Bjørn-Jostein Singstad

sustech-logo My research is on machine learning for medical applications. I mainly focus on medical application that can be implemented as a tool in the clinic and help doctors and clinicians with decision support in diagnostics and patient triage. An important part of developing systems for medical applications is to validate the it correctly, doing prospective and randomized cotrol studies. Finaly, I also think such systems should be understood by humans, and therfore I also focus on explainable AI (XAI) in my research. This can be helpful both in terms of transparency and for debugging the system, but also this can be used to discover new medical information which can be applied and further studied by clinicians.

I work partly (50%) as a health informatician and advisor at the innovation and research department at Vestfold Hospital Trust in Tønsberg. Here I help clinicians and researchers with their projects, ranging from starting a new health register to data wrangling and statistics. In addition to helping other researchers with their project I also work on my own research projects (VO2max estimation in obese patients), and supervise bachelor and master students.

Akershus Universit Hospital The other 50% of my work time I am doing a PhD, funded by Nasjonalforeningen for Folkehelsen, at the department of e-health and technology at Akershus University Hospital in Lørenskog. The title of my project is: Artificial intelligence-enabled ECG interpretation for improved detection of patients with myocardial infarction. Here we use electrocardiograms (ECG) recorded in a clinical or prehospital setting and other clinical information to train a deep neural network to predict whether a patient has a acute myocardial infarction (total occluded coronary blood vessel) or not.

Selected publications

Estimation of heart rate variability from finger photoplethysmography during rest, mild exercise and mild mental stress

Bjørn-Jostein Singstad, Naomi Azulay, Andreas Bjurstedt, Simen S Bjørndal, Magnus F Drageseth, Peter Engeset, Kari Eriksen, Muluberhan Y Gidey, Espen O Granum, Matias G Greaker, and others

Journal of Electrical Bioimpedance (2021)

Abstract DOI PubMed

Due to the possibilities in miniaturization and wearability, photoplethysmography (PPG) has recently gained a large interest not only for heart rate measurement, but also for estimating heart rate variability, which is derived from ECG by convention. The agreement between PPG and ECG-based HRV has been assessed in several studies, but the feasibility of PPG-based HRV estimation is still largely unknown for many conditions. In this study, we assess the feasibility of HRV estimation based on finger PPG during rest, mild physical exercise and mild mental stress. In addition, we compare different variants of signal processing methods including selection of fiducial point and outlier correction. Based on five minutes synchronous recordings of PPG and ECG from 15 healthy participants during each of these three conditions, the PPG-based HRV estimation was assessed for the SDNN and RMSSD parameters, calculated based on two different fiducial points (foot point and maximum slope), with and without outlier correction. The results show that HRV estimation based on finger PPG is feasible during rest and mild mental stress, but can give large errors during mild physical exercise. A good estimation is very dependent on outlier correction and fiducial point selection, and SDNN seems to be a more robust parameter compared to RMSSD for PPG-based HRV estimation.
Convolutional neural network and rule-based algorithms for classifying 12-lead ecgs

Bjørn-Jostein Singstad, and Christian Tronstad

In 2020 Computing in Cardiology (2020)

Abstract DOI

The objective of this study was to classify 27 cardiac abnormalities based on a data set of 43101 ECG recordings. A hybrid model combining a rule-based algorithm with different deep learning architectures was developed. We compared two different Convolutional Neural Networks, a Fully Convolutional Neural Network and an Encoder Network, a combination of both, and with the addition of another neural network using age and gender as input. Two of these combinations were finally combined with a rule-based model using derived ECG features. The performance of the models was evaluated on validation data during model development using hold-out validation. Finally, the models were deployed to a Docker image, trained on the provided development data, and tested on the Challenge validation set. The model that performed best on the Challenge validation set was then deployed and tested on the full Challenge test set. The performance was evaluated based on a particular Challenge score. Our team, TeamUIO, achieved a Challenge validation score of 0.377, and a full test score of 0.206 for our best model. The score on the full test set placed us at 20th out of 41 teams in the official ranking.