HSI Wizard

See Beyond the Visible: The Magic of Hyperspectral Imaging for Medical & Bioinformatics Applications

Introduction

Welcome to the hsi-wizard package! Designed primarily for medical hyperspectral imaging and bioinformatics workflows, this Python package provides a straightforward environment for hyperspectral imaging (HSI) analysis, supporting everything from basic spectral analysis to advanced machine learning and AI methods. Whether you're working with raw sensor data or pre-processed datasets, hsi-wizard offers a suite of tools to simplify and enhance your analysis workflow.

If you're new here, the best place to start is the documentation, where you'll find detailed instructions on how to begin.

Features

• DataCube Class for managing and processing HSI data
• Spectral plotting and visualization
• Clustering and spectral analytics
• Tools for merging and processing HSI data
• Data loaders for various file formats (e.g., NRRD, Pickle, TDMS, and XLSX)
• Decorators for method tracking, input validation, and execution time logging

Geospatial Limitations

This package does not track or process geographical coordinates. If you require full remote-sensing or GIS integration, you may need complementary tools. hsi-wizard focuses on medical and lab-based spectral analysis, so spatial georeferencing is restricted.

Comparison with Existing Tools

| Attribute | HSI-Wizard | PySptools | HyDe | Spectral Python (SPy) | ENVI | | -------------------- | ---------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------- | ----------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------- | ---------------------------------------------------------------- | | Category | Medical Imaging | Geospatial & Research Spectral Analysis | Preprocessing & Denoising | General-Purpose Data I/O & Basic Handling | Commercial All-rounder | | Key Features | End-to-end medical pipeline, DataCube, Merge & analytics | Endmember extraction, Spectral unmixing, Low-rank wavelet & DNN denoising, Energy-efficient implementations | Read/display/manipulate HSI files, Bip/Bil/Bsq interleaves, Basic classification routines | Advanced spectral processing & classification, Full GUI & Python API, ArcGIS integration | | | File Support | ENVI (.hdr/.img), Images, CSV, FSM, NRRD, Pickle (.pkl), TDMS, XLSX | ENVI (.hdr/.img) via Spectral Python integration | None built-in (operates on NumPy arrays from user’s loader) | ENVI (.hdr/.img), BIL, BIP, BSQ interleaves | ENVI (.hdr/.img), GeoTIFF, HDF, ASCII – broad commercial support | | Licensing / Cost | Open-source (MIT) | Open-source (MIT) | Open-source (BSD-3-Clause) | Open-source (BSD-3) | Proprietary, paid |

Installation

Requirements

• Python >3.10

Via pip

You can install the package via pip:

bash pip install hsi-wizard

Compile from Source

Alternatively, you can compile HSI Wizard from source:

bash python -m pip install -U pip setuptools wheel # Install/update build tools git clone https://github.com/BlueSpacePotato/hsi-wizard # Clone the repository cd hsi-wizard # Navigate into the directory python -m venv .env # Create a virtual environment source .env/bin/activate # Activate the environment pip install -e . # Install in editable mode pip install wheel # Install wheel pip install --no-build-isolation --editable . # Compile and install hsi-wizard

Usage

After installing the package, you can import the DataCube, read function, and plotter for quick HSI data analysis:

```python3 import wizard

Load an HSI datacube from a file

dc = wizard.read('pathtofile')

process DataCube

dc.resize(xnew=500, ynew=500) dc.remove_background()

Visualize the datacube

wizard.plotter(dc) ```

For more examples visist the documentation.

Contributing

We welcome contributions from the medical imaging and bioinformatics communities! To make the collaboration smooth and efficient, please follow the guidelines below:

Reporting Issues

If you encounter bugs, unexpected behavior, or have feature requests, please:

1. Search existing issues to see if your problem or idea has already been reported.
2. Open a new issue with a clear title and description:
   1. Steps to reproduce the problem
   2. Expected vs. actual behavior
   3. Version of hsi-wizard, Python, and operating system
3. Label the issue appropriately (e.g., bug, feature request, documentation).

Seeking Support

For usage questions, help with examples, or general discussion: - GitHub Discussions: Post your questions under the Support category in our Discussions board.

Contributing Code

If you have ideas or fixes, we’d love your help! Just fork the repo, create your branch, and make your changes. When you’re ready, push your branch up and open a Pull Request against main. We’ll review it and merge it in!