diff --git a/piximi-documentation/_toc.yml b/piximi-documentation/_toc.yml
index 49923e8..9f82572 100644
--- a/piximi-documentation/_toc.yml
+++ b/piximi-documentation/_toc.yml
@@ -11,6 +11,7 @@ parts:
- caption: Tutorials
chapters:
- file: translocation_tutorial
+ - file: translocation_tutorial_ES
- file: classify-example-eukaryotic-image
- file: classify-example-eukaryotic-object
- caption: How-to Guides
diff --git a/piximi-documentation/img/tutorial_images/Figure1.png b/piximi-documentation/img/tutorial_images/Figure1.png
index 6ee9809..2b866a2 100644
Binary files a/piximi-documentation/img/tutorial_images/Figure1.png and b/piximi-documentation/img/tutorial_images/Figure1.png differ
diff --git a/piximi-documentation/img/tutorial_images/Figure8.png b/piximi-documentation/img/tutorial_images/Figure8.png
index a493015..fd9ade4 100644
Binary files a/piximi-documentation/img/tutorial_images/Figure8.png and b/piximi-documentation/img/tutorial_images/Figure8.png differ
diff --git a/piximi-documentation/translocation_tutorial.md b/piximi-documentation/translocation_tutorial.md
index 593d08b..adbdc3c 100644
--- a/piximi-documentation/translocation_tutorial.md
+++ b/piximi-documentation/translocation_tutorial.md
@@ -1,6 +1,6 @@
# Piximi: Installation-free segmentation and classification in the browser
-## A computer exercise using webtool \- Piximi
+## A computer exercise using webtool - Piximi
Beth Cimini, Le Liu, Esteban Miglietta, Paula Llanos, Nodar Gogoberidze
@@ -12,7 +12,7 @@ Broad Institute of MIT and Harvard, Cambridge, MA.
Piximi is a modern, no-programming image analysis tool leveraging deep learning. Implemented as a web application at [https://piximi.app/](https://piximi.app/), Piximi requires no installation and can be accessed by any modern web browser. Its client-only architecture preserves the security of researcher data by running all\* computation locally.
-Piximi is interoperable with existing tools and workflows by supporting import and export of common data and model formats. The intuitive researcher interface and easy access to Piximi allows biological researchers to obtain insights into images within just a few minutes. Piximi aims to bring deep learning-powered image analysis to a broader community by eliminating barriers to entry.
+Piximi is interoperable with existing tools and workflows by supporting import and export of common data and model formats. The intuitive interface and easy access to Piximi allows biological researchers to obtain insights into images within just a few minutes. Piximi aims to bring deep learning-powered image analysis to a broader community by eliminating barriers to entry.
\* except for the segmentations using Cellpose, which are sent to a remote server (with the permission of the user).
@@ -20,12 +20,10 @@ Core functionalities: **Annotator, Segmentor, Classifier, Measurments.**
#### **Goal of the exercise**
-In this exercise, you will familiarize yourself with Piximi’s main functionalities of annotation, segmentation, classification, measurement and visualization and use it to analyze a sample image dataset from a translocation experiment. The goal of this experiment is to determine the **lowest effective dose** of Wortmannin required to induce GFP-tagged FOXO1A nuclear localization (Figure 1\)**.** You will segment the images using one of the deep learning models available in Piximi, check and curate the segmentation, then train an image classifier to classify the individual cells as having “nuclear-GFP”, “cytoplasmic-GFP” or “no-GFP”. Finally, you will make measurements and plot them to answer the biological question.
+In this exercise, you will familiarize yourself with Piximi’s main functionalities of annotation, segmentation, classification, measurement and visualization and use it to analyze a sample image dataset from a translocation experiment. The goal of this experiment is to determine the **lowest effective dose** of Wortmannin required to induce GFP-tagged FOXO1A nuclear localization (Figure 1). You will segment the images using one of the deep learning models available in Piximi, check and curate the segmentation, then train an image classifier to classify the individual cells as having “nuclear-GFP”, “cytoplasmic-GFP” or “no-GFP”. Finally, you will make measurements and plot them to answer the biological question.
