bart ter haar romeny


Part V: 2005-2017

A visual story.

This period covers:
- Northeastern University, Shenyang, China
- RetinaCheck

China & RetinaCheck

It all started with Philips

Prof. Schuurmans and me discussing
with a new teaching professor

Prof. Martin Schuurmans, who was president of the famous Philips Natlab Laboratories in Eindhoven, later established, as Board Member of Philips, new lines for Philips Medical Systems. One line was a substantial expansion in China. In 2006 he founded a Joint Venture with Neusoft, one of China’s largest Medical Imaging companies in China, and TU/e. Neusoft, with more than 20.000 employees (2015) is located in Shenyang, the capital of Liaoning Province in Northeast China. This JV soon needed more biomedical engineers, and, together with the CEO of Neusoft, prof. Jiren Liu, he established a new Biomedical and Information Engineering School, as part of Northeastern University, a large and good local university with 50.000 students. As the first appointed Chinese Dean of BMIE could not do the job due to a major role in the Olympic Games in Beijing, a new dean was needed, and Martin (just retired from Philips in 2006) and his wife Victoire took the task to get the BMIE School through the first infancy year, with great success (see also his autobiography).

I was asked by the Board of TU/e to assist the growing school, and was appointed at BMIE as vice dean for education. A 20-year contract was signed with TU/e. So from the beginning of the BMIE School I was, together with my colleague Peter Hilbers, the TU/e representative, and gave regular courses on medical image analysis and informatics.
In the beginning, I visited Shenyang 1-2 times a year for teaching and advice. Later, I was appointed as professor at Northeastern University in Shenyang, and became project leader of the RetinaCheck project (see below). I frequently was in Shenyang, and had a private apartment on NEU campus.

I represented TU/e in the Joint Management Team (with BMIE, NEU, Philips, Neusoft and TU/e). Here I discuss the progress of the BMIE School of Northeastern University with prof. Jiren Liu (CEO Neusoft).

Prof. Yan Kang, the next dean of BMIE and heading the medical imaging software research at Neusoft, became a close friend. Yan was instrumental in getting things done in China, by introducing me anywhere and oiling many wheels.

BMIE Joint Management Committee in 2007. From left to right: Zhao Yue (Vice-dean), myself, Martin Schuurmans (Dean), Jicheng He (President Norteastern University), Frans Greidanus (Philips - Head of Research and CTO Philips Asia), Jireng Liu (CEO Neusoft) and 3 legal representatives of NEU.             

In the early days of the new BMIE school, students were dressed in uniform, and I really was the 'tall Dutchman'. 
Here students of a Master Class on 14-04-2007.

Design-centered learning

As design-centered learning (tackling a problem in a group) was a proven and highly successful educational concept in Eindhoven, we also introduced this at the BMIE School. The students really liked it. The groups were getting smaller every year, the problems more difficult, and the students more independent and self-confident.

Lecturer Han van Triest (a former MSc student of me in Eindhoven) was moved to China (due to his Chinese love), spoke Chinese and had a background in design-centered learning. He was instrumental in getting it to work in Shenyang. In China all strategies are dictated from above, and experimenting with this new educational form was a challenge, but, with the support of the Deat, the teachers and the unirsity it got a solid foot on the ground.

In 2016 I published a paper with our experiences with DCL in China.

BMIE Shenyang - Medical Image Analysis class - May 2016


I was project leader of a large Sino-Dutch collaborative project on screening project for diabetes in China: RetinaCheck.

In 2013 prof. Wei He, president of the He Vision group, asked us to assist in a eye-fundus screening program for early detection of damage by diabetic retinopathy. In China 11.6% has diabetes, due to genetic predisposition, and a too fast change in lifestyle. The goal was to finally screen 24 million people, the older half of the province of Liaoning. I started the Sino-Dutch RetinaCheck project.

The project was funded by the Dutch Research Council (NWO), the Dutch company i-Optics, the European Foundation for the Study of Diabetes (EFSD), the Chinese Diabetes Society, the EU Marie Curie project, the Chinese Scholarship Council and family support.

