Collaborating Authors

Could artificial intelligence predict the outcomes of patients with TBI in real time?


In a collaboration project between Helsinki University Hospital (HUS), Kuopio University Hospital and Turku University Hospital (all Finland), a team of researchers have presented the first artificial intelligence (AI) based algorithm that has the potential to assist in treating patients with severe TBI in intensive care units (ICUs). Patients with the most severe cases of TBI are usually treated in ICUs, however, despite the high-quality care, recent observational studies have reported mortality rates of approximately 30%. Patients who suffer from severe TBI are unconscious, therefore, it is a challenge to accurately monitor their condition. In ICUs many tens of variables, such as intercranial pressure and mean arterial pressure, are continuously monitored to assess the patient's condition. One variable alone could yield hundreds of thousands of data points per day, making it impossible for ICU staff to fully analyze.

Artificial intelligence-based algorithm for intensive care of traumatic brain injury


A recent Finnish study published in Scientific Reports presents the first artificial intelligence (AI)-based algorithm designed for use in intensive care units for treating patients with severe traumatic brain injury. The project is a collaborative project between three Finnish university hospitals: Helsinki University Hospital, Kuopio University Hospital and Turku University Hospital. Traumatic brain injury (TBI) is a significant global cause of mortality and morbidity with an increasing incidence, especially in low-and-middle income countries. The most severe TBIs are treated in intensive care units (ICU), but in spite of the proper and high-quality care, about one in three patients dies. Patients that suffer from severe TBI are unconscious, which makes it challenging to accurately monitor the condition of the patient during intensive care.

Latent Gaussian process with composite likelihoods for data-driven disease stratification Machine Learning

Data-driven techniques for identifying disease subtypes using medical records can greatly benefit the management of patients' health and unravel the underpinnings of diseases. Clinical patient records are typically collected from disparate sources and result in high-dimensional data comprising of multiple likelihoods with noisy and missing values. Probabilistic methods capable of analysing large-scale patient records have a central role in biomedical research and are expected to become even more important when data-driven personalised medicine will be established in clinical practise. In this work we propose an unsupervised, generative model that can identify clustering among patients in a latent space while making use of all available data (i.e. in a heterogeneous data setting with noisy and missing values). We make use of the Gaussian process latent variable models (GPLVM) and deep neural networks to create a non-linear dimensionality reduction technique for heterogeneous data. The effectiveness of our model is demonstrated on clinical data of Parkinson's disease patients treated at the HUS Helsinki University Hospital. We demonstrate sub-groups from the heterogeneous patient data, evaluate the robustness of the findings, and interpret cluster characteristics.

Deep learning takes on tumours


As cancer cells spread in a culture dish, Guillaume Jacquemet is watching. The cell movements hold clues to how drugs or gene variants might affect the spread of tumours in the body, and he is tracking the nucleus of each cell in frame after frame of time-lapse microscopy films. But because he has generated about 500 films, each with 120 frames and 200–300 cells per frame, that analysis is challenging to say the least. "If I had to do the tracking manually, it would be impossible," says Jacquemet, a cell biologist at Åbo Akademi University in Turku, Finland. So he has trained a machine to spot the nuclei instead.

AI and Ownership for Educators OEB


Jari Multisilta is the director of Cicero Learning Network in the University of Helsinki, Finland and the professor of multimedia at the Tampere University of Technology, Information Technology at Pori, Finland. He did his doctoral thesis on hypermedia based learning environments for mathematics. Prof. Multisilta has studied learning and modern communication and information technologies and has taken part in several research projects on this area. Currently, his research interests include games for learning, mobile video storytelling, and mobile social video applications. Professor Multisilta has published over 100 international conference papers and journal articles on his research area.