#### **Context of the sample experiment**
-
-
In this experiment, researchers imaged fixed U2OS osteosarcoma (bone cancer) cells expressing a FOXO1A-GFP fusion protein and stained DAPI to label the nuclei. FOXO1 is a transcription factor that plays a key role in regulating gluconeogenesis and glycogenolysis through insulin signaling. FOXO1A dynamically shuttles between the cytoplasm and nucleus in response to various stimuli. Wortmannin, a PI3K inhibitor, can block nuclear export, resulting in the accumulation of FOXO1A in the nucleus.
@@ -40,7 +38,7 @@ In this experiment, researchers imaged fixed U2OS osteosarcoma (bone cancer) cel
#### **Materials necessary for this exercise**
-The materials needed in this exercise can be downloaded from: [PiximiTutorial](./downloads/Piximi_Translocation_Tutorial_RGB.zip). The “Piximi Translocation Tutorial RGB.zip” file contains a Piximi project, including all the images, already labeled with the corresponding treatment (Wortmannin concentration or Control). Download this file but **do NOT unzip it**\!
+The materials needed in this exercise can be downloaded from: [PiximiTutorial](./downloads/Piximi_Translocation_Tutorial_RGB.zip). The “Piximi Translocation Tutorial RGB.zip” file contains a Piximi project, including all the images, already labeled with the corresponding treatment (Wortmannin concentration or Control). Download this file but **do NOT unzip it**!
#### **Exercise instructions**
@@ -54,20 +52,29 @@ Read through the steps below and follow instructions where stated. Steps where y
* Load the example project: Click “Open” \- “Project” \- “Project from Zip”, as shown in figure 2 to upload a project file for this tutorial from Zip, and you can optionally change the project name in the top left panel, such as “Piximi Exercise”. As it is loaded, you can see the progression in the top left corner logo
.
-
+```{figure} ./img/tutorial_images/Figure2.png
+:width: 600
+:align: center
+
+**Figure 2**: Loading a project file.
+```
2. ##### **Check the loaded images and explore the Piximi interface**
-These 17 images represent Wortmannin treatments at eight different concentrations (expressed in nM), as well as mock treatments (0uM). Note the DAPI channel (Nuclei) is shown in magenta and that the GFP channel (FOXOA1) is shown in green.
+These 17 images represent Wortmannin treatments at eight different concentrations (expressed in nM), as well as mock treatments (0nM). Note the DAPI channel (Nuclei) is shown in magenta and that the GFP channel (FOXOA1) is shown in green.
As you hover over the image, color labels are displayed on the left corner of the images. These annotations are from metadata in the zipped file we just uploaded. In this tutorial, the different colored labels indicate the concentration of Wortmannin, while the numbers represent the number of images in each category.
Optionally, you can annotate the images manually by clicking “+ Category”, entering your label, and then selecting the image by clicking the images annotating the selected images by clicking **“Categorize”**. In this tutorial, we’ll skip this step since the labels were already uploaded at the beginning.
-
+```{figure} ./img/tutorial_images/Figure3.png
+:width: 600
+:align: center
+**Figura 3**: Exploring the images and labels.
+```
-3. ##### **Segment Cells \- find out the cells from the background**
+3. ##### **Segment Cells - find out the cells from the background**
🔴 TO DO
@@ -79,8 +86,12 @@ Optionally, you can annotate the images manually by clicking “+ Category”, e
* It will take a few minutes to finish the segmentation.
-
+```{figure} ./img/tutorial_images/Figure4.png
+:width: 600
+:align: center
+**Figura 4**: Loading a segmentation model.
+```
Please note that the previous steps were performed on your local machine, meaning your images are stored locally. However, Cellpose inference runs in the cloud, which means your images will be uploaded for processing. If your images are highly sensitive, please exercise caution when using cloud-based services.
@@ -88,9 +99,14 @@ Please note that the previous steps were performed on your local machine, meanin
🔴 TO DO
-* Click on the **CELLPOSE\_CELLS** tab to check the individual cells that have been segmented Click on the “IMAGE” tab and then “Annotate”, you can check the segmentation on the whole image.
+* Click on the **CELLPOSE_CELLS** tab to check the individual cells that have been segmented. Click on the “IMAGE” tab and then “Annotate”, you can check the segmentation on the whole image.
+
+```{figure} ./img/tutorial_images/Figure5.png
+:width: 600
+:align: center
-
+**Figura 5**: Piximi's annotator tool.
+```
* Optionally, here you can manually refine the segmentation using the annotator tools. The Piximi annotator provides several options to **add**, **subtract**, or **intersect** annotations. Additionally, the **selection tool** allows you to **resize** or **delete** specific annotations. To begin editing, select specific or all images by clicking the checkbox at the top.
* Optionally, you can adjust channels: Although there are two channels in this experiment, the nuclei signal is duplicated in both the red and green channels. This design is intended to be **color-blind friendly** and to produce a **magenta color** for nuclei. The **green channel** also includes cytoplasmic signals.
@@ -105,19 +121,24 @@ Reason for doing this: We want to classify the 'CELLPOSE\_CELLS' based on GFP di
🔴 TO DO
-* Go to the **CELLPOSE\_CELLS** tab that displays the segmented objects (arrow 1, figure 6\)
+* Go to the **CELLPOSE_CELLS** tab that displays the segmented objects (arrow 1, figure 6)
* Click on the **Classification** tab on the left panel (arrow 2, figure 6).
-* Create new categories by clicking **“+ Category”**. Adding “Cytoplasmatic\_GFP”, “Nuclear \_GFP”, “No GFP” three categories (Arrow 3, Figure 6).
+* Create new categories by clicking **“+ Category”**. Adding “Cytoplasmatic_GFP”, “Nuclear_GFP”, “No_GFP” three categories (Arrow 3, Figure 6).
* Click on the images that match your criteria. You can select multiple cells by holding **Command (⌘)** on Mac or **Shift** on Linux. Aim to assign **\~20–40 cells per category**. Once selected, click **“Categorize”** to assign the labels to the selected cells.
-
+```{figure} ./img/tutorial_images/Figure6.png
+:width: 600
+:align: center
+
+**Figura 6**: Classifying individual cells based on GFP presence and localization.
+```
6. ##### **Train the Classifier model**
🔴 TO DO
-* Click the ”
- fit model” icon to open the model hyperparameter settings. For today’s exercise, we’ll adjust a few parameters:
-* Click on “Architecture Settings” and set the Model Architecture to SimpleCNN.
+* Click the ”
- Fit Model” icon to open the model hyperparameter settings. For today’s exercise, we’ll adjust a few parameters:
+* Click on “Architecture Settings” and set the Model Architecture to **SimpleCNN**.
* Update the Input Dimensions to:
- Input rows: 48
- Input cols: 48
@@ -125,8 +146,12 @@ Reason for doing this: We want to classify the 'CELLPOSE\_CELLS' based on GFP di
(You can change to other numbers such as 64, 128)
-
+```{figure} ./img/tutorial_images/Figure7.png
+:width: 600
+:align: center
+**Figura 7**: Classifier model setup.
+```
* Click on the “Dataset Setting” tab and set the Training Percentage to 0.75, which reserves 25% of the labeled data for validation.
* When you click **"Fit Classifier"** in Piximi, two training plots will appear “**Accuracy vs Epochs”** and **“Loss vs Epochs”**. Each plot shows curves for both **training** and **validation** data.
@@ -139,12 +164,17 @@ These plots help you understand how the model is learning and whether adjustment
🔴 TO DO
-
+```{figure} ./img/tutorial_images/Figure8.png
+:width: 400
+:align: center
+
+**Figura 8**: Classifier training and validation.
+```
-* Click **“Predict Model” (figure 8, arrow 1\)** to apply the model we just trained. This step will generate predictions on the cells we did not annotate.
-* You can review the predictions in the CELLPOSE\_CELLS tab and delete any wrongly assigned categories.
+* Click **“Predict Model” (figure 8, arrow 1)** to apply the model we just trained. This step will generate predictions on the cells we did not annotate.
+* You can review the predictions in the CELLPOSE_CELLS tab and delete any wrongly assigned categories.
* Optionally, you can continue using the labels to refine the ground truth and improve the classifier. This process is part of the **Human-in-the-loop classification**, where you iteratively correct and train the model based on human input.
-* Click **“Evaluate Model” (figure 8, arrow 2\)** to evaluate the model we just trained. The confusion metrics and evaluation metrics can be compared to the ground truth.
+* Click **“Evaluate Model” (figure 8, arrow 2)** to evaluate the model we just trained. The confusion metrics and evaluation metrics can be compared to the ground truth.
* Click "Accept Prediction (Hold)”, to assign the predicted labels to all the objects.
8. ##### **Measurement**
@@ -154,13 +184,17 @@ Once you are satisfied with the classification, we will proceed to measure the o
🔴 TO DO
* Click “Measurement” in the top right corner.
-* Click Tables (Arrow 1\) and select Image and click “Confirm” (Arrow 2).
+* Click Tables (Arrow 1) and select Image and click “Confirm” (Arrow 2).
* Choose "MEASUREMENT" in the left panel, note the measurement step may take some time to process.
* Click on 'Category' to include all categories in the measurement.
-* "Under 'Total', click on 'Channel 1' (Arrow 3\) to select the measurement for GFP. You will see the measurement in the “DATA GRID” tab. Measurements are presented as either mean or median values, and the full dataset is available upon exporting the .csv file.
+* "Under 'Total', click on 'Channel 1' (Arrow 3) to select the measurement for GFP. You will see the measurement in the “DATA GRID” tab. Measurements are presented as either mean or median values, and the full dataset is available upon exporting the .csv file.
-
+```{figure} ./img/tutorial_images/Figure9.png
+:width: 600
+:align: center
+**Figura 9**: Add measurements.
+```
9. ##### **Visualization**
@@ -168,13 +202,18 @@ After generating the measurements, you can plot the measurements.
🔴 TO DO
-* Click on 'PLOTS' (Arrow 1\) to visualize the measurements.
+* Click on 'PLOTS' (Figure 10, Arrow 1) to visualize the measurements.
* Set the plot type to 'Swarm' and choose a color theme based on your preference.
-* Select 'Y-axis' as 'intensity-total-channel-1' and set 'SwarmGroup' to 'category'; this will generate a curve showing how GFP intensity varies across different categories (Arrow 2).
-* Selecting 'Show Statistics' will display the mean, as well as the upper and lower quality bounds, on the plot.
+* Select 'Y-axis' as 'intensity-total-channel-1' and set 'SwarmGroup' to 'category'; this will generate a curve showing how GFP intensity varies across different categories (Figure 10, Arrow 2).
+* Selecting 'Show Statistics' will display the mean, as well as the upper and lower confidence boundaries, on the plot.
* Optionally, you can experiment with different plot types and axes to see if the data reveals additional insights.
-
+```{figure} ./img/tutorial_images/Figure10.png
+:width: 600
+:align: center
+
+**Figura 10**: Plot results.
+```
10. ##### **Export results and save the project**
diff --git a/piximi-documentation/translocation_tutorial_ES.md b/piximi-documentation/translocation_tutorial_ES.md
new file mode 100644
index 0000000..c07121d
--- /dev/null
+++ b/piximi-documentation/translocation_tutorial_ES.md
@@ -0,0 +1,229 @@
+# Piximi: Segmentación y clasificación sin instalación en el navegador
+
+## Un ejercicio informático utilizando webtool - Piximi
+
+Beth Cimini, Le Liu, Esteban Miglietta, Paula Llanos, Nodar Gogoberidze
+
+Instituto Broad del MIT y Harvard, Cambridge, MA.
+
+### **Información general:**
+
+#### **¿Qué es Piximi?**
+
+Piximi es una herramienta moderna de análisis de imágenes tomando ventaja de varios métodos de *deep learning*, sin requerir conocimientos de programación. Implementado como una aplicación web en [https://piximi.app/](https://piximi.app/), Piximi no requiere instalación y se puede acceder desde cualquier navegador web moderno. Su arquitectura de cliente único preserva la seguridad de los datos del investigador ejecutando todos los cálculos localmente.
