BactiScan

Are you a passionate researcher seeking a powerful and efficient solution to count bacteria in microscope images? Look no further! Introducing BactiScan - the cutting-edge app designed exclusively for scientists like you.

🔍 Explore Features:

✔️ Import images from your gallery or capture them with your camera for instant analysis.
✔️ Effortlessly rename and organize your scans, making retrieval a breeze.
✔️ Swiftly search for specific scans based on names for enhanced productivity.
✔️ Utilize the power of ImageJ.JS for unrivaled image processing, all within your device.
✔️ Rest assured, your data remains secure - no uploads to external servers!

📸 Seamless Export: After processing, effortlessly export your results to the gallery for further analysis and seamless integration with your workflow.

🌟 User-Friendly Interface: Experience the epitome of user-friendly design with BactiScan's minimalist and sleek UI. Navigate through functions with ease and concentrate on what truly matters - your research!

Unleash the full potential of your scientific endeavors with BactiScan. Download now and experience the joy of precise bacterial analysis at your fingertips!

Design Decisions:

API:

API provided by imjoy is a WASM based JavaScript port of the ImageJ Java API. It is a very powerful API and can be used to perform almost everything that ImageJ can do.
The API uses CheerpJ to compile Java code to JavaScript. The best way to use the API is to refer to the Java documentation for ImageJ and then use the JavaScript API to perform the tasks by using CheerpJ's functions to run Java code.

Mobile App:

The app is built using React Native. The app uses a webview to load the Imjoy API.

Why webview?

I decided to use a webview for the app because the ImageJ.JS API does not have a React Native implementation. The other option was to use a server to run the API and then make calls to the server from the app. However, this would require the server to be running constantly and would also require the images to be sent to the server. Furthermore, it wouldn't have made full use of the API as the API is very powerful and can be used to perform a lot of tasks and is designed to run in the browser.

Why React Native?

React Native is a cross-platform framework that allows us to build native apps for both Android and iOS using JavaScript.
It is a very powerful framework and is used by a lot of companies to build their apps.
It is also very easy to use and has a lot of community support.

Communication between React Native and webview:

The communication between React Native and the webview is done using the react-native-webview library. Transferring data between the webview and React Native can be tricky. The react-native-webview library provides a postMessage function that can be used to send data from the webview to React Native. However, the postMessage function can only send strings. To send objects, we need to use the injectJavaScript function. This function can be used to run JavaScript code in the webview. We can use this function to run a function in the webview that can then use the postMessage function to send the data to React Native.

The react-native-webview library provides an onMessage function that can be used to listen for messages from the webview. However, this function can only listen for strings. To listen for objects, we need to use the onMessageReceived function. This function can be used to run a function in the webview that can then use the postMessage function to send the data to React Native.

An RPC-like design was chosen for communication between React Native and the webview. At first, the react native side injects a javascript function into the webview which listens for messages posted to the webview. Now we have essentially created a channel of communication capable of sending messages from react native and receiving messages on the webview end. This function is designed in such a way that responses from the function are sent back to the webview using the "postMessage" function and the react native app listens for these messages using the "onMessageReceived" function.

The documentation for the Imjoy API is extensive but can be challenging to navigate. Furthermore, some of the early version of the API are very buggy. The API is available as a javascript module as well as a python package that you can simply pip install. However, I faced some challenges with using the Python API.

Some future improvements to the app are listed below:

UI Improvements:

1. Create a folder structure so that researchers can create a folder for every bacterium group they are monitoring instead of having all scans in one folder
2. Allow researchers to add tags to scans and allow for searching for tags
3. The order of images displayed after the scan is processed needs to be fixed so that the particle analysis output is displayed first instead of the mask image
4. Ability to save processed images to Google Drive as backup
5. Ability to modify the macro that the API runs in the backend -- ideally we want to allow researchers to record a macro they would like to run and simply copy-paste the macro into the app and have to run the macro for input images and also capture intermediate images and display them on the app

Performance improvements:

1. As of Jul 25, 2023, a new webview is created every time a new scan is opened in the app and only one image is processed at a time. This is done only for development purposes and ease of debugging and is not a limitation of the API. More work needs to be put in to parallelize the image processing
2. The imageJ webview can be kept permanently connected and hidden in the background. This way operations can be performed on the webview in the background.
3. The entire app can be rewritten as a progressive web app (PWA), installed on devices using Safari (for iPhones), and can completely avoid having to go through the App store. PWA capabilities need to be evaluated carefully before making this decision.
4. Currently, the app does not run when the device is not connected to the internet. But, with a few modifications, it should be possible to make the app run offline.