• No results found

DOI: 10.1007/978-3-030-11723-8 11

Abstract: Accurate volumetric change estimation of glioblastoma is very important for post-surgical treat-ment follow-up. In this paper, an interactive segtreat-mentation method was developed and evaluated with the aim to guide volumetric estimation of glioblastoma. U-Net based fully convolutional network is used for ini-tial segmentation of glioblastoma from post contrast MR images. The max flow algorithm is applied on the probability map of U-Net to update the initial segmentation and the result is displayed to the user for interac-tive refinement. Network update is performed based on the corrected contour by considering patient specific learning to deal with large context variations among different images. The proposed method is evaluated on a clinical MR image database of 15 glioblastoma patients with longitudinal scan data. The experimental re-sults depict an improvement of segmentation performance due to patient specific fine-tuning. The proposed method is computationally fast and efficient as compared to state-of-the-art interactive segmentation tools.

This tool could be useful for post-surgical treatment follow-up with minimal user intervention.

14. Training-Free and Segmentation-Free Word Spotting using Feature Matching and Query Expansion Authors:Ekta Vats, Anders Hast, Alicia Forn´es(1)

(1) Universitat Aut`onoma de Barcelona

In Proceedings: 15th International Conference on Document Analysis and Recognition, pp. 1294–1299 DOI: 10.1109/ICDAR.2019.00209

Abstract: Historical handwritten text recognition is an interesting yet challenging problem. In recent times, deep learning based methods have achieved significant performance in handwritten text recognition. How-ever, handwriting recognition using deep learning needs training data, and often, text must be previously segmented into lines (or even words). These limitations constrain the application of HTR techniques in doc-ument collections, because training data or segmented words are not always available. Therefore, this paper proposes a training-free and segmentation-free word spotting approach that can be applied in unconstrained scenarios. The proposed word spotting framework is based on document query word expansion and re-laxed feature matching algorithm, which can easily be parallelised. Since handwritten words posses distinct shape and characteristics, this work uses a combination of different keypoint detectors and Fourier-based descriptors to obtain a sufficient degree of relaxed matching. The effectiveness of the proposed method is empirically evaluated on well-known benchmark datasets using standard evaluation measures. The use of informative features along with query expansion significantly contributed in efficient performance of the proposed method.

5. Texture-based oral cancer detection: A performance analysis of deep learning approaches Authors:Jo Gay, Hugo Harlin, Elisabeth Wetzer, Joakim Lindblad, and Nataˇsa Sladoje In Proceedings: 3rd NEUBIAS Conference, Luxembourg

Comment: Abstract review

6. Visualization of convolutional neural network class activations in automated oral cancer detection for interpretation of malignancy associated changes

Authors:Nadezhda Koriakina, Nataˇsa Sladoje, Ewert Bengtsson, Eva Darai Ramqvist(1), Jan M. Hirsch(2), Christina Runow Stark(3),Joakim Lindblad

(1) Pathology and Cytology, Karolinska Institute, Stockholm, Sweden (2) Oral and Maxillofacial Surgery, UU

(3) Public Dental Service, S¨odersjukhuset, Stockholm, Sweden In Proceedings: 3rd NEUBIAS Conference, Luxembourg Comment: Abstract review

7. CBA Annual Report 2018

Editors:Gunilla Borgefors, Filip Malmberg, Ingela Nystr¨om, Leslie Solorzano, Robin Strand Publisher: Centre for Image Analysis, 107 pages

7 Activities

In this section, we list all the “other” work we do, apart from teaching and research. This is not a small part of our activities, and is an important way of keeping in touch with other scientists and also to showcase what we do to the general public.

Among the various conferences and workshops we organised (Section 7.1), extra noteworthy is the activity at the Technical Museum in Stockholm where 600 14-year olds were introduced to the work of engineers. We held 28 seminars outside CBA (Section 7.2), the closest at the IT department UU, the farthest at University of Victoria in Canada. We are also really proud of our own long-standing well attended seminar series (Section 7.3), with a CBA seminar almost all Monday afternoons. In 2019, we had 33 seminars, of which 16 were held by external scientists from Sweden or abroad. The average number of attendees was 21, ranging from 12 to 30. As usual, we attended many international and national meetings, where we presented our work as invited speakers or giving oral or poster presentations (Section 7.4). We also spoke at very many smaller, non-reviewed meetings (Section 7.5). In total, CBA people presented 50 times at various meetings, or, if seminars are counted, 95 times, equaling almost two per week. We also attended many conferences just to listen and learn (Section 7.6). In fact, almost everybody participated in the 2019 Swedish Symposium on Image Analysis in G¨oteborg. In 2019, we had five international visiting scientists that stayed from a week two seven months (Section 7.7). The longest stay was by Guest Professor Fred Hamprecht from Heidelberg University, Germany. We ourselves visited five other labs, again from a few days to up to three months (Section 7.8).

