File size: 3,219 Bytes
6640272
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
---
title: CellsiLAB
emoji: 🔬
colorFrom: purple
colorTo: pink
sdk: static
pinned: false
---

# CellsiLAB: Cells and Image Analysis

**CellsiLAB** is the biomedical image analysis area of the **Codalab** research group at the **Universitat Politècnica de Catalunya (UPC)**.

We work on **applied artificial intelligence**, focused mainly (though not exclusively) on **biomedical image analysis**. We develop **deep learning** and **computer vision** methods to support clinical decision-making, with a strong emphasis on **explainable** and **trustworthy AI**: we care not only about *what* a model predicts, but *why*.

## Research lines

- **Explainable AI (xAI)** for biomedical imaging: new methodologies, evaluation strategies and clinical applications.
- **Computer vision for medical imaging**: classification, segmentation and quality assessment.
- **Generative and vision-language models** applied to medical data.
- **Federated learning** for robust multi-center generalization.

## Application domains

- **Hematology**: blood cell classification, myelodysplastic syndromes, leukemia and red cell morphology.
- **Obstetrics**: fetal ultrasound quality assessment (nuchal translucency).
- **Pediatric ophthalmology**: retinopathy of prematurity.

## Research projects

- **Explainable deep learning in medical image analysis**: novel methods, evaluation strategies and clinical applications *(PID2023-146261OB-I00, 2024 to 2027)*.
- **Computational system for the diagnosis of acute leukemia and lymphoma** from peripheral blood images, including a proof of concept and technological valorization *(PDC2022-133514-I00, 2022 to 2024)*.
- **CellsiMaticDeep**, computational hematology: deep learning solutions for the diagnosis of hematological diseases from peripheral blood cell images *(PID2019-104087RB-I00, 2020 to 2023)*.
- **Characterization and automatic classification of leukemic cells** by means of digital image processing and pattern recognition for diagnosis support *(DPI2015-64493-R, 2016 to 2018)*.

## Institutions and collaborations

- **Universitat Politècnica de Catalunya (UPC)**
- **Hospital Clínic de Barcelona**, CORE Laboratory
- **Hospital Sant Joan de Déu**, Barcelona
- **Centro de Diagnóstico Biomédico (CDB)**
- International partners across Europe, Latin America and Asia

## Team

- **Dr. José Rodellar.** Professor (Emeritus), UPC. Signal processing and machine learning for blood cell morphology. [Profile](https://futur.upc.edu/178882)
- **Santiago Alférez.** Assistant Professor, Dept. of Mathematics, UPC. Machine learning, statistics and explainable AI for medical image analysis. [Profile](https://futur.upc.edu/EdwinSantiagoAlferezBaquero)
- **Kevin Barrera.** Assistant Professor, UPC. Deep learning for cell morphology. [Profile](https://futur.upc.edu/KevinIvanBarreraLlanga)
- **Dr. Anna Merino.** Clinical lead, CORE Laboratory, Hospital Clínic de Barcelona. Hematology and cell morphology.

## Datasets and Models

Our public datasets and models will be published here soon. Stay tuned.

## Links

- Website: [cellsilab.com](https://cellsilab.com/)

---

*For collaborations or inquiries, please reach out through our [website](https://cellsilab.com/).*