> For the complete documentation index, see [llms.txt](https://docs.biodock.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.biodock.ai/start/path-1-ai-assisted-analysis.md).

# Path 1: AI-assisted analysis

*What you'll get: your objects labeled with AI help and measured, on a few images, without training a model.*

{% hint style="info" %}
Want to run the same analysis on many images, or on whole slides? [Train a model](https://docs.biodock.ai/start/path-2-train-a-fully-automated-ai-model.md) instead. In AI-assisted analysis you check every object yourself, so it suits small image sets.
{% endhint %}

## 1. Upload your images

*Already uploaded? Skip this step.*

Click **Files** in the sidebar, open the folder you want, and click **Upload files**. You can upload from your computer or from Google Drive, Dropbox, Box or OneDrive. See [Uploading images](https://docs.biodock.ai/files/uploading-images.md).

## 2. Start the analysis

<figure><img src="https://3806122466-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fk3DfNQKoMy1JXdaDLvpG%2Fuploads%2Fgit-blob-34f5f1312be4d3746e399022e629762041f1042f%2Fdocs-start-ai-assisted-new-run.gif?alt=media" alt="Analysis is opened, Start new analysis is clicked, the AI-assisted card is picked, Add files opens Choose images, two images are ticked and Confirm selection is clicked, the run is named in the Job manifest and Start labeling session is clicked"><figcaption><p><strong>Start new analysis</strong> → <strong>AI-assisted</strong> → <strong>Add files</strong> → name the run → <strong>Start labeling session</strong>.</p></figcaption></figure>

1. Click **Analysis** in the sidebar, then **Start new analysis**.
2. Under **Analysis type**, click **AI-assisted**. No trained model is needed.
3. Under **Input files**, click **Add files**, tick your images or folders, and click **Confirm selection**.
4. In the **Job manifest** on the right, type a name for the run.
5. Click **Start labeling session**.

The run opens on its **Label** tab, where you label your objects.

## 3. Add your first class

A class is the kind of object you are looking for, such as "Nucleus".

1. In the left sidebar, click the **+** button next to **Classes**.
2. Type a **Class name** and pick a **Class color**.
3. Under **Segmentation type**, choose **Detect each object** (things you count one by one, such as cells) or **Detect each region** (areas, such as inflamed tissue). Click **Next**.
4. Drag a tight box around one typical, average-sized example of the object. This sets the tile size, so the AI sees your objects at the right scale.
5. Click **Create class**.

Objects of very different sizes, such as organoids and the cells inside them, should be separate classes, because each class has its own tile size.

## 4. Label your objects

Click your class in the sidebar. Three tools appear: **AI Detect**, **AI Select** and **Pen**.

* **AI Detect** is the fastest: box one example and it outlines the look-alikes in the tile. Filter its guesses, then click **Accept**. It works for object classes only.
* **AI Select** fills the gaps: click an object, and the AI draws its outline.
* **Pen** is for anything the AI gets wrong: draw the outline yourself.

Hold **Space** and drag to move around the image. Press **Esc** to put a tool down. Everything saves as you go.

How to use each tool, and how to select, edit and delete objects, is in [Labeling tools](https://docs.biodock.ai/projects-and-models/labeling-tools.md). Every keyboard shortcut is in [Labeling shortcuts](https://docs.biodock.ai/projects-and-models/labeling-shortcuts.md).

#### Other things you can change while labeling

* **Switch image:** use the image picker under **Images** at the top of the sidebar, or **← Prev** and **Next →**.
* **Add images:** click **Add images** next to **Images**, then **Upload from local or cloud** or **Choose from Biodock filesystem**.
* **Change a class:** click the gear on the class for **Change tile size**, **Rename class**, **Change color** or **Delete class**. Deleting a class deletes every object labeled with it.
* **Brightness and contrast:** click **Channels**. This only changes your view, not what is measured.

## 5. Measure your objects

When every object is labeled, click **Quantify objects** under **Actions** in the sidebar.

1. In the **Choose analysis metrics** window, tick what to measure for your object and region classes, such as size, shape and channel intensity. More metrics take longer.
2. Click **Run analysis**.

You can label more objects later and quantify again. Each new run replaces the results of the last one. Quantifying uses credits; see [Credits, storage and limits](https://docs.biodock.ai/account/credits-storage-and-limits.md).

## 6. See your results

When the run finishes, the sidebar shows **Results ready**. Click **View** to open the **Dashboard** tab. Download your numbers from the **Export data** tab. Both tabs stay greyed out until you have run **Quantify objects**.

To read the dashboard, see [Results dashboard overview](https://docs.biodock.ai/analysis-and-results/results-dashboard-overview.md).

{% hint style="info" %}
**Coming back later?** Open **Analysis**, click your run, and open its **Label** tab to carry on labeling.
{% endhint %}


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.biodock.ai/start/path-1-ai-assisted-analysis.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