+
+Piximi es interoperable con herramientas y flujos de trabajo existentes, ya que admite la importación y exportación formatos de datos y modelos comunes. La interfaz intuitiva y el fácil acceso a Piximi permiten a los biólogos obtener información sobre las imágenes en tan sólo unos minutos. Piximi tiene como objetivo llevar el análisis de imágenes basado en *deep learning* a una comunidad más amplia mediante la eliminación de las barreras de entrada.
+
+\* excepto las segmentaciones mediante Cellpose, que se envían a un servidor remoto (con el permiso del usuario).
+
+Funciones básicas: **Anotador, Segmentador, Clasificador, Mediciones.**
+
+#### **Objetivo del ejercicio**
+
+En este ejercicio, se familiarizará con las principales funcionalidades de Piximi de anotación, segmentación, clasificación, medición y visualización y lo utilizará para analizar un conjunto de imágenes de muestra de un experimento de translocación. El objetivo de este experimento es determinar la **dosis efectiva más baja** de Wortmannin requerida para inducir la localización nuclear de FOXO1A etiquetada con GFP (Figura 1). Segmentará las imágenes utilizando uno de los modelos de *deep learning* disponibles en Piximi. Comprobará y curará la segmentación y luego entrenará un clasificador de imágenes para clasificar las células individuales como teniendo «GFP nuclear», «GFP citoplasmática» o «sin GFP». Por último, realizará mediciones y las representará gráficamente para responder a la pregunta biológica.
+
+#### **Contexto del experimento de muestra**
+
+En este experimento, los investigadores tomaron imágenes de células U2OS de osteosarcoma (cáncer de hueso) fijadas que expresaban una proteína de fusión FOXO1A-GFP y tiñeron con DAPI para marcar los núcleos. FOXO1 es un factor de transcripción que desempeña un papel clave en la regulación de la gluconeogénesis y la glicogenólisis a través de la señalización de insulina. FOXO1A se desplaza dinámicamente entre el citoplasma y el núcleo en respuesta a diversos estímulos. Wortmannin, un inhibidor de PI3K, puede bloquear la exportación nuclear, lo que resulta en la acumulación de FOXO1A en el núcleo.
+
+```{figure} ./img/tutorial_images/Figure1.png
+:width: 300
+:align: center
+
+**Figura 1**: Representación esquemática del mecanismo de acción de FOXO1A.
+```
+
+
+#### **Materiales necesarios para este ejercicio**
+
+Los materiales necesarios para este ejercicio pueden descargarse de: [PiximiTutorial](./downloads/Piximi_Translocation_Tutorial_RGB.zip). El archivo «Piximi Translocation Tutorial RGB.zip» contiene un proyecto de Piximi que incluye todas las imágenes, ya etiquetadas con el tratamiento correspondiente (concentración de Wortmannin o Control). ¡Descargue este archivo pero **NO lo descomprima**!
+
+#### **Instrucciones para el ejercicio**
+
+Lea los pasos que se indican a continuación y siga las instrucciones donde se indican. Los pasos en los que debe averiguar una solución están marcados con 🔴 PARA HACER.
+
+1. ##### **Cargar el proyecto Piximi**
+
+🔴 PARA HACER
+
+* Inicia Piximi en:[https://piximi.app/](https://piximi.app/)
+
+* Cargar el proyecto de ejemplo: Haga clic en «Abrir» \- “Proyecto” \- «Proyecto desde Zip», como se muestra en la figura 2 para cargar un archivo de proyecto para este tutorial desde Zip, y opcionalmente puede cambiar el nombre del proyecto en el panel superior izquierdo, como «Ejercicio Piximi». A medida que se carga, se puede ver la progresión en la esquina superior izquierda logotipo
.
+
+
+```{figure} ./img/tutorial_images/Figure2.png
+:width: 600
+:align: center
+
+**Figura 2**: Cargando el archivo de proyecto.
