Skip to main content
Department of Information Technology

Seminar series of the Centre for Image Analysis (CBA)

The seminar series hosted by the Centre for Image Analysis (CBA) is open to everybody interested. This is also the seminar of the Image Analysis Unit at Vi3. Join us in Theatrum Visuale, room 100155, building 10, Ångström Laboratory.

It is also possible to follow the seminar via zoom, https://uu-se.zoom.us/j/68686452556.

Each Monday at 14:15-15:00, seminars on current research in image analysis and related topics are given by internal and external speakers. Occasionally, extra seminars may be scheduled at different times. For details, see schedule below.

Welcome!

Coming Seminars, Spring 2024

Date Time Title Speaker
2024-03-11 No Seminar
SSBA/SSDL Symposium
2024-03-18 No seminar
IT Staff meeting
2024-03-25 14:15-15:00
Theatrum Visuale
Title
Abstract
Sara Hamis
2024-04-01 No Seminar
Easter Monday
2024-04-08 14:15-15:00
Theatrum Visuale
TBA
Half-Time Presentation

Abstract
Axel Andersson
2024-04-15 14:15-15:00
Theatrum Visuale
Title
Abstract
2024-04-22 14:15-16:00
Theatrum Visuale
TBA
Half-Time Presentation
Abstract
Can Deniz Bezek
2024-04-29 14:15-15:00
Theatrum Visuale
Application of Deep learning for Detection of Retinal Abnormalities from fundus images
Deep learning has emerged as a powerful tool in medical image analysis, particularly in the domain of retinal imaging for diagnosing diabetic retinopathy. Diabetic retinopathy is a common complication of diabetes and a leading cause of blindness in the working population worldwide. Early detection and timely intervention are crucial in preventing vision loss among diabetic patients. Here I will present some deep learning-based solutions for detection and staging of diabetic retinopathy. I will also focus on the automated segmentation of diabetic retinopathy associated retinal lesions like microaneurysms, haemorrhages, hard exudates and soft exudates from fundus images. I will conclude my talk by discussing the limitations and challenges that are critical to realize the full potential of deep learning for improving the delivery of eye care services.
Nitigya Sambyal
2024-05-06 14:15-15:00
Theatrum Visuale
Cell Detection by Functional Inverse Diffusion and Nonnegative Group Sparsity
On August 28, 2018, a Stockholm-based biotech company launched a new product: The Mabtech IRIS: a next-generation FluoroSpot and ELISpot reader. The reader is a machine designed to analyze biomedical image-based assays that are commonly used in immunology to study cell responses. A contemporary use case involves the development of vaccines for SARS-CoV-2 or the study of T-cells in the immune system. A core technology of the overall solution is a positivity constrained groups sparsity regularized least squares optimization problem, solved with large-scale convex optimization methods. The presentation will outline the problem of analyzing FluoroSpot assays from a signal processing and optimization perspective and explain the methods we designed to solve it. The problem essentially amounts to counting, localizing, and quantifying heterogeneous diffuse spots in an image. The solution involves the development of a tractable linear model of the physical properties that govern the reaction-diffusion-adsorption-desorption process in the assay; the formulation of an inverse problem in function spaces and its discretized approximation; the role of group sparsity in finding a plausible solution to an otherwise ill-posed problem; and how to efficiently solve the resulting 40 million variable optimization problem on a GPU.
Joakim Jaldén
2024-05-13 14:15-15:00
Theatrum Visuale
Title
Abstract
2024-05-20 14:15-15:00
Theatrum Visuale
Title
Abstract
2024-05-27 14:15-15:00
Theatrum Visuale
Title
Abstract
2024-06-03 14:15-15:00
Theatrum Visuale
Title
Abstract
2024-06-10 14:15-15:00
Theatrum Visuale
Title
Abstract
2024-06-17 14:15-15:00
Theatrum Visuale
Title
Abstract
2024-06-24 14:15-15:00
Theatrum Visuale
Title
Abstract

