Radiation Imaging Apparatus Noise Removal via FFT Signal Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing radiation imaging apparatuses face erroneous detection of radiation irradiation due to high-frequency noise from static electricity, which is not effectively removed by current noise reduction methods.
Innovation Solution
A radiation imaging apparatus with a pixel array, bias line, drive lines, and processing unit that acquires and processes signal values to identify outliers in bias current, determining radiation irradiation based on non-outlier signal values, thereby reducing erroneous detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If noise reduction methods (sample-and-hold circuit or difference calculation) are used to remove low-frequency and medium-frequency noise, then noise removal effectiveness is improved for those frequency ranges, but high-frequency noise (several MHz to several GHz) caused by static electricity cannot be removed
Solution Approach 1:
The patent applies periodic action by cyclically switching the switch element between conductive and non-conductive states at a specific frequency (e.g., 1 kHz). This periodic switching generates a modulation frequency that is distinct from both low-frequency drift noise and high-frequency static electricity noise. By sampling the bias line current at multiple time points within each cycle and performing frequency-domain analysis (FFT), the system can selectively filter out noise components while preserving the radiation detection signal.
Solution Approach 2:
The patent implements preliminary action by performing multiple signal acquisitions and preprocessing operations before final radiation detection. The system acquires signals at multiple time points within each switching cycle, performs FFT transformation to convert to frequency domain, identifies and removes noise components at specific frequencies, and then performs inverse FFT to reconstruct the cleaned signal. This preliminary noise removal process ensures that high-frequency static electricity noise is eliminated before the final radiation detection decision is made.
2Measurement precision
If the bias line current is monitored continuously to detect radiation irradiation, then detection sensitivity is improved, but high-frequency noise causes erroneous detection
Solution Approach 1:
The patent implements feedback by continuously monitoring the bias line current and comparing it against dynamically updated reference values. The system performs real-time FFT analysis to identify noise components, adjusts the signal processing parameters based on detected noise characteristics, and feeds this information back into the detection algorithm. This feedback mechanism allows the system to maintain high detection sensitivity while dynamically adapting to and compensating for high-frequency noise interference.
Solution Approach 2:
The patent introduces an intermediary processing stage between raw current measurement and final radiation detection decision. The FFT transformation acts as an intermediary that converts the time-domain current signal into the frequency domain, where noise components can be identified and separated from the radiation signal. This intermediary frequency-domain representation allows selective filtering of high-frequency noise while preserving the radiation-induced current changes, thereby eliminating erroneous detection.
3Productivity
If signal processing is performed on all acquired signal values including those with high-frequency noise, then processing speed is maintained, but detection accuracy deteriorates due to erroneous radiation detection
Solution Approach 1:
The patent applies taking out by extracting and removing noise components from the signal processing pipeline. Through FFT transformation, the system identifies high-frequency noise components at specific frequencies (several MHz to several GHz) and extracts them from the total signal. By removing only the noise portions and retaining the radiation-related signal components, the system maintains processing efficiency while significantly improving detection accuracy. This selective extraction approach avoids the need to reprocess entire signal datasets.
Solution Approach 2:
The patent implements parameter changes by transforming the signal from time domain to frequency domain using FFT, which changes the representation parameters of the signal. This parameter transformation allows the system to view noise and signal components in a different dimensional space where they can be more easily separated. By changing the analysis parameter from time-based sampling to frequency-based spectral analysis, the system can efficiently identify and remove high-frequency noise while preserving the radiation detection signal, thereby maintaining processing speed while improving accuracy.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Effectively reduces erroneous detection of radiation irradiation by identifying and excluding high-frequency noise, ensuring accurate detection of radiation presence.
Implementation Method 1
a conversion element for converting radiation into electric charge
Data Source
AI summary
A radiation imaging apparatus includes a pixel array, a bias line, a plurality of drive lines, and a driving unit configured to cyclically supply an ON voltage to the drive lines. The radiation imaging apparatus also includes an acquiring unit configured to acquire a plurality of signal values by acquiring a signal value representing a current flowing through the bias line at each of a plurality of times within a period in which the ON voltage is continuously supplied to at least one of the plurality of drive lines, and a processing unit configured to specify an outlier in the plurality of signal values and determine whether or not there is a radiation irradiation with respect to the pixel array based on a signal value among the plurality of signal values that is not an outlier, and without being based on the outlier.