+```
+
+2. ##### **Compruebe las imágenes cargadas y explore la interfaz Piximi**
+
+Estas 17 imágenes representan tratamientos con Wortmannin a ocho concentraciones diferentes (expresadas en nM), así como tratamientos con sólo vehículo (0nM). Observe que el canal DAPI (Núcleos) se muestra en magenta y que el canal GFP (FOXOA1) se muestra en verde.
+
+Al pasar el cursor por encima de la imagen, aparecen etiquetas de color en la esquina izquierda de las imágenes. Estas anotaciones proceden de los metadatos del archivo comprimido que acabamos de cargar. En este tutorial, las etiquetas de diferentes colores indican la concentración de Wortmannin, mientras que los números representan el número de imágenes en cada categoría.
+
+Opcionalmente, puede anotar las imágenes manualmente haciendo clic en «+ Category», introduciendo su etiqueta, y luego seleccionando la imagen haciendo clic en las imágenes anotando las imágenes seleccionadas haciendo clic en **«Categorize»**. En este tutorial, nos saltaremos este paso ya que las etiquetas ya estaban cargadas al principio.
+
+```{figure} ./img/tutorial_images/Figure3.png
+:width: 600
+:align: center
+
+**Figura 3**: Explorando las imágenes y etiquetas.
+```
+
+3. ##### **Segmentar Células - diferenciar las células del *background***.
+
+ 🔴 PARA HACER
+
+* Para iniciar la predicción en todas las imágenes, haga clic en «Seleccionar todas las imágenes» en el panel superior, como se muestra en la Figura 3.
+* Cambie la Tarea de Aprendizaje a «SEGMENTATION» (Figura 4, Flecha 1).
+
+* Haga clic en «+ LOAD MODEL» (Flecha 2) y aparecerá una ventana que le permitirá elegir un modelo pre-entrenado (Flecha 3). Para el ejercicio de hoy, seleccione «Cellpose» (Flecha 4). Puede encontrar más información sobre el modelo admitido [aquí](https://documentation.piximi.app/segmentation.html).
+* Haga clic en «Open Segmentation Modeln» (Flecha 5\) para cargar su modelo y seleccionarlo. Por último, haga clic en «Predict model» (Flecha 5). Verá el progreso de la predicción en la esquina superior izquierda debajo del logo de Piximi
.
+* Tardará unos minutos en finalizar la segmentación.
+
+
+```{figure} ./img/tutorial_images/Figure4.png
+:width: 600
+:align: center
+
+**Figura 4**: Cargando un modelo de segmentación.
+```
+
+Tenga en cuenta que los pasos anteriores se realizaron en su computadora local, lo que significa que sus imágenes se almacenan localmente. Sin embargo, la inferencia de Cellpose se ejecuta en la nube, lo que significa que sus imágenes se cargarán para su procesamiento. Si sus imágenes son altamente sensibles, por favor tenga cuidado cuando utilice servicios basados en la nube.
+
+4. ##### **Visualice el resultado de la segmentación y corrija los errores de segmentación**
+
+ 🔴 PARA HACER
+
+* Haga clic en la pestaña **CELLPOSE_CELLS** para comprobar las células individuales que se han segmentado. Haga clic en la pestaña «IMAGE» y luego en «Annotate», puede comprobar la segmentación en toda la imagen.
+
+```{figure} ./img/tutorial_images/Figure5.png
+:width: 600
+:align: center
+
+**Figura 5**: La herramienta de anotación de Piximi.
+```
+
+* Opcionalmente, aquí puede refinar manualmente la segmentación utilizando las herramientas del anotador. El anotador de Piximi ofrece varias opciones para **añadir**, **restar** o **interseccionar** anotaciones. Además, la **herramienta de selección** le permite **redimensionar** o **eliminar** anotaciones específicas. Para empezar a editar, seleccione imágenes específicas, o todas las imágenes, haciendo clic en la casilla de verificación de la parte superior.
+* Opcionalmente, puede ajustar los canales: Aunque hay dos canales en este experimento, la señal de los núcleos se duplicó en los canales rojo y verde. Este diseño está pensado para ser **color-blind friendly** y para producir un **color magenta** para los núcleos. El **canal verde** también incluye señales citoplasmáticas.
