NHS Parliamentary Award voor AI-onderzoeker Rob van der Geest 

Op 5 juli 2023 vond in London de uitreiking plaats van de NHS Parliamentary Awards. De eerste prijs in de categorie ‘The Future NHS Award’  werd uitgereikt aan een team van onderzoekers van het Sheffield Teaching Hospital en het LUMC.

Het onderzoeksteam kreeg de prijs vanwege de succesvolle ontwikkeling van een AI-algoritme voor de automatische analyse en kwantificatie van cardiale MRI-beelden en de gerealiseerde implementatie van de ontwikkelde techniek in de klinische workflow. Het AI-algoritme werd ontwikkeld door LUMC-onderzoekers van het Laboratorium voor Klinische en Experimentele Beeldverwerking van de afdeling Radiologie, onder leiding van dr. ir. Rob van der Geest.

Het AI-algoritme maakt het mogelijk om volledig automatisch de cardiale structuren te herkennen in de bewegende MRI-beelden, zodat afwijkingen in de dimensies of de bewegingspatronen van het hart kunnen worden gedetecteerd. Het jury rapport vermeldt onder andere het volgende: “The super-fast analysis could be a ‘game changer’ in future heart disease care – speeding up diagnosis, treatment and care.”

In vervolgonderzoek gaan ze in samenwerking met het universiteitsziekenhuis in Sheffield onderzoeken hoe ze de AI-techniek verder kunnen optimaliseren, zodat deze breder inzetbaar wordt.

Olaf Neve receives price for vestibular schwannoma algorithm

Drs. Olaf Neve received the price¹ for his groundbreaking paper on an innovative deep learning method to automatically segment vestibular schwannomas (brughoektumor in Dutch) on MRI.

Vestibular schwannomas are rare intracranial tumors. Some remain stable and do not require intervention, others progress and do require surgery or radiotherapy. The algorithm gives fast and accurate automated measurements and can detect tumor growth, which has important implications for the decision to treat.

The work is a close collaboration between the Departments of Otorhinolaryngology and Radiology of Leiden University Medical Center, with deep learning contributions from Yunjie Chen and dr. Qian Tao (LKEB), and clinical input from dr. Berit Verbist (Radiology) and dr. Erik Hensen (Otorhinolaryngology). In an observer study² the team showed that the deep learning tool was similar to human delineation in 85% – 92% of cases from a very diverse cohort (from multiple hospitals, scanned on various MR machines), indicating that the tool is robust enough for clinical deployment. With help of the LUMC CAIRELab, clinical implementation is indeed currently pursued.

Example deep learning delineations on challenging cases with cystic components.

The award was created by hearing aid company “Beter Horen” for publications by young Otorhinolaryngology physicians, as a recognition for their work.

¹ https://www.beterhoren.nl/wetenschapsprijs

² https://doi.org/10.1148/ryai.210300

PhD student in AI research Radiology receives Best Abstract Award

Yanli Li, PhD student at the division of image processing (LKEB) of the Radiology department, received in Milan during the international rheumatology congress, EULAR, the  Best Abstract Award for Clinical Science. In collaboration with the Rheumatology department, he studied whether his deep learning models would be able to predict the early development of rheumatoid arthritis from MRI scans of the hands, wrists and feet of patients with clinically suspect arthralgia.

https://www.lumc.nl/over-het-lumc/actueel/2023/juni/prijzen-subsidies-en-benoemingen/

Can a computer replace human interpretation?

EULAR – the European Alliance of Associations for Rheumatology – begins its 2023 annual congress on 31st May in Milan, Italy. Among the data being presented, researchers from the Netherlands ask whether artificial intelligence interpretation of MRI images can provide more accurate predictions than visual scoring in early rheumatoid arthritis (RA).

Go to Press Release

Looking for a challenge in medical imaging?

We have two PhD positions available for our new ICAI lab AI4MRI, that aims to accelerate and improve MR imaging.

PhD Candidate in AI-based Reconstruction of Cardiac MRI

PhD Position on the management of artificial intelligence (AI) in healthcare & policy

Artificial intelligence project om MRI-scans te versnellen ontvangt 2 miljoen euro

Het Leids Universitair Medisch Centrum (LUMC), de Universiteit Leiden en Philips ontvangen samen ruim 2 miljoen euro van NWO om een artificial intelligence (AI) laboratorium op te zetten. Het doel van dit lab is om MRI-scans te versnellen en te verbeteren met AI. Fijn voor de patiënt, én het helpt MRI toegankelijker te maken.

Voor meer informatie: LUMC, Universiteit Leiden, NWO

Job Opportunity: PhD Candidate Machine Learning for Brain Tumor Management

Vestibular schwannomas (VS) are rare intracranial tumors that are (typically) benign, but may cause invalidating symptoms such as hearing loss, balance disturbance or even intracranial hypertension and brainstem compression in advanced cases. Typically, patients are monitored closely with periodic MR imaging, and only in case of tumor progression, treatment is opted for. Computational tools based on Machine Learning may allow prediction of tumor progression in an early stage, when less invasive treatment strategies are still an option. To support this clinical decision, in this PhD project you will develop Artificial Intelligence and Machine Learning methods to predict a patient-specific risk of progression, growth rate, and optimal monitoring frequency. These predictions are predominantly based on longitudinal MR imaging, and can be augmented with relevant clinical information, such as audiometry, symptom-based tests, age and sex. You will likely use and develop multi-task learning strategies, convolutional neural networks and transformers, and spatio-temporal image registration methods. Since LUMC is a national referral center for VS, a very large cohort of multi-center patient data is available, which will be extended with data shared by Erasmus MC. We recently developed a volumetric segmentation model for VS, showing excellent performance on a diverse dataset, which can be used during the project. The methods you develop will be integrated in the clinical workflow by a clinical fellow, to establish their added clinical value, already during the course of your PhD.

For more info see: https://www.lumc.nl/en/over-het-lumc/werken-bij/vacancies/b.23.pv.jh.10-phd-candidate-machine-learning-for-brain-tumor-management/

Grant for development of AI-based tumor growth prediction models

A team of the Departments of Radiology and  Otorhinolaryngology and Head & Neck Surgery received a grant from the Hanarth Fonds to develop and clinically implement machine learning models for growth estimation of vestibular schwannomas.

Vestibular schwannomas (VS) are rare intracranial tumors that are (typically) benign, but may cause invalidating symptoms such as hearing loss, balance disturbance or even intracranial hypertension and brainstem compression in advanced cases. Typically, patients are monitored closely with periodic MR imaging, and only in case of tumor progression, treatment is opted for. Computational tools based on Machine Learning may allow prediction of tumor progression in an early stage, when less invasive treatment strategies are still an option.

The Hanarth Fonds awarded € 400 000 to the MLSCHWAN project, to enable the group of prof. Marius Staring to develop the computational tools. Dr. Erik Hensen, dr. Berit Verbist and dr. Willem Grootjans will then translate these tools to the clinic, and execute retrospective and prospective studies to determine their added value, comparing a workflow with and without AI assistance. In collaboration with Erasmus MC the world’s largest patient cohort is available for training the models, since both centers are national referral centers for VS. The LUMC CAIRELAB supports the translational AI aspects of the project.

The Hanarth Fonds awarded this year in total 13 projects on machine learning for rare tumors, for a total of € 4.4 million: https://www.hanarthfonds.nl/nl/gehonoreerde-aanvragen/2022-call.

LUMC gaat meebouwen aan meest gedetailleerde atlas van menselijk brein tot nu toe

Een toonaangevend team van internationale hersenonderzoekers wil de ongeveer 200 miljard cellen in het menselijk brein in kaart brengen volgens hun type en functie. Een project waar zo’n 110 miljoen dollar mee gemoeid is. Het Leids Universitair Medisch Centrum (LUMC) ontwikkelt technieken voor datavisualisatie om de enorme bak aan complexe data die hieruit voorkomt inzichtelijker te maken. De verwachting is dat deze atlas het onderzoek naar de oorzaken en behandelingen van hersenziektes in een stroomversnelling brengt.

Lees het hele bericht op lumc.nl. 

New ‘Machine Learning’ Professor will focus on Medical Imaging

Marius Staring has been appointed Professor of ‘Machine Learning’ for Medical Imaging as of 1 April, 2022. Staring hopes to contribute to faster and better image processing techniques with relatively new algorithms, such as deep learning [lumc.nl]

SURF Research Support Champions 2022

Congratulations to Michèle Huijberts for being one of the SURF Research Support Champions 2022. “Michèle goes above and beyond to support all the digital needs in our group.” . https://www.surf.nl/en/news/surf-research-support-champions-2022-sarah-coombs-serkan-girgin-michele-huijberts

PhD defence: Mohamed S. Elmahdy

Mohamed S. Elmahdy

Date:             Tuesday March 15, 2022
Time:             15:00
Venue:           online|Academy Building, Rapenburg 73, Leiden
Thesis title:   Deep learning for online adaptive radiotherapy

PhD defence: Antonios Somarakis

Antonios Somarakis

Date:             Thursday 20 january 2022         
Time:             11.15      
Venue:           online|Academy Building, Rapenburg 73, Leiden
Thesis title:   Visual Analytics for Spatially Resolved OMICS Data at Single Cell Resolution: Methods & Application

PhD defence: Hessam Sokooti Oskooyi

Hessam Sokooti Oskooyi

Date:             Thursday 25 November 2021         
Time:            16:15 – 17:00 
Venue:           online|Academy Building, Rapenburg 73, Leiden
Thesis title:  Supervised learning for medical image registration

Zwaartekracht-consortium draagt bij aan het in kaart brengen van de motorische cortex van de hersenen

Twee co-leiders van het Brainscapes-project van NWO Zwaartekracht hebben succesvol bijgedragen aan het in kaart brengen van verschillende celtypen van corticale gebieden van het brein, waaronder de motor cortex. Nature publiceerde een speciale editie over de resultaten van het internationale BRAIN initiative Cell Census Network. De resultaten laten zien hoe genetici, bioinformatici en neurowetenschappers samenwerken om betere behandelingen te ontwikkelen voor hersenziekten.

[nwo][brainscapes]

CZI funds further development of Elastix

The Chan Zuckerberg Initiative (CZI) has funded several open source software projects in the field of biomedicine and Elastix is one of the recipients.

https://www.lumc.nl/over-het-lumc/nieuws/2021/september/een-zomer-vol-prijzen-en-subsidies/

https://github.com/SuperElastix/elastix/

Job Opportunity: PhD Candidate AI-based MR Reconstruction Methods Radiology

The Radiology department, in collaboration with Philips Electronics, is looking for an ambitious PhD candidate with affinity for MRI techniques and artificial intelligence. Is this you? Then don’t hesitate to apply!

PhD-kandidaat-AI-gebaseerde-MR-reconstructie-Radiologie

Successful collaboration between LUMC and LIACS on AI for radiotherapy

Daily-adapted radiotherapy can help to more precisely target radiation dose to tumors compared to the current clinical practice, while avoiding radiosensitive organs-at-risk in the surrounding area. A main obstacle however is that new treatment plans need to be created every day, which is a manual and time-consuming process. A team from LUMC and LIACS recently created AI technology that can do this fully automatically with promising accuracy and in real-time.

https://www.universiteitleiden.nl/en/sails/news/2021/lumc–liacs-collaboration-success

Elastix/Napari plugin

There is now a @napari_imaging plugin for fast image registration based on ITKElastix, the python wrapper for Elastix.

Job opportunity: Postdoc Computational Oncology

We are looking for a postdoc for the division of Image Processing (LKEB), in collaboration with the department of Pathology (Molecular Tumour Genetics, research programs Bone and Soft Tissue Tumours and Immunogenomics). Do you want to contribute to groundbreaking cancer research, and are you experienced in computational approaches? Then you might be the postdoc we are looking for.

PhD defence: Paulien Stegehuis

Paulien Stegehuis

Date:              Tuesday, 20 April 2021        
Time:             11:15 AM     
Venue:           online|Academy Building, Rapenburg 73, Leiden
Thesis title:   The use of optical techniques to guide cancer diagnosis and radical surgical resections

PhD defence: Arlin Keo

Arlin Keo

Date:             Thursday, 3 December 2020
Time:             11:15 AM
Venue:           online (https://www.universiteitleiden.nl/wetenschappers/livestream-promotie)
Thesis title:  Analyzing Spatial Transcriptomics and NeuroImaging data in Neurodegenerative Diseases

PhD defence: Zhiwei Zhai

Date:             Tuesday, 10 March 2020
Time:             11:15
Venue:           Academy Building, Rapenburg 73, Leiden
Thesis title:  Automatic Quantitative Analysis of Pulmonary Vessels in CT: Methods and Applications

PhD defence: Qing Cao

Date:             Tuesday, 21 January 2020
Time:             10:00 AM
Venue:           Academy Building, Rapenburg 73, Leiden
Thesis title:  Automated Analysis Approaches for Coronary CT Angiography

Challenge won

Nature Immunology

PhD defence: Nicola Pezzotti

Date:             Monday, 8th April, 2019
Time:            10:00 h
Venue:          Senaatszaal of the Auditorium, Mekelweg 5, Delft
Thesis title: Dimensionality-Reduction Algorithms for Progressive Visual Analytics

PhD defence: Evgeni Aizenberg

Date:             Tuesday, 14 March 2019
Time:            13:45 h
Venue:          Academy Building, Rapenburg 73, Leiden
Thesis title: Computer-aided techniques for assessment of MRI-detected inflammation for early identification of inflammatory arthritis

Nvidia

PhD defence: Yingguang Li

Date:             Tuesday, 9 October 2018
Time:            11:15h
Venue:          Academy Building, Rapenburg 73, Leiden
Thesis title: Fusion of X-ray angiography and Optical Coherence Tomography for coronary flow simulation

PhD defence: Shengan Liu

Date:             Tuesday 4 September 2018
Time:            16:15h
Venue:          Academy Building, Rapenburg 73, Leiden
Thesis title: Optical coherence Tomography for Coronary Artery Disease: Analysis and Applications

PhD defence: Yuchuan Qiao

Date:             Wednesday, November 1, 2017
Time:            10:00 h
Venue:          Academy Building, Rapenburg 73, Leiden
Thesis title: Fast Optimization Methods for Image Registration in Adaptive Radiation Therapy