No abstract

We applied three different model observers (non-prewhitening matched filter with an eye filter, Hotelling and channelized Hotelling) to predict the effect ofJPEG image compression on human visual detection of a simulated lesion (clinically known as thrombus) in single fmme digital x-ray coronary angiograms. Since the model observers' absolute performance is better than human, model performance was degraded to match human performance by injecting internal noise proportional to the external noise. All three modelobservers predicted reasonably well the degradation in human performance as a function of JPEG image compression, although the NPWEW and the channelized Hotelling models (with internal noise proportional to the external noise) were better predictors than the Hotelling model.

No abstract

A method is presented for directly estimating the weights, or "linear template" used by an observer performing a signal-known-exactly detection task in a two-alternative forced-choice (2-AFC) experiment. The approach generalizes prior work by Ahumada, and Beard and Ahumada, to 2-AFC experiments and correlated image noise, and yields an unbiased estimate of the observer template. The estimation procedure is checked against a known linear detection strategy, and human-observer templates estimated from some preliminary psychophysical experiments are shown.

FLASH radiation therapy (FLASH-RT) reference dosimetry to obtain traceability, repeatability and stability of irradiations cannot be performed with conventional dosimetric methods, such as monitor chambers or ionization chambers. Until now, only passive dosimeters have provided the necessary dosimetric data. Alanine dosimetry is accurate; however, to be used for FLASH-RT in biological experiments and for clinical transfer to humans, the reading time needs to be reduced, while preserving a maximum deviation to the reference of ±2%. Optimization of alanine dosimetry was based on the acquisition of electron paramagnetic resonance (EPR) spectra with a Bruker spectrometer. Reading parameters such as the conversion time, the number of scans, the time constant, the microwave power and the modulation amplitude of the magnetic field were optimized as a trade-off between the signal-to-noise ratio (SNR) and the reading time of one measurement using the reference 10.1 Gy alanine pellet. After optimizing the parameters, we compared the doses measured with alanine pellets up to 100 Gy with the reference doses, and then determined the number of measurements necessary to get a difference lower than ±2%. A low-dose alanine pellet of 4.9 Gy was also measured to evaluate the quality of the optimization for doses lower than 10 Gy. The optimization of the Bruker default parameters made it possible to reduce the reading time for one measurement from 5.6 to 2.6 min. That reduction was not at the cost of the SNR because it was kept comparable to the default parameters. Three measurements were enough to obtain a maximum dose deviation to the reference of 1.8% for the range of 10–100 Gy. The total reading time for the three measurements was 7.8 min (3 × 2.6 min). For lower doses such as 4.9 Gy, three measurements led to a deviation greater than 5%. By increasing the number of measurements to five, the average difference to the reference dose was reduced to less than 5% with a total reading time increased to 13.0 min. For doses between 10 Gy and 100 Gy, the optimized acquisition parameters made it possible to keep the average differences between the reference and the measured doses below ±2%, for a reading time of 7.8 min. This enabled an accurate and fast dose determination for biological preparations as part of FLASH-beam irradiations.