+
+Otra razón para duplicar los canales es que algunos modelos (como **Cellpose** que usamos hoy) requieren que las imágenes de entrada tengan **tres canales**.
+
+* Puede optar por segmentar manualmente las células para generar máscaras para los datos de 'verdad de referencia' (*ground truth*).
+
+5. ##### **Clasificar células**
+
+Razón para hacer esto: Queremos clasificar las “CELLPOSE_CELLS” basándonos en la distribución de la GFP (en Núcleos, citoplasma, o sin GFP) sin etiquetarlas todas y cada una manualmente. Para ello, podemos utilizar la función de clasificación en Piximi, que nos permite entrenar un clasificador utilizando un pequeño subconjunto de datos etiquetados y luego clasificar automáticamente las células restantes.
+
+ 🔴 PARA HACER
+
+* Ir a la pestaña **CELLPOSE_CELLS** que muestra los objetos segmentados (flecha 1, figura 6).
+* Hacer clic en la pestaña **Clasificación** del panel izquierdo (flecha 2, figura 6).
+* Cree nuevas categorías haciendo clic en «+ Category». Añadir «Cytoplasmatic_GFP», «Nuclear_GFP», «No_GFP» tres categorías (flecha 3, figura 6).
+* Haga clic en las imágenes que coincidan con sus criterios. Puede seleccionar varias células manteniendo pulsado al tecla **Comamnd (⌘)** en Mac o **Shift** en Linux. Intenta asignar **~20-40 células por categoría**. Una vez seleccionadas, haz clic en **«Categorize»** para asignar las etiquetas a las células seleccionadas.
+
+```{figure} ./img/tutorial_images/Figure6.png
+:width: 600
+:align: center
+
+**Figura 6**: Clasificando células individuales en base a la presencia de GFP y su localización.
+```
+
+6. ##### **Entrenar el modelo clasificador**
+
+ 🔴 PARA HACER
+
+* Haz clic en el icono «
- Fit Model» para abrir la configuración de los hiperparámetros del modelo. Para el ejercicio de hoy, ajustaremos algunos parámetros:
+* Haga clic en «Architecture Settings» y ajuste la *Model Architecture* a **SimpleCNN**.
+* Actualice las dimensiones de entrada a
+ - Filas de entrada: 48
+ - Columnas de entrada: 48
+ - Canales: 3 (ya que nuestras imágenes están en formato RGB)
+
+ (Puede cambiar a otros números como 64, 128)
+
+```{figure} ./img/tutorial_images/Figure7.png
+:width: 600
+:align: center
+
+**Figura 7**: Configuración del modelo clasificador.
+```
+
+* Haga clic en la pestaña «Dataset Setting» y establezca el porcentaje de entrenamiento (*training percentage*) en 0,75, que reserva el 25% de los datos etiquetados para la validación.
+* Cuando haga clic en **"Fit Classifier"** en Piximi, aparecerán dos gráficos de entrenamiento "**Precisión vs Épocas "** y **"Pérdida vs Épocas "**. Cada gráfico muestra curvas para datos de **entrenamiento** y **validación**.
+* En el gráfico de **precisión**, verás lo bien que está aprendiendo el modelo. Lo ideal es que tanto la precisión de entrenamiento como la de validación aumenten y se mantengan cercanas.
+* En el gráfico de pérdidas, los valores más bajos significan un mejor rendimiento. Si la pérdida de validación empieza a aumentar mientras la pérdida de entrenamiento sigue cayendo, el modelo podría estar sobreajustándose.
+
+Estos gráficos le ayudan a comprender cómo está aprendiendo el modelo y si es necesario realizar ajustes.
+
+7. ##### **Evaluar el modelo:**
+
+🔴 PARA HACER
+
+```{figure} ./img/tutorial_images/Figure8.png
+:width: 400
+:align: center
+
+**Figura 8**: Entrenamiento y validación del clasificador.
