective

Accelerating Sensitivity Analysis in Microscopy Image Segmentation Workflows

Spatial Pattern Analysis using Closest Events (SPACE)—A Nearest Neighbor Point Pattern Analysis Framework for Assessing Spatial Relationships from Digital Images

Squidpy: a scalable framework for spatial omics analysis

SCIP: A scalable, reproducible and open-source pipeline for morphological profiling of image cytometry and microscopy data

Work-stealing prefix scan: Addressing load imbalance in large-scale image registration