Michael Berthold | Universität Konstanz (original) (raw)

Uploads

Papers by Michael Berthold

Classifying large datasets without any a-priori informa- tion poses a problem especially in the f... more Classifying large datasets without any a-priori informa- tion poses a problem especially in the field of bioinformat- ics. In this work, we explore the problem of classifying hun- dreds of thousands of cell assay images obtained by a high- throughput screening camera. The goal is to label a few selected examples by hand and to automatically label the rest of

Objective: It has been suggested that laser fluorescence close to the dental pulp shows different... more Objective: It has been suggested that laser fluorescence close to the dental pulp shows different values than more distant measurements. The aim of the present study was to compare fluorescence assessed with a device for fluorescence aided caries excavation (FACE) on the cavity floor of medium and deep depth dentin cavities after conventional caries excavation. Method: 41 carious lesions (22 front teeth, 19 premolars/molars) were excavated with a bur in vivo. The endpoint of caries removal was determined by visual, tactile and auditory means. After excavation, 19 cavities were classified as deep, 22 as medium depth cavities. Fluorescence was induced employing a FACE device with an irradiation wavelength of 405 nm. Cavities were observed through a 515-nm high-pass filter. Result: Comparing conventionally excavated medium and deep depth cavities, a statistically significant difference in fluorescence could be observed (p<0.05). Medium depth cavities showed a predominantly green flu...

Objectives: The aim of the study was to assess the impact of laser induced antimicrobial photodyn... more Objectives: The aim of the study was to assess the impact of laser induced antimicrobial photodynamic therapy (aPDT) on the viability of Streptococcus mutans cells employing an aritificial biofilm model. Methods: Employing sterile chambered coverglasses, a salivary pellicle layer formation was induced in 19 chambers. Streptococcus mutans cells were inoculated in a sterile culture medium. Using a live/dead bacterial viability kit, bacteria with intact cell membranes stain fluorescent green. Test chambers containing each the pellicle layer and 0.5 ml of the bacterial culture were analyzed using a confocal laser scan microscope within a layer of 10 m at intervals of 1 m from the pellicle layer. A photosensitizer was added to the test chambers and irradiated with a diode laser (wavelength: 660 nm, output power: 100 mW, Helbo) for 2 min each. Results: Comparing the baseline fluorescence (median: 13.8 [U], min: 3.7, max: 26.2) with the values after adding the photosensitizer (median: 3.7,...

Objectives: Tooth discoloration is described as possible side effect after antimicrobial photodyn... more Objectives: Tooth discoloration is described as possible side effect after antimicrobial photodynamic therapy (aPDT) for root canal disinfection. The aim of the study was to assess the possibility to prevent the photosensitizer dye from staining dental hard tissues after aPDT using different treatment procedures. Methods: 80 freshly extracted human single rooted teeth were included. After preparing the access cavities, coronal flaring was carried out with Gates Glidden burs. Root canals were cleaned and shaped using machine driven rotary instruments (FlexMaster) with a .04 tapered master apical file of ISO size 35. Then, the root canals were dried and an aPDT procedure (Helbo) was performed according to the manufacturer's instructions, using phenothiazine chloride as photosensitizer. Teeth were randomly assigned to four groups: (I) no rinsing after performing the aPDT procedure, (II) rinsing with sodium chloride for 90 s with a hand syringe, (III) pneumatic rinsing with sodium c...

IEEE International Conference on Data Mining, 2002

Proceedings of the 1st international workshop on open source data mining frequent pattern mining implementations - OSDM '05, 2005

International Symposium on Neural Networks, 1995

An algorithm is presented to train a special kind of a local basis function classifier. The so-ca... more An algorithm is presented to train a special kind of a local basis function classifier. The so-called “rectangular basis function network” (RecBFN) consists of hidden units, each covering a rectangular area in the input space, using a trapezoidal activation function. The underlying training algorithm allows easy and fast construction of these types of networks and no parameters need to be

Neural Information Processing Systems, 1995

Radial Basis Function (RBF) Networks, also known as networksof locally--tuned processing units (s... more Radial Basis Function (RBF) Networks, also known as networksof locally--tuned processing units (see [6]) are well known for theirease of use. Most algorithms used to train these types of networks,however, require a fixed architecture, in which the numberof units in the hidden layer must be determined before trainingstarts. The RCE training algorithm, introduced by Reilly, Cooperand Elbaum (see [8]), and

Neurocomputing, 1998

... ted from pictures or Fourier transformed pieces of sound) and produce class informa-tion ... ... more ... ted from pictures or Fourier transformed pieces of sound) and produce class informa-tion ... This could be the slope and bias of a linear decision boundary, or the ... Therefore, techniques for dynamically constructing flexible classifiers without requir-ing a priori knowledge continue ...

Proceedings of the 16th ACM SIGKDD international conference on Knowledge discovery and data mining - KDD '10, 2010

Lecture Notes in Computer Science, 2006

In the previous chapter we have discussed methods that find patterns of different shapes in data ... more In the previous chapter we have discussed methods that find patterns of different shapes in data sets. All these methods needed measures of similarity in order to group similar objects. In this chapter we will discuss methods that address a very different setup: instead of finding structure in a data set, we are now focusing on methods that find explanations

Lasers in Dentistry XVII, 2011

The aim of the study was to assess the impact of laser irradiation during antimicrobial photodyna... more The aim of the study was to assess the impact of laser irradiation during antimicrobial photodynamic therapy. Test chambers containing each a salivary pellicle layer and a Stretococcus mutans culture were analyzed using confocal laser microscopy after adding a photosensitizer. Half of the chambers were irradiated with a diode laser. Comparing baseline fluorescence with the values after laser irradiation, a

Lecture Notes in Computer Science, 1999

ACM SIGKDD Explorations Newsletter, 2009

Lecture Notes in Computer Science, 2011

Proceeding of the 18th ACM conference on Information and knowledge management - CIKM '09, 2009

Proceedings of International Conference on Neural Networks (ICNN'96), 1996

Many algorithms to train radial basis function (RBF) networks have already been proposed. Most of... more Many algorithms to train radial basis function (RBF) networks have already been proposed. Most of them, however, concentrate on building function approximators and only few specialized algorithms are known that concentrate on RBFs for classification. They are based on heuristics that focus on finding areas where relatively few (or no) conflicts occur, but do not try to approximate the underlying