Active pixel imagers incorporating pixel-level amplifiers based on polycrystalline-silicon thin-film transistors

Area of Science:

• Materials Science
• Electrical Engineering
• Medical Imaging Technology

Background:

• Active matrix, flat-panel imagers (AMFPIs) using amorphous silicon (a-Si) TFTs are standard in x-ray imaging.
• Their signal-to-noise ratio (SNR) is limited by modest system gain and electronic noise, especially at low exposures or high resolutions.
• Existing designs face constraints in achieving higher performance metrics crucial for advanced diagnostic imaging.

Purpose of the Study:

• To investigate the efficacy of incorporating in-pixel amplification circuits, termed active pixel (AP) architectures, using polycrystalline-silicon (poly-Si) TFTs.
• To overcome the signal-to-noise limitations of conventional a-Si based AMFPIs.
• To evaluate the performance of prototype arrays with varying levels of in-pixel amplification.

Main Methods:

• Developed three prototype indirect detection arrays using poly-Si TFTs and a continuous a-Si photodiode structure.
• Prototypes included a reference array (PSI-1) with a single TFT and two AP arrays (PSI-2, PSI-3) with three and five TFTs, respectively.
• Characterized x-ray sensitivity, charge trapping, modulation transfer function (MTF), and noise performance.

Main Results:

• Poly-Si AP architectures (PSI-2, PSI-3) demonstrated significant signal gains of approximately 10.7x and 20.9x compared to the reference (PSI-1).
• PSI-3 exhibited substantially lower charge trapping compared to PSI-1 and PSI-2, largely independent of bias voltage.
• Despite high additive noise in early prototypes, PSI-3 achieved a significantly lower noise level of ~560 e (rms) due to faster readout and multi-sampling capabilities.

Conclusions:

• Poly-Si TFTs enable sophisticated AP designs capable of achieving tailored signal gain for improved x-ray imaging.
• The PSI-3 design shows promise for future low-noise, high-performance imaging systems by mitigating charge trapping and leveraging advanced readout techniques.
• Further optimization is needed to fully realize the potential of poly-Si AP architectures for net SNR improvement in practical applications.