An Ultra-High Frame Rate Ion Imaging Platform Using ISFET Arrays With Real-Time Compression
An Ultra-High Frame Rate Ion Imaging Platform Using ISFET Arrays With Real-Time Compression
复制标题
使用具有实时压缩功能的 ISFET 阵列的超高帧率离子成像平台
DOI：
10.1109/tbcas.2021.3105328
复制
发表时间：
2021
影响因子：
5.1
通讯作者：
P. Georgiou
中科院分区：
文献类型：
--
作者：
Junming Zeng;Lei Kuang;Miguel Cacho;P. Georgiou
本文提出了一种用于高速离子成像的具有超高通量和实时图像压缩的芯片实验室平台。传感前端由CMOS ISFET阵列组成，传感器在速度饱和下偏置，用于线性pH到电流转换以及高空间和时间分辨率。设计了一个128 × 128像素的阵列，像素尺寸为13. 5 $\boldsymbol{\mu}$m × 10. 5 $\boldsymbol{\mu}$m。通过将ISFET的浮置栅极异步复位到已知的固定电位，像素内复位开关被应用于偏移补偿。此外，每行像素由具有自动调零功能的电流模式信号管道处理，以消除固定模式噪声，然后由片内1 MS/s 8位行并行单斜率ADC处理。采用标准TSMC 180 nm BCD工艺制造，整个片上系统占用2 mm × 2 mm的硅面积，并实现6100 fps的帧速率（模拟为7800 fps）。提出了一种基于USB3.0接口和标准JPEG的25 ms延迟的高速读出平台，分别用于实时离子成像和图像压缩，同时还设计和验证了优化的JPEG算法，以在不牺牲图像质量的情况下获得更高的压缩比。我们通过以6100 fps的速度感测高速离子扩散来展示实时离子图像可视化，这比当前最先进的速度快两倍多。
In this paper, a Lab-on-Chip platform with ultra-high throughput and real-time image compression for high speed ion imaging is presented. The sensing front-end comprises of a CMOS ISFET array with sensors biased in velocity saturation for a linear pH-to-current conversion and high spatial and temporal resolution. An array of 128 × 128 pixels is designed with a pixel size of 13.5 $\boldsymbol{\mu}$m × 10.5 $\boldsymbol{\mu}$m. In-pixel reset switches are applied for offset compensation, by asynchronously resetting the floating gate of the ISFET to a known fixed potential. Additionally, each row of pixels is processed by a current mode signal pipeline with auto zeroing functionality to remove fixed pattern noise, followed by an on-chip 1 MS/s 8-bit row-parallel single slope ADC. Fabricated in standard TSMC 180 nm BCD process, the entire system-on-chip occupies a silicon area of 2 mm × 2 mm, and achieves a frame rate of 6100 fps (7800 fps from simulation). A high speed 25 ms-latency readout platform based on a USB 3.0 interface and standard JPEG is presented for real-time ion imaging and image compression respectively, while an optimised JPEG algorithm is also designed and verified for a higher compression ratio without sacrificing image quality. We demonstrate real-time ion image visualisation by sensing high speed ion diffusion at 6100 fps, which is more than two times faster than the current state-of-the-art.