Transforming healthcare with AI

Are you going?

38 spots left


Technoparkstrasse 1

Technoparkstrasse 1 · Zürich

How to find us

Enter Technopark through main entrance, take elevator 2 (to Edison building, in the middle of entrance hall), get off at 5th floor, walk the corridor to the right (10 meter) until you reach SCS.

Location image of event venue


17:30 - Doors Open
18:00 - Welcome address
18:05 - Pascal Kaiser
18:35 - Dr. Matteo Manica
19:05 - Closing address
19:10 - Apéro

Pascal Kaiser, Super Computing Systems, Zürich

Pascal received a M.Sc. in Statistics from ETH Zurich and M.Sc. in Biology from University of Zurich. Since 2016 he is working as a software development engineer at Supercomputing Systems. His work focusses mainly on applied artificial intelligence, machine learning and deep learning in the field of life science. In particular, segmentation and classification of biomedical imagery, analysis of medical patient data and predictive maintenance of in-service equipment.

In this talk we present the development and deployment of a cloud-based, Deep Learning (DL) algorithm for the semantic segmentation of Optical Coherence Tomography (OCT) images. OCT is an imaging technique that yields 3D images of the eye. We present the neural network architecture, the ground truth data and the evaluation against benchmark segmentations. Furthermore, we focus on the architecture of the cloud-based web application that hosts the DL algorithm and we explain the deployment of the DL algorithm into production. The DL algorithm is being used for predicting semantic segmentation maps for OCT images by various doctors and researchers through a web interface. This project has been achieved in collaboration between Supercomputing Systems (SCS) and the IOB Basel.

Dr. Matteo Manica, IBM Research, Zürich

Dr. Manica is a Research Staff Member in Cognitive Health Care and Life Sciences at IBM Research Zürich. He's currently working on the development of multimodal deep learning models for drug discovery using chemical features and omic data. He also researches in multimodal learning techniques for the analysis of pediatric cancers in a H2020 EU project, iPC, with the aim of creating treatment models for patients. He received his degree in Mathematical Engineering from Politecnico di Milano in 2013. After getting his MSc he worked in a startup, Moxoff spa, as a software engineer and analyst for scientific computing. In 2019 he obtained his doctoral degree at the end of a joint PhD program between IBM Research and the Institute of Molecular Systems Biology, ETH Zürich, with a thesis on multimodal learning approaches for precision medicine.

Designing a drug is an extremely complex process involving many interconnected phases and it requires an intensive experimental validation of a candidate compound before finally entering clinical trials. The costs of this experimental phase can be prohibitive and any solution that helps to decrease the number of required experimental assays can provide an incredible competitive advantage and reduce time to market. In this talk, I will describe the efforts of IBM Research’s Computational Systems Biology group in the space of in-silico drug sensitivity screening and de novo drug design. In the first part, I will describe PaccMann, a model for compound screening based on the most recent advances in AI for computational biochemistry. The model developed implements a holistic multimodal approach to drug sensitivity combining three key data modalities: anticancer compound structure in the form of SMILES, molecular profile of cell lines in the form of gene expression data and prior knowledge in the form of biomolecular interactions. PaccMann predicts drug sensitivity (IC50) on drug-cell line pairs while highlighting the most informative genes and compound sub-structures using a contextual attention mechanism. I will present PaccMann-RL, a framework built on top of state-of-the-art models for molecules generation. Using an actor critic reinforcement learning scheme, the model is able to bias compound generation towards the design of novel anticancer drug candidates effective on specific biomolecular profiles.