Samulski.nl - Maurice Samulski » Research

Promoted!

Maurice Samulski — Thu, 15 Dec 2011 15:25:24 +0000

On 14 december 2011, I successfully defended my PhD thesis against a committee consisting of Nico Karssemeijer and Peter Lucas (as my promoters), Tom Heskes, Mathias Prokop, Boudewijn Lelieveldt, Roland Holland, Frans Vos, Marina Velikova, and Bram van Ginneken.

Just send me an email if you are interested in my work (msamulski at gmail dot com) and want to request a hardcopy of my thesis (for free). Or get it online here or here

Thanks to everyone that was at my defense, and thanks for all the presents. It was an unforgettable day. Click on more to see the photos my fiancée made.

Thesis defense

Maurice Samulski — Tue, 15 Nov 2011 13:09:00 +0000

Finally! After five years of research and article writing, I am glad to announce that I am defending my Ph.D. thesis “Computer Aided Detection as a Decision Aid in Medical Screening” on wednesday, 14th december 2011 at 10.30 in the aula of the Radboud University Nijmegen. You are all welcome to attend this public defense. Afterwards, a reception will take place in the aula.

The thesis has been printed by Ipskamp Drukkers and is currently being shipped out. If you didn’t receive a copy, please send me an email with your address and I’ll make sure you will get a copy. Meanwhile, my thesis and invitation is also available online:

Download thesis and invitation

Multiview CAD in TMI

Maurice Samulski — Wed, 23 Feb 2011 21:24:45 +0000

My paper entitled “Optimizing case-based detection performance in a multi-view CAD system for mammography” has been accepted for publishing in IEEE Transactions on Medical Imaging and you can get it from here. It is my second publication, and I’m proud that it is going to appear in the highly ranked journal TMI. This article describes our approach to improve the clinically important case-based detection performance detecting masses in mammograms using a multi-view classifier with an alternative learning rule. Read further for the technical abstract.

When reading mammograms, radiologists combine information from multiple views to detect abnormalities. Most computer-aided detection (CAD) systems, however, use primitive methods for inclusion of multi-view context or analyze each view independently. In previous research it was found that in mammography lesion-based detection performance of CAD systems can be improved when correspondences between MLO and CC views are taken into account. However, detection at case level detection did not improve.

In this paper, we propose a new learning method for multi-view CAD systems, which is aimed at optimizing case-based detection performance. The method builds on a single-view lesion detection system and a correspondence classifier. The latter provides class probabilities for the various types of region pairs and correspondence features. The correspondence classifier output is used to bias the selection of training patterns for a multi-view CAD system. In this way training can be forced to focus on optimization of case-based detection performance. The method is applied to the problem of detecting malignant masses and architectural distortions.

Experiments involve 454 mammograms consisting of 4 views with a malignant region visible in at least one of the views. To evaluate performance, 5-fold cross validation and FROC analysis was performed. Bootstrapping was used for statistical analysis.

A significant increase of case-based detection performance was found when the proposed method was used. Mean sensitivity increased by 4.7% in the range of 0.01-0.5 false positives per image.

My first journal article

Maurice Samulski — Mon, 14 Jun 2010 14:19:56 +0000

Yes!! After almost three years of hard work, I have my first journal article accepted for publication by European Radiology. It is published as open-access, so the article is free for everyone to download as a PDF. The article is entitled: “Using computer-aided detection in mammography as a decision support”. I’m very thrilled with my first publication, and I hope that I am on the right track getting my PhD. My second article about a multi-view CAD system is almost ready to be sent to an other journal for consideration. Read further for the abstract.

OBJECTIVE: To evaluate an interactive computer-aided detection (CAD) system for reading mammograms to improve decision making.
METHODS: A dedicated mammographic workstation has been developed in which readers can probe image locations for the presence of CAD information. If present, CAD findings are displayed with the computed malignancy rating. A reader study was conducted in which four screening radiologists and five non-radiologists participated to study the effect of this system on detection performance. The participants read 120 cases of which 40 cases had a malignant mass that was missed at the original screening. The readers read each mammogram both with and without CAD in separate sessions. Each reader reported localized findings and assigned a malignancy score per finding. Mean sensitivity was computed in an interval of false-positive fractions less than 10
RESULTS: Mean sensitivity was 25.1% in the sessions without CAD and 34.8% in the CAD-assisted sessions. The increase in detection performance was significant (p = 0.012). Average reading time was 84.7 +/- 61.5 s/case in the unaided sessions and was not significantly higher when interactive CAD was used (85.9 +/- 57.8 s/case).
CONCLUSION: Interactive use of CAD in mammography may be more effective than traditional CAD for improving mass detection without affecting reading time.