+```
+
+* Haga clic en **«Predict model» (figura 8, flecha 1)** para aplicar el modelo que acabamos de entrenar. Este paso generará predicciones en las células que no hemos anotado.
+* Puede revisar las predicciones en la pestaña CELLPOSE_CELLS y eliminar cualquier categoría mal asignada.
+* Opcionalmente, puede seguir utilizando las etiquetas para refinar la verdad de referencia (*ground truth*) y mejorar el clasificador. Este proceso es parte de la clasificación **Human-in-the-loop**, donde se corrige iterativamente y entrenar el modelo basado en la entrada humana.
+* Haga clic en **«Evaluate model» (figura 8, flecha 2)** para evaluar el modelo que acabamos de entrenar. Las métricas de confusión y de evaluación pueden compararse con la verdad de referencia (*ground truth*).
+* Haga clic en «Accept Prediction (Hold)» (deberás mantener presionado el cursor unos segundos), para asignar las etiquetas predichas a todos los objetos.
+
+8. ##### **Medición**
+
+Una vez que esté satisfecho con la clasificación, procederemos a medir los objetos. El objetivo del ejercicio de hoy es determinar la concentración mínima de Wortmannin necesaria para bloquear la exportación de FOXO1A-GFP desde los núcleos. Para ello, podemos medir la intensidad total de GFP a nivel de imagen o a nivel de objeto.
+
+🔴 PARA HACER
+
+* Haga clic en «Measurement» en la esquina superior derecha.
+* Hacer clic en Tablas (Flecha 1) y seleccionar Imagen y hacer clic en «Confirm» (Flecha 2).
+* Elija «MEASUREMENT» en el panel izquierdo, tenga en cuenta que el paso de medición puede tardar algún tiempo en procesarse.
+* Haga clic en «Category» para incluir todas las categorías en la medición.
+* En «Total», haga clic en «Channell 1» (Flecha 3) para seleccionar la medición para GFP. Verá la medición en la pestaña «DATA GRID». Las mediciones se presentan como valores medios o medianos, y el conjunto de datos completo está disponible al exportar el archivo .csv.
+
+```{figure} ./img/tutorial_images/Figure9.png
+:width: 600
+:align: center
+
+**Figura 9**: Agregar mediciones.
+```
+
+9. ##### **Visualización**
+
+Después de generar las mediciones, puede trazar las mediciones.
+
+🔴 PARA HACER
+
+* Haga clic en “PLOTS” (Figura 10, Flecha 1) para visualizar las mediciones.
+* Establezca el tipo de trazado en “Swarm” y elija un tema de color basado en su preferencia.
+* Seleccione “Y-axis” como “intensity-total-channel-1” y establezca “SwarmGroup” como “category”; esto generará una curva mostrando cómo varía la intensidad de GFP a través de diferentes categorías (Figura 10, Flecha 2).
+* Seleccionando “Show Statistics” se mostrará la media, así como los límites de confianza superior e inferior, en el gráfico.
+* Opcionalmente, puede experimentar con diferentes tipos de gráficos y ejes para ver si los datos revelan información adicional.
+
+```{figure} ./img/tutorial_images/Figure10.png
+:width: 600
+:align: center
+
+**Figura 10**: Graficar los resultados.
+```
+
+10. ##### **Exportar los resultados y guardar el proyecto**
+
+🔴 PARA HACER
+
+* Haz clic en «SAVE» en la esquina superior izquierda para guardar todo el proyecto. Verás la animación del logo de Piximi a medida que avanza el guardado
.
+
+11. ##### **Información adicional**
+
+Consulta el paper de Piximi: [https://www.biorxiv.org/content/10.1101/2024.06.03.597232v2](https://www.biorxiv.org/content/10.1101/2024.06.03.597232v2)
+
+Consulta la documentación de Piximi:[Documentación de Piximi](https://documentation.piximi.app/intro.html):[https://documentation.piximi.app/intro.html](https://documentation.piximi.app/intro.html)
+
+Informar de fallos/errores o solicitar características [https://github.com/piximi/documentation/issues](https://github.com/piximi/documentation/issues)
\ No newline at end of file