Current progress in artificial intelligence (AI) paves the way towards intelligent systems that enable automation with unprecedent precision in tasks such as computer vision and action planning. Adapting these models in critical applications both in industry and healthcare is still under debate due to the lack of model transparency and methods to reliably quantify how much we can trust them in unforeseen circumstances. There is also a risk that with increased automation, humans become just observers and they are not actively engaged in the loop. This causes a loss of awareness, a lack of ability to comprehend the complexity of the system and to be empowered to rapidly controlled it. In this talk, Deligianni will highlight current challenges and opportunities in human-machine collaborative frameworks and present paradigms on how AI technology along with wearable/ambient sensing can improve decision making processes.
Dr Fani Deligianni’s holds a PhD in Medical Image Computing (Imperial College London), an MSc in Advanced Computing (Imperial College London), an MSc in Neuroscience (University College London) and a MEng (equivalent) in Electrical and Computer Engineering (Aristotle University, Greece).
Her PhD work was on augmenting 3D reconstructed models of the bronchial tree with 2D video images acquired during bronchoscopy. Bronchial deformation was modelled based on Active Shape Models (ASM) and a predictive tracking algorithm was incorporated to improve tracking of the endoscopic camera.
She was awarded an MRC Special Research Training Fellowship in Biomedical Informatics to explore links between structural connectivity as it is measured with Diffusion Weighted Imaging (DWI) and functional brain connectivity captured with resting-state (rs)-fMRI. She was based at the Biomedical Image Analysis group in Computing Department of Imperial College London. Her research work suggests a prediction framework to study the link between structural brain connectivity and functional brain connectivity.
She developed sophisticate computational approaches in machine learning, statistics and network analysis for the investigation of human brain structure and function. She applied her approach in functional data derived from simultaneous resting-state EEG-fMRI and microstructural indices obtained from neurite orientation dispersion and density imaging of the human brain. In particular, she uses graph theory, machine learning and statistics to describe and characterise complex interconnections between multi-modal brain networks.
Recently, her work has been focused on workload assessment based on neurophysiological signals. She has also done work on human motion analysis with wearable sensors and single rgb(d) camera.