Research

PhD student at Western Norway University of Applied Sciences, Department of Computer science, Electrical engineering and Mathematical sciences and Mohn Medical Imaging and Visualization Center, Department of Radiology, Haukeland University Hospital . My research actives are mainly related to machine learning and medical image analysis, with a particular focus on design methodologies in deep learning for efficient use of data.

Conference posters

Brain age versus chronological age: A large scale MRI and deep learning investigation
with A. Lundervold and A.S Lundervold. EPOS scientific poster at ECR 2020, online, Jul. 2020

BrainAge: From DICOM to age
with A. Lundervold and A.S Lundervold. Poster at MMIV Conference 2019, Bergen, Norway, Dec. 2019

fastai for 3D MRI deep learning & explainable AI
with A. Lundervold and A.S Lundervold. Poster at MMIV Conference 2019, Bergen, Norway, Dec. 2019

Transfer learning for medical images: a case study
with A.S Lundervold. Poster at GTC Europe 2018, Munich, Germany, Oct. 2018

Preprints, submitted and in preparation

Submitted: 2D and 3D U-Nets for skull stripping in a large and heterogeneous set of head MRI using fastai
S. Kaliyugarasan, M. Kocinski, A. Lundervold, A.S Lundervold

In prep: Brain age versus chronological age: A large scale MRI and deep learning investigation
S. Kaliyugarasan, A. Lundervold, A.S. Lundervold

Talks and travels

Latest and upcoming

• Medisinsk bildebehandling og maskinlæring. Title: Kunstig intelligens ved MMIV, Radiologisk avdeling, Haukeland universitetssykehus., Streaming, October 12, 2020
• Seminar, Bergen gynekologisk kreft - Voss 2020: Title: Artificial intelligence in image diagnostics: design methodologies for efficient use of data and radiologist’s expertise., Scandic Voss, Norway, March 5-6, 2020
• Posters at MMIV 2019: Convergence of medical data science for improved patient care. , Bikuben, Bergen, Norway, December 9-11, 2019
• Bergen Data Science Meetup. Title: Artificial intelligence in image diagnostics: design methodologies for efficient use of data., Bouvet ASA, Bergen, Norway, November 19, 2019
• TekPRAT Førde - Maskinlæring og kunstig intelligens. Title: MMIV@HUS: Kunstig intelligens ved radiologisk avdeling., Førde Sentralsjukehus, Norway, September 23, 2019
• MMIV Seminar September 2019. Title: Artificial intelligence in image diagnostics – transfer learning and active learning for efficient use of data and radiologist’s expertise., Haraldsplass Diakonale Sykehus, Bergen, Norway, September 20, 2019
• NordBioMedNet Summer School 2019 in Computational Biomedicine - Imaging, machine learningn and precision medicine. Title: Deep Learning in medical image analysis., Seili, Finland, August 11-16, 2019
• Exhibition at Christiekonferansen 2019. Title: Mohn Medical Imaging and Visualization Center.,   Universitetsaulaen i Bergen, Norway, April 29, 2019
• Bergen AI & Machine Learning Symposium 2019. Title: Deep transfer learning: a case study., Solstrand Hotel, Bergen, Norway, March 25-26, 2019
• Exhibition at EHiN 2018., Oslo Spektrum, Norway, November 14, 2018
• Machine learning seminar: A machine learning mini-conference. Title: Deep transfer learning: Can a network trained to do a task be reused for other tasks?, Haukeland universitetssjukehus, Bergen, Norway, October 17, 2018
• Poster at GTC Europe 2018. Title: Transfer learning for medical images: a case study., Munich, Germany, October 9-11, 2018

Other research activites

Lumbar spine segmention
Manually segmented 5 T1-weighted scans in collaboration with Ansgar Espeland, and trained a deep neural network to automatically segment new scans with promising results., Haukeland University Hospital, Bergen, Norway, January-August, 2019

Master thesis

Deep transfer learning in medical imaging
S. Kaliyugarasan, University of Bergen and Western Norway University of Applied Sciences, 2019