EIBIR Winter School 2010 – Viladrau

Maurice Samulski — Tue, 26 Jan 2010 15:05:47 +0000

I’ve been to the EIBIR Winter School that took place from January 17 – 22, 2010 in Viladrau, Spain to learn a lot about biomedical imaging research focused on the two leading causes of death worldwide: cardiovascular and oncologic diseases. During the five days, there were a lot of intense lectures hosted in the Hostal de la Gloria, Viladrau. To break the week, we also had a very nice excursion: a walk through the forest with a guide that took us to the house of one of the most famous bandits: Serrallonga. After the heavy week of studying, we finished our stay in Viladrau and went up into the Turo de l’Home ( ‘mountain man’ in Castilian), which is the highest mountain in the Montseny massif with a height of 1712 meters.

SPIE 2009

Maurice Samulski — Sun, 15 Feb 2009 12:33:57 +0000

For the third time I went to the SPIE Medical Imaging conference, being held this year in Orlando, Florida. The meeting was from 6-12 February 2009 at the Disney Coronado Springs Resort in Lake Buena Vista. It is one of my favourite conferences, as it has an overall informal character, which enables one to network and share ideas among colleagues in the field. This year I’ve presented my work at a different session than usually, in the Image Perception, Observer Performance and Technology Assessment session. It was a very sunny week, and we had also opportunities to relax and enjoy the nice weather and food.

RSNA 2008

Maurice Samulski — Mon, 15 Dec 2008 10:06:19 +0000

I attended world’s biggest radiology conference in Chicago, Illinois to present our interactive CAD system for detection of masses in mammograms. The technical exhibit was a great success, more than 110 people have done (part of) the observer experiment – and the comments about using CAD in a different way were overall very positive.

Click here to get the exhibit posters, and AuntMinnie’s Insider Exclusive article.

Click here to see some photo’s.

AIA Award

Maurice Samulski — Tue, 28 Nov 2006 09:13:24 +0000

Yesterday, I received an email from Marko van Eekelen, president of the AIA Awards jury that I won the AIA Masters Award of the academic year 05/06 for my Master Thesis! I’m very honored, and this award mades me even more proud of my work. At December 15th there will be an official ceremony where I present a part of my thesis.

Some sources that published the press release:
Radboud University newsitem (dutch)
Press Center (dutch)

Master Thesis

Maurice Samulski — Mon, 19 Jun 2006 12:39:00 +0000

Breast cancer is the most common life-threatening type of cancer affecting women in The Netherlands. The success of the treatment of breast cancer largely depends on the stage of a tumor at the time of detection. Therefore, early detection of breast cancer is essential. Although mammography screening is currently the most effective tool for early detection of breast cancer, up to one-fifth of women with invasive breast cancer have a mammogram that is interpreted as normal, i.e., a false-negative mammogram result. To overcome such limitations, Computer-Aided Diagnosis (CAD) systems for automatic classification of breast lesions as either benign or malignant are being developed. CAD systems help radiologists with the interpretation of lesions, such that they refer less women for further examination when they actually have benign lesions. In this thesis we constructed several types of classifiers, i.e., Bayesian networks and support vector machines, for the task of computer-aided diagnosis of breast lesions.

Classification of Breast Lesions in Digital Mammograms, Master Thesis (1.3 MB PDF)

Image compression: Fourier vs Wavelets

Maurice Samulski — Wed, 19 Apr 2006 14:46:15 +0000

Digital images have become an important source of information in the modern world of communication systems. Thinks about FBI fingerprints, MRI scans, ECG data, digital camera photos, building plans etc. In their raw form, these images require an enormous amount of memory. With the introduction of multimedia computing, the demand for processing, storing and transmitting images has increased exponentially. In the last two decades a considerable amount of research has been devoted to tackle the problem of image compression.

During the Research Lab 4 course, I wrote a paper about the similarities and differences, advantages and disadvantages of the two well known compression techniques: Fourier and Wavelets. JPEG, JPEG2000, DiVX and MPEG, all of them are based on one of these two techniques.

Research Paper (3.8 MB PDF) (5.5 MB PS)

Presentation (1.9MB PPT) (2.2MB PDF) (HTML)