Loci

Product Introduction

1. Definition: Loci is a free, open-source, local-first desktop application for biomedical image analysis, specifically categorized as a scientific image analysis workspace. It is built using modern desktop application frameworks and integrates a Python-based analysis engine for deterministic processing.
2. Core Value Proposition: Loci exists to provide life science researchers with a powerful, privacy-focused, and traceable image analysis platform that operates entirely offline, eliminating dependency on cloud subscriptions, accounts, and internet connectivity for core workflows. Its primary value is enabling secure, reproducible analysis of sensitive microscopy, histology, and 3D medical image data directly on a researcher's local machine.

Main Features

1. Local-First, Offline-Capable Architecture: All core image viewing, navigation, and built-in analysis algorithms execute entirely on the local host machine. The application is designed for air-gapped environments, with zero telemetry, silent network calls, or mandatory cloud accounts. Advanced features like remote compute require explicit user configuration.
2. Multi-Dimensional, Multi-Format Scientific Image Viewer: The workspace supports direct navigation of channel (C), Z-stack (Z), and time series (T) dimensions. It natively reads a wide array of formats including multi-series OME-TIFF, OME-Zarr (NGFF), Aperio SVS whole-slide images, DICOM, and NIfTI volumes, handling gigapixel pyramids and 3D data without flattening or pre-processing.
3. Integrated Classical and AI-Powered Analysis Engine: Loci includes deterministic classical image analysis algorithms (Otsu/Yen/Li thresholding, adaptive watershed segmentation) and provides a managed framework for user-provisioned deep learning models. It supports official Cellpose-SAM checkpoints and compatible ONNX model packages, running them locally with explicit version and digest verification for reproducibility.
4. Traceable, Audit-Ready Workflow & Export: Every analysis step is logged with parameters and software version metadata. Original source images are treated as immutable. Exports include high-bit-depth TIFF label maps, CSV measurement tables, and PNG figures, all bundled with SHA-256 source file fingerprints to create a complete provenance manifest for scientific audit and reproducibility.
5. Interactive Research Workbench with Quantitative Review: Features direct object selection in 2D/3D viewports, a linked quantitative review table for metrics (area, volume, intensity), and reusable, validated analysis recipe presets. Annotations (points, polygons, calibrated lines) are drawn directly on raw pixel coordinates.

Problems Solved

1. Pain Point: Researchers lack accessible, powerful desktop software for advanced bioimage analysis that respects data privacy, avoids vendor lock-in via subscriptions, and ensures long-term reproducibility without ongoing costs.
2. Target Audience: Primary users are academic and industry biomedical researchers, lab scientists, and clinician-scientists working with microscopy (fluorescence, confocal), digital pathology, and 3D medical imaging (CT, MRI). Secondary users include core facility managers and bioimage informaticians seeking standardized, traceable analysis pipelines.
3. Use Cases: Essential for: 1) Analyzing sensitive human tissue image data under strict data governance (GDPR, HIPAA) in an air-gapped lab; 2) Reproducible cell counting and morphology measurement in fluorescence microscopy for publication; 3) Qualitative review and annotation of whole-slide histology images at native resolution; 4) Exploratory analysis of 3D volumetric time-series data (e.g., light-sheet microscopy) with linked MPR views.

Unique Advantages

1. Differentiation: Unlike cloud-based platforms (e.g., some commercial AI tools) or traditional, costly closed-source desktop software (e.g., Imaris, Volocity), Loci is completely free, open-source, and local-first. Unlike other open-source tools like ImageJ/Fiji, it offers a modern, integrated desktop workspace with explicit support for multi-dimensional data, reproducible recipe presets, and managed AI model execution out-of-the-box.
2. Key Innovation: The integration of a managed, verifiable AI model runtime within a local-first architecture is a key innovation. Researchers can bring their own compatible ONNX or Cellpose-SAM models, with the software verifying cryptographic digests against known-good versions, enabling cutting-edge AI analysis while maintaining full data control, auditability, and offline operation—a combination typically not found in unified desktop tools.

Frequently Asked Questions (FAQ)

1. Is Loci really free to use for academic and commercial research? Yes, Loci is completely free and open-source, released under a permissive license that allows unrestricted use in academic, commercial, and government research. There are no subscriptions, tiered plans, or hidden costs.
2. How does Loci handle data privacy and security for sensitive clinical images? Loci is architected as a local-first application. It makes no network connections for core functionality, keeps all data on the user's machine, and strips local filesystem paths from exported manifests. This makes it suitable for analyzing sensitive data under HIPAA or GDPR in secure/air-gapped environments.
3. What image formats are supported for digital pathology and 3D imaging? Loci natively supports whole-slide images like Aperio SVS and Hamamatsu NDPI, and 3D medical volumes in DICOM and NIfTI formats. It renders 3D data using direct GPU ray-casting and linked multi-planar reconstruction (MPR) views.
4. Can I use my own trained AI models in Loci? Yes, through its managed model system. You can use officially released Cellpose-SAM checkpoints or bring your own compatible ONNX model packages. The software verifies model integrity against a known cryptographic digest to ensure reproducibility.
5. Why is there a Gatekeeper warning on macOS, and is it safe to install? The initial public beta is signed with an Apple developer ad-hoc signature but is not notarized by Apple, which triggers macOS security. It is safe to install; you must approve it via System Settings → Privacy & Security → Open Anyway. This is a standard procedure for beta software from independent developers.