Projects in the Laboratory are (or have been) Sponsored by:
1. Decision Support System for Intra Surgical Localization of Subthalamic Nucleus in Parkinson Patients
During deep brain stimulation (DBS) treatment of Parkinson disease, the target of the surgery is the subthalamic nucleus (STN).
This anatomical structure is small (9 × 7 × 4 mm) and poorly visible using Computer Tomography (CT) or Magnetic Resonance Imaging
(MRI) scans. A multi-electrode micro recording system is used intra surgically for better localization of this target nucleus.
My PhD student, Konrad Ciecierski, built decision support system to help neurosurgeons in precise localization of STN.
The work was done in collaboration with Dr. A. Przybyszewski from the
Dept. of Neurology, UMass Medical School and two internationally recognized neurosurgeons from the Medical School in Warsaw, Poland.
The system was successfuly tested in one of the hospitals in Warsaw (Poland) during 30+ Parkinson surgeries.
Our Decision Support System used during Parkinson Surgery
"Foundations of automatic system for intrasurgical localization
of Subthalamic Nucleus in Parkinson patients", K. Ciecierski, Z.W. Ras, A. Przybyszewski, in Web
Intelligence and Agent Systems, International Journal, IOS Press, Vol. 12, No. 1, 2014, 63-82
"Computer Aided Subthalamic Nucleus (STN)
Localization during Deep Brain Stimulation (DBS) Surgery in Parkinson's Patients", Ciecierski, K., Mandat, T., Rola, R., Ras, Z.W., Przybyszewski, A.W.,
Annales Academiae Medicae Silesiensis, Vol. 68, No. 5, 2014, 275-283
"Spike Sorting based upon PCA over DWT Frequency Band Selection", Ciecierski, A., Ras, Z.W., Przybyszewski, A.,
Proceedings of ISMIS 2014 in Roskilde, Denmark, LNAI, Vol. 8502, Springer, 2014, 154-163
"Intraoperative Decision Making with Rough Set Rules for STN DBS in Parkinson Disease", Ciecierski, K., Ras, Z.W.,
Przybyszewski, A.W., Proceedings of the International Conference on Brain Informatics
and Health, in Warsaw, Poland, LNCS, Vol. 8609, Springer, 2014, 323-334
of the micro electrode recordings for STN localization during DBS surgery in
Parkinson's patients", K. Ciecierski, Z.W. Ras, A. Przybyszewski,
Proceedings of FQAS 2013 in Granada, Spain, LNAI, Vol. 8132, Springer, 2013, 328-339
of recommender system for STN localization during DBS surgery in Parkinson's
patients", K. Ciecierski, Z.W. Ras, A. Przybyszewski, in Foundations of
Intelligent Systems, Proceedings of ISMIS 2012 Symposium, LNAI, Vol. 7661,
Springer, 2012, 234-243
of the optimal electrode in Parkinson's disease DBS treatment", K.
Ciecierski, Z.W. Ras, A. Przybyszewski, in Foundations of Intelligent Systems,
Proceedings of ISMIS 2011 Symposium, LNAI, Vol. 6804, Springer, 2011, 554-564
2. MIRAI - System for Automatic Indexing of Polyphonic Music.
Hierachically structured cascade classifiers are used in the system called MIRAI to estimate multiple timbre information from the
polyphonic sound - classification is based on acoustic features and short-term power spectrum matching. MIRAI
system makes a first estimate on the higher level decision attribute which stands for the musical instrument family. Then, the
further estimation is done within that specific family range. Experiments showed better performance of a hierarchical system
than the traditional flat classification method which directly estimates the instrument without higher level of family information
analysis. Traditional hierarchical structures (Hornbostel Sachs, Play Methods) are constructed in human semantics, which
is meaningful from human perspective
but not appropriate for a cascade system. We introduced a new hierarchical instrument schema according to the clustering results
of the acoustic features. This new schema better describes the similarity among different instruments or among different playing
techniques of the same instrument. The classification results show the higher accuracy of cascade system with the new schema
compared to the traditional schemas. Project was sponsored by two NSF Grants: IIS-0414815, IIS-0968647
System MIRAI for Automatic Indexing of Music by Hierarchically Structured Cascade Classifiers
in Music Information Retrieval, Z.W. Ras, A. Wieczorkowska (editors), Studies in Computational Intelligence, Vol. 274, Springer, 2010, 420 pages
"From Personalized to Hierarchically Structured Classifiers for Retrieving Music by Mood", Mostafavi, A., Ras, Z.W., Wieczorkowska, A.,
in "New Frontiers in Mining Complex Patterns",
Post-proceedings of NFMCP 2013, ECML/PKDD Workshop, Prague, Czech Republic, LNAI, Vol. 8399, Springer, 2014, 231-245
Automatic Indexing of Music by Cascade Classifiers", W. Jiang, Z.W. Ras,
Intelligence and Agent Systems, International Journal, IOS Press, Vol. 11,
No. 2, 2013, 149-170
of instrument timbres in real polytimbral audio recordings", E.
Kubera, A. Wieczorkowska, Z.W. Ras, M. Skrzypiec, in Proceedings of 2010
ECML/PKDD Conference, LNCS, Springer, 2010, 97-110
Driven Cascade Classifiers for Multi-Indexing of Polyphonic Music by
Instruments", W. Jiang, Z.W. Ras, A. Wieczorkowska, in Advances in Music
Information Retrieval, Studies in Computational Intelligence, Vol. 274,
Springer, 2010, 19-38
"Music Instrument Estimation in Polyphonic Sound Based on Short-Term Spectrum
Match", W. Jiang, A. Wieczorkowska, Z.W. Ras, in "Foundations of
Computational Intelligence Volume 2", A.-E. Hassanien et al (Eds.), Studies in
Computational Intelligence, Vol. 202, Springer, 2009, 259-273
3. Hierarchically Structured Recommender System for Improving the Efficiency of a Company’s Growth Engine
It is built from knowledge extracted from very large datasets (NPS datasets) containing answers to a set of queries (called questionnaire)
sent to a randomly chosen groups of customers. The purpose of the questionnaire is to check customer satisfaction in using services
of these companies which have repair shops all involved in a similar type of business (in our project - fixing heavy equipment).
These shops are located in 29 states in the US and Canada. Some of the companies have their shops located in more than one state.
They can compete with each other only if they target the same group of customers. The performance of a company is evaluated using
Net Promoter System (NPS). For that purpose, the data from the completed questionnaires are stored in NPS datasets.
We have 34 such datasets, one for each company. Knowledge extracted from them, especially action rules and their triggers,
can be used to build recommender systems giving hints to companies how to improve their NPS ratings. We introduced the concept
of semantic similarity between companies. More semantically similar the companies are, the knowledge extracted from their joined NPS
datasets has higher accuracy and coverage. Our hierarchically structured recommender system is a collection of recommender systems
organized as a tree. Lower the nodes in the tree, more specialized the recommender systems are and the same the classifiers, action rules,
and their triggers used to build their recommendation engines have higher precision and accuracy.
PPT Presentation, POSTER.
(1) One of the system modules generating recommended sets of actions. (2) Kasia Tarnowska and Doug Fowler at the 2016 ROSE-HUB Conference
"In Search for Best Meta-Actions to Boost Businesses Revenue", Kuang, J., Ras, Z.W., in Proceedings of the Conference on Flexible
Query Answering Systems 2015, in Krakow, Poland, Advances in Intelligent Systems and Computing, Vol. 400, Springer, 2015, 431-443
"Personalized Meta-Action Mining for NPS Improvement", Kuang, J., Ras, Z.W., Daniel, A., in Foundations of Intelligent Systems, Proceedings
of ISMIS 2015 in Lyon, France, LNAI, Vol. 9384, Springer, 2015, 73-80
"Hierarchical Agglomerative Method for Improving NPS", Kuang, J., Ras, Z.W., Daniel, A., in
Proceedings of the International Conference on Pattern Recognition and Machine Intelligence, LNCS, Vol. 9124, Springer, 2015, 54-64