Professor Robert Henderson talks us through how he and his team are developing state-of-the-art detectors for the Proteus project.
Robert you’ve been working on CMOS Sensors for many years. How did you get started in this area? And what triggered your interest in biomedical applