> For the complete documentation index, see [llms.txt](https://userguide.deepq.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://userguide.deepq.ai/use-cases/train-test-ai-model/advanced-tips.md).

# Advanced Tips

## Dicom Specialized Training (SSDL) <a href="#id-414-dicom-specialized-training-ssdl" id="id-414-dicom-specialized-training-ssdl"></a>

#### Smart Scan Deep Learning (SSDL) <a href="#smart-scan-deep-learning-ssdl" id="smart-scan-deep-learning-ssdl"></a>

Smart Scan Deep Learning (SSDL) is a advanced technique that can optimize the training results of Dicom images. SSDL will automatically activate when Dicom format in the dataset are detected.

#### Prevent Data Leakage <a href="#prevent-data-leakage" id="prevent-data-leakage"></a>

Data Leakage during training happens when the train/validate split is not handled carefully, causing some of the validation data contains information from the training data. As a result the validation score might overperform and be misleading.

In medical scans with image series, information might be shared between slices in a single scan. Therefore, splitting data by study prevents data leakage, as shown in the figure below.

<figure><img src="https://2559985325-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FEU1efCcxGpwM5PoOPuQP%2Fuploads%2Fgit-blob-db2cbeafa2093bd8bace1e6db8fc16a49dcb7a39%2FAI_Training_Model_SSDL_1.jpg?alt=media" alt="SSDL prevents data leakage in sequential images through &#x22;split by study&#x22;"><figcaption><p>SSDL prevents data leakage in sequential images through "split by study"</p></figcaption></figure>

#### Activating SSDL <a href="#activating-ssdl" id="activating-ssdl"></a>

When the user selects a dataset in training tasks, the system will detect whether the dataset contains sequential information and activate SSDL automatically.

## AI Assisted Annotation <a href="#id-431-ai-assisted-annotation" id="id-431-ai-assisted-annotation"></a>

An algorithm trained to classify Chest X-ray images cannot help in segmenting tumors in brain MRI since each Deep learning algorithm is task-oriented. AI assisted annotation must start with an neural network trained specifically for the same purpose.

![](https://2559985325-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FEU1efCcxGpwM5PoOPuQP%2Fuploads%2Fgit-blob-256e0052c23d4ce69d3fe7e6085e52b44ea6f0f4%2FAI_Training_Inf_AIA_0.png?alt=media)

#### AI Training <a href="#ai-training" id="ai-training"></a>

* Create an prediction only batch inference.
* Select the dataset from the inference list, and click on the "EXPORT TO ANNOTATE" tab to start the AI assisted annotation task.

![](https://2559985325-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FEU1efCcxGpwM5PoOPuQP%2Fuploads%2Fgit-blob-350b30a3ff7228f93f056acc90742aea95a66c7a%2FAI_Training_Inf_AIA_1.png?alt=media)

* Name the annotation data that you want to export.

![](https://2559985325-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FEU1efCcxGpwM5PoOPuQP%2Fuploads%2Fgit-blob-9c9675097c462cb557ff58a1f004d41338ae5aab%2FAI_Training_Inf_AIA_2.png?alt=media)

* The column below shows the dataset details, where you can make sure if you have select the right one.

![](https://2559985325-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FEU1efCcxGpwM5PoOPuQP%2Fuploads%2Fgit-blob-f7e8c5a4e5af49987b3c6cded936a60c6cec118e%2FAI_Training_Inf_AIA_3.png?alt=media)

#### DeepCap <a href="#deepcap" id="deepcap"></a>

* Back to DeepCap, and create an annotation project.
* Select the dataset from the drop list, enable the option "Include annotation data", and choose the one that have been annotated by AI.

![](https://2559985325-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FEU1efCcxGpwM5PoOPuQP%2Fuploads%2Fgit-blob-e4485a7874f03d52633ce318464d5b120fabb90e%2FAI_Training_Inf_AIA_4.png?alt=media)