MOTIVATION: Spatial omics data provides unprecedented insights into disease biology, yet its complexity introduces significant challenges in data analysis. Comprehensive analysis requires frameworks that integrate diverse modalities and enable joint processing of multiple datasets and corresponding metadata. RESULTS: To address these challenges, we introduce InSituPy, a versatile and scalable framework for analyzing spatial omics data from the multi-sample level down to the cellular and subcellular level. Its hierarchical data structure organizes all relevant data modalities per sample and links them to their corresponding metadata, enabling scalable analysis of large patient cohorts using spatial omics technologies. Interactive visualization tools within InSituPy enable seamless integration of histopathological expertise, promoting collaborative hypothesis generation in translational research. Additionally, InSituPy includes built-in analytical algorithms and interfaces with external tools, establishing a standardized workflow for multi-sample spatial omics data analysis. AVAILABILITY: The Python package InSituPy is publicly available on GitHub (https://github.com/SpatialPathology/InSituPy) and PyPi (https://pypi.org/project/insitupy-spatial/), and archived on Zenodo (DOI: 10.5281/zenodo.18459471). Tutorials and documentation for InSituPy are available at https://insitupy.readthedocs.io/. All code to replicate the results shown in this manuscript can be found in the GitHub repository. Scripts to connect QuPath and InSituPy can be found at https://github.com/SpatialPathology/InSituPy-QuPath. All data required to complete the tutorials is publicly available, and functions to download the data have been implemented. A Zulip community chat for user support and discussion is accessible at https://insitupy.zulipchat.com. CONTACT: [email protected], [email protected].

Registering pre-operative modalities, such as magnetic resonance imaging or computed tomography, to ultrasound images is crucial for guiding clinicians during surgeries and biopsies. Recently, deep-learning approaches have been proposed to increase the speed and accuracy of this registration problem. However, all of these approaches need expensive supervision from the ultrasound domain. In this work, we propose a multitask generative framework that needs weak supervision only from the pre-operative imaging domain during training. To perform a deformable registration, the proposed framework translates a magnetic resonance image to the ultrasound domain while preserving the structural content. To demonstrate the efficacy of the proposed method, we tackle the registration problem of pre-operative 3D MR to transrectal ultrasonography images as necessary for targeted prostate biopsies. We use an in-house dataset of 600 patients, divided into 540 for training, 30 for validation, and the remaining for testing. An expert manually segmented the prostate in both modalities for validation and test sets to assess the performance of our framework. The proposed framework achieves a 3.58 mm target registration error on the expert-selected landmarks, 89.2% in the Dice score, and 1.81 mm 95th percentile Hausdorff distance on the prostate masks in the test set. Our experiments demonstrate that the proposed generative model successfully translates magnetic resonance images into the ultrasound domain. The translated image contains the structural content and fine details due to an ultrasound-specific two-path design of the generative model. The proposed framework enables training learning-based registration methods while only weak supervision from the pre-operative domain is available.

Histopathology as a diagnostic mainstay for tissue evaluation is strictly a 2D technology. Combining and supplementing this technology with 3D imaging has been proposed as one future avenue towards refining comprehensive tissue analysis. To this end, we have developed a laboratory-based X-ray method allowing for the investigation of tissue samples in three dimensions with isotropic volume information. To assess the potential of our method for micro-morphology evaluation, we selected several kidney regions from three patients with cystic kidney disease, obstructive nephropathy and diabetic glomerulopathy. Tissue specimens were processed using our in-house-developed X-ray eosin stain and investigated with a commercial microCT and our in-house-built NanoCT. The microCT system provided overview scans with voxel sizes of [Formula: see text] and the NanoCT was employed for higher resolutions including voxel sizes from [Formula: see text] to 210 nm. We present a methodology allowing for a precise micro-morphologic investigation in three dimensions which is compatible with conventional histology. Advantages of our methodology are its versatility with respect to multi-scale investigations, being laboratory-based, allowing for non-destructive imaging and providing isotropic volume information. We believe, that after future developmental work this method might contribute to advanced multi-modal tissue diagnostics.

Staining and scanning of tissue samples for microscopic examination is fraught with undesirable color variations arising from differences in raw materials and manufacturing techniques of stain vendors, staining protocols of labs, and color responses of digital scanners. When comparing tissue samples, color normalization and stain separation of the tissue images can be helpful for both pathologists and software. Techniques that are used for natural images fail to utilize structural properties of stained tissue samples and produce undesirable color distortions. The stain concentration cannot be negative. Tissue samples are stained with only a few stains and most tissue regions are characterized by at most one effective stain. We model these physical phenomena that define the tissue structure by first decomposing images in an unsupervised manner into stain density maps that are sparse and non-negative. For a given image, we combine its stain density maps with stain color basis of a pathologist-preferred target image, thus altering only its color while preserving its structure described by the maps. Stain density correlation with ground truth and preference by pathologists were higher for images normalized using our method when compared to other alternatives. We also propose a computationally faster extension of this technique for large whole-slide images that selects an appropriate patch sample instead of using the entire image to compute the stain color basis.