A rewarding and necessary part of being an international scientist is serving the scientific community by working for professional organisations, being Editors of scientific journals, serv-ing in program committees for international and national conferences, reviewserv-ing for interna-tional journals (which often goes undocumented), being members of dissertation committees, and functioning as evaluators of projects and positions. Many of the CBA seniors have many such engagements, which are listed in Section 7.9.

7.1 Conference organisation

1. Special session on Correlated Multimodal Imaging in Life Sciences (COMULIS) Organisers:Nataˇsa Sladoje

Address: Glasgow, UK Date: 20190319

Comment: EMIM Molecular Imaging meeting; Co-organiser together with the core group of COST CO-MULIS project.

2. 3rd Swedish Symposium on Deep Learning (SSDL 2019) Organisers:Ida-Maria Sintorn, Robin Strand, Joakim Nivre (1) (1) Dept. of Linguistics and Philology, UU

Address: Norrk¨oping Visualization Center, Norrk¨oping Date: 20190610–20190611

3. 21st Scandinavian Conference on Image Analysis (SCIA 2019)

Organisers: Jonas Unger (1),Ida-Maria Sintorn, Michael Felsberg (2), Per-Erik Forss´en (2) (1) Dept. of Science and Technology, Link¨oping University

(2) Dept. of Electrical Engineering, Link¨oping University Address: Norrk¨oping Visualization Center, Norrk¨oping Date: 20190611–20190613

4. Analysis Day in Memory of Mikael Passare

Organisers: Mats Andersson (1),Christer Oscar Kiselman, Pavel Kurasov (2) (main organiser) (1) Dept. of Mathematics at the Chalmers University of Technology and University of G¨oteborg

(2) Institute of Mathematics, Stockholm University Address: Stockholm University

Date: 20190911

Comment: Annual conference in memory of Mikael Passare (1959–2011).

5. IVA 100 years: Experience, Experiment, Examine – as an Engineer (IVA 100 ˚ar: Upplev, uppt¨ack, utforska – som en ingenj¨or)

Organisers:Carolina W¨ahlby, Ewert Bengtsson, and IVA

Address: Tekniska Museet (the Stockholm Museum of Science), Stockholm Date: 20190916–20190917

Comment: This popular science meeting was organised to celebrate the 100th birthday of the Royal Academy of Engineering Sciences, IVA. Six hundred 7th graders were invited to a day of lectures and hands-on exper-iments guided by PhD students and researchers at Tekniska Museet (TM) in Stockholm. Carolina W¨ahlby acted as coordinator of the full event, in collaboration with staff from IVA and TM, and Ginevra Castel-lano organised one of the workshops, where the students could interact with robots. Ewert Bengtsson was responsible for coordinating the event-room.

6. Tutorial at XXIV IberoAmerican Congress on Pattern Recognition (CIARP 2019) Organisers:Filip Malmberg, Fredrik Nysj¨o, Ingela Nystr¨om

Address: Hotel Nacional, Havana, Cuba Date: 20191028

Comment: Arranged a half-day tutorial for the CIARP participants on ”Interactive Segmentation and Visu-alization of Medical Images – from theory to practice”.

7. Swedish Society for Automated Image Analysis workshop

Organisers:Ida-Maria Sintorn, Robin Strand, and rest of SSBA board Address: Storg˚arden, Rimforsa

Date: 20191105–20191106

8. COMULIS conference on Multimodal Imaging in Life Sciences Organisers:Nataˇsa Sladoje

Address: Natural History Museum of Vienna, Austria Date: 20191121–20191122

Comment: Co-organiser together with the core group of COST COMULIS project.

9. Medical Image Analysis and Medical Engineering: How do we provide solutions for tomorrow’s clin-ical needs?

Organisers:Ingela Nystr¨om, Fredrik Nikolajeff (1) (1) Dept. of Materials Science and Engineering, UU Address: Noor Castle, Knivsta

Date: 20191203–20191204

Comment: Medtech Science & Innovation and CBA arranged this workshop to build networks. The 37 participants presented research activities, brainstormed on joint research interests, funding opportunities, etc.