Previous CBA Seminars

Date Time Title Speaker
2024-01-15 14:15-15:00
Theatrum Visuale
Zoom: https://uu-se.zoom.us/j/68686452556
Next-generation bioimage file formats in spatial biology
Methods for biological imaging and in situ profiling of RNA and/or protein expression in biological tissue are rapidly evolving, pushing the limits of existing computational analysis infrastructure. To cope with increased computational demands, novel approaches for storing, accessing and sharing bioimaging data are currently emerging. In this practically-minded seminar, I will discuss ongoing community efforts in standardizing the next generation of bioimage file formats, facilitating the integration of bioimaging data with spatially resolved (e.g., single-cell) data.
Jonas Windhager, SciLifeLab BioImage Informatics Facility (BIIF)
2024-01-22 No seminar due to Staff meeting
2024-01-29 14:15-15:00
Theatrum Visuale
Zoom: https://uu-se.zoom.us/j/68686452556
Advancing Molecular Insights: Report on the EMBL-SLL Internship in Rome
This seminar encapsulates an internship experience at the European Molecular Biology Laboratory (EMBL) in Rome, conducted in September 2023. My primary project, supervised by Alvaro Crevenna in collaboration with the Boulard group, aimed to profile the expression patterns of 200 murine olfactory receptor genes within the nasal epithelium of wild-type and mutated mice using a microscopy-based method and image analysis pipeline. The seminar will include the interdisciplinary nature of the research, hands-on laboratory work with cutting-edge in situ sequencing techniques, and the invaluable skills and knowledge acquired during the internship. My journey of learning, growth, and scientific exploration emphasizes the potential for significant contributions to the field of molecular biology.
Andrea Behanova
2024-02-05 14:15-15:00
Theatrum Visuale
Participating at SciFest - why and how
Last fall (2023) I for the first time participated with a booth/activity at SciFest -the science festival in Uppsala open for schools and the general public. In this seminar I will share my experience and view of the whole process of participating from coming up with the idea, registering the activity, creating the activity and running it at SciFest. And perhaps most importantly what did I get out of it, which corners I think can be cut and whether it was worth it.
Ida-Maria Sintorn
2024-02-12 14:15-15:00
Theatrum Visuale
Methods for Atherosclerotic Plaque Phenotyping
Cardiovascular disease (CVD) is mainly related to complications from atherosclerosis and accounts for more than 30% of overall global mortality. Atherosclerosis leads to plaques in the arteries with lipids, inflammation, and fibrosis. The disease can impair blood flow and cause organ ischemia, and when lesions become unstable, plaque rupture or erosions can lead to atherothrombosis, block blood flow and cause myocardial infarction or cerebral embolization and stroke from plaques in the carotid bifurcation. The main goal is to improve treatment of patients with CVD using the Biobank of Karolinska Endarterectomies (BiKE) and associated clinical and molecular analyses of plasma and plaques from patients with carotid disease. One aspect of the work involves the segmentation of relevant morphologies (phenotypes) and the association of Bulk RNA of plaques to the accompanying histological slides. Using self supervised learning, graph clustering and partial labeling for evaluation, the project so far has successfully segmented the slides from the dataset and the current work involves performing Gene Set Enrichment Analysis, to determine whether a defined set of genes shows statistically significant differences in symptomaticity
Leslie Solorzano
2024-02-19 14:15-15:00
Theatrum Visuale
Equivariant Neural Networks for Biomedical Image Analysis
PhD Thesis Defense Rehearsal
While artificial intelligence and deep learning have revolutionized many fields in the last decade, one of the key drivers has been access to data. This is especially true in biomedical image analysis where expert annotated data is hard to come by. The combination of Convolutional Neural Networks (CNNs) with data augmentation has proven successful in increasing the amount of training data at the cost of overfitting. In this thesis, equivariant neural networks have been used to extend the equivariant properties of CNNs to more transformations than translations. The networks have been trained and evaluated on biomedical image datasets, including bright-field microscopy images of cytological samples indicating oral cancer, and transmission electron microscopy images of virus samples. By designing the networks to be equivariant to e.g. rotations, it is shown that the need for data augmentation is reduced, that less overfitting occurs, and that convergence during training is faster. Furthermore, equivariant neural networks are more data efficient than CNNs, as demonstrated by scaling laws. These benefits are not present in all problem settings and which benefits will occur is somewhat unpredictable. We have identified that the results to some extent depend on architectures, hyperparameters and datasets. Further research may broaden the performed studies to explain how the results occur with new theory.
Karl Bengtsson Bernander
2024-02-26 14:15-15:00
Theatrum Visuale
Detecting key instances within sparsely populated positive bags through self-supervised one-class representation learning, applied to the analysis of cytology images towards early detection of cancer
PhD half-time seminar
Classification of whole slide pathological images using slide-level labels is a scenario where deep multiple-instance learning is predominantly applied. When confronted with images containing a very large number of instances, such as in the context of whole slide cytology images where each instance (patch) is containing roughly one single cell, the algorithm faces challenges in pinpointing key instances when they are limited in number. In this seminar, I will present our evaluation of the efficacy of utilizing representations learned exclusively from patches extracted from normal slides for making decisions at the instance level. The goal is to achieve interpretable decisions at the slide level for whole slide cytology images without overlooking any key instances. Specifically, we explore the effectiveness of a self-supervised contrastive learning framework known as SimCLR in a one-class-classifier setup, evaluating its capability for domain generalization from a limited number of normal slides. We evaluate the proposed approach on one publicly available cytology dataset and one oral cancer dataset collected in collaboration with Folktandvården Stockholms län AB. I will also present work on exploring the role of contextual information, towards improving deep CNN based oral cancer screening on whole slide cytology images.
Swarnadip Chatterjee

External reviewer: Rodrigo Moreno
2024-02-29 15:15-16:00
room 104150
Benchmarking and challenges in biomedical image analysis
Extra seminar by opponent at Karl Bengtsson Bernander's PhD defense on March 1
Benchmarking of image analysis methods using publicly available annotated datasets has become the standard way of comparing new methods to the state-of-the-art. New benchmarks are typically introduced and propagated via competitions (so-called challenges) associated to famous conferences in the field. This holds with some delay also for biomedical image analysis in spite of the lack of annotated datasets. The talk will summarize endeavors in this field and show an example of an own challenge for cell segmentation and tracking.
Michal Kozubek, Masaryk University, Brno
2024-03-04 14:15-15:00
Theatrum Visuale
Collecting Social Interaction Datasets: A School Adventure
The project that has taken most of my research time this last year has been a user study in a local primary school, where we collected a dataset of children playing a collaborative game together. In this presentation, I will talk about why we care about this kind of data, what we planned to do, and all the things that went wrong along the way. Strap in!
Marc Fraile
Updated  2024-03-15 23:18:11 by Joakim Lindblad.