dorsal/arxiv
View SchemaThe dual-gain mode: a way to enhance the dynamic range of X-ray detectors
| Authors | Evangelos Matsinos, Wolfgang Kaissl |
|---|---|
| Categories | |
| ArXiv ID | physics/0607021 |
| URL | https://arxiv.org/abs/physics/0607021 |
Abstract
Varian Medical Systems has manufactured and recently put into operation a clinically-applicable solution for image-guided radiation therapy. Cone-beam CT imaging, one of the operation modes of the imaging unit of this device, aims at high-quality volumetric reconstruction. To boost the image quality, the dual-gain mode, a successful means for enhancing the dynamic range of the flat-panel detectors and obtaining better results in the contrast of the reconstructed image, was developed and successfully tested during the last few years. The important steps in the calibration of this mode involve a correction to the pulse widths associated with the X-ray production, the assessment of the detector signal above which nonlinear effects become significant and the determination of some properties of the detector pixels, namely, of dark fields, flatness corrections, etc. Finally, the defect-pixel map is obtained, containing dead and flickering pixels, as well as pixels with properties which are sufficiently `out of range'. An effect observed in the offset correction of raw images seems to originate in the mismatch between the measured and extrapolated values of the dark field; a correction scheme is proposed to take account of this effect. In the data processing, which is achieved on the basis of the files produced in the calibration of the mode, the high-gain signal is used whenever meaningful; otherwise, it is substituted by the low-gain signal, properly scaled. Defect pixels are interpolated from their good neighbours; a procedure achieving a correction in case of small defect-pixel clusters is outlined.
{
"annotation_id": "2a5e0d50-7136-4c5e-b504-53a49f160d93",
"date_created": "2026-03-02T18:01:11.440000Z",
"date_modified": "2026-03-02T18:01:11.440000Z",
"file_hash": "0a548e19c5cfbcdedf3bfb28437f68e685f77fdd5eb5a568b6bb19a9ab4109e9",
"private": false,
"record": {
"abstract": "Varian Medical Systems has manufactured and recently put into operation a\nclinically-applicable solution for image-guided radiation therapy. Cone-beam CT\nimaging, one of the operation modes of the imaging unit of this device, aims at\nhigh-quality volumetric reconstruction. To boost the image quality, the\ndual-gain mode, a successful means for enhancing the dynamic range of the\nflat-panel detectors and obtaining better results in the contrast of the\nreconstructed image, was developed and successfully tested during the last few\nyears. The important steps in the calibration of this mode involve a correction\nto the pulse widths associated with the X-ray production, the assessment of the\ndetector signal above which nonlinear effects become significant and the\ndetermination of some properties of the detector pixels, namely, of dark\nfields, flatness corrections, etc. Finally, the defect-pixel map is obtained,\ncontaining dead and flickering pixels, as well as pixels with properties which\nare sufficiently `out of range\u0027. An effect observed in the offset correction of\nraw images seems to originate in the mismatch between the measured and\nextrapolated values of the dark field; a correction scheme is proposed to take\naccount of this effect. In the data processing, which is achieved on the basis\nof the files produced in the calibration of the mode, the high-gain signal is\nused whenever meaningful; otherwise, it is substituted by the low-gain signal,\nproperly scaled. Defect pixels are interpolated from their good neighbours; a\nprocedure achieving a correction in case of small defect-pixel clusters is\noutlined.",
"arxiv_id": "physics/0607021",
"authors": [
"Evangelos Matsinos",
"Wolfgang Kaissl"
],
"categories": [
"physics.med-ph"
],
"title": "The dual-gain mode: a way to enhance the dynamic range of X-ray detectors",
"url": "https://arxiv.org/abs/physics/0607021"
},
"schema_id": "dorsal/arxiv",
"source": {
"execution_id": "7ce12767-6868-400a-b622-1680588fddff",
"id": "arXiv Dataset IDs",
"type": "Model",
"variant": "snapshot-2026-03-01",
"version": "0.1.0"
},
"user_id": 1000002
}