Diabetic retinopathy

With the acquired funding, I could hire a postdoc, 4 PhD students and a range of MSc students, and purchase retinal laser-scanning fundus cameras, compute-servers and a dedicated GPU server with a total of 29440 cores. 

Behdad Dashtbozorg, postdoc

Behdad, with a PhD in retinal Image analysis from prof. Aurélio Campilho of Porto University, lead the group in Eindhoven. He was my solid support when I was in China.

RetinaCheck PhD students

- Erik Bekkers
- Samaneh Abbasi
- Fan Huang
- Jiong Zhang

Show more 


Our goal was to first develop a software suite for fully automated retinal image analysis: 

- vessel segmentation, 
- artery-vein classification, 
- vessel diameters, 
- bifucations and crossings, 
- illumination normalozation,
- crossings-preserving denoising,
- optic nerve head detection, 
- fractal dimensions, 
- micro-aneurysm detection, 
- exudates and drusen detection. 

The multi-orientation analysis, pioneered by Remco Duits, was highly rewarding.

In China we collected fundus images from 50 He Vision optometry centers, the He Eye Hospital and the Shengjing Hospital, both in Shenyang. Further images came from a variety of public databases.

Fundus image acquisition in one of the many He Vision optometry stores in Shenyang, China. Left: Mengment Tong, right: the author.

Fundus image acquisition in He Eye Hospital, Shenyang, China

We wrote 56 scientific publications on the project, 4 PhD theses, 1 patent, and 60 student reports: 
Output of the Retinacheck project
(list of scientific publications & MSc reports)
Output of the Retinacheck project
(PDF's of scientific publications & PhD theses)

The RetinaCheck team in Shenyang.

The RetinaCheck team in Eindhoven.

With the whole He Vision team at He Eye Hospital, Shenyang, China.

Life in China, and the great support I received

I frequently went to Shenyang, to teach and supervise the RetinaCheck project. It is not easy to survive in China when you do'nt speak the language and almost nobody speaks English. I had a Chinese bank account, and could arrange my own train tickets, hotel and theater bookings, pay with WeChat etc. I had my private couzy apartment on the old NEU campus, next to an abundance of small restaurants and Sanhao Street, famous for the dozens of large computer outlet stores. With my electric bike (and the excellent metro and bus system) I could easily travel downtown. Only a few Westerners are working/living in Shenyang. Several studytour visits were paid by our TU/e BME students to Shenyang, and I could always arrange a nice program for them. I also was able to discover a lot of beautiful China by traveling
I was tremendously assisted by many good friends:

- prof. Yan Kang, Dean of the BMIE School in Shenyang, and ​General Manager of Clinical Application Business Unit at Neusoft Medical Systems, was my mentor in everything. He joined me at so many meetings, we held lots of brainstorm sessions, and he arranged for my housing (I had a wonderful apartment on NEU old campus), and my appointment at NEU as professor. Hetty and I went on a fantastic cruise on the Yangtze River with Yan and his wife Meixia.

- Prof. Tianzi Jiang is Director of the Key Laboratory of Brainnetome at the Institute of Automation of the Chinese Academy of Sciences (CAS) in Beijing. He is also professor at Queensland Brain Institute, University of Queensland, Australia. We collaborated on brain atlases, and optimizing deep brain stimulation for Parkinson patients. He appointed my group as Associate Member of LIAMA, and me as Visiting Professor at CAS. 

Read More about the
Chinese Academy of Sciences 


In 2019 the project was concluded (see the RetinaCheck final report). The RetinaCheck project was very successful. We published 55 peer-reviewed scientific papers, 4 PhD theses, a patent on our deep learning technology, and a wealth of retinal image analysis software.
The project was converted into the TU/e startup 'RetinaCheck', which was fully acquired by the He Vision Group in 2020. It is very rewarding that the screening program will now, with our RetinaCheck software, be rolled out at scale in China.
During the whole project we were well supported by the TU/e (Rob Debey, Leon Luwijs, Annelies Kroon-van Kuijk, Bart Nelissen). 

And many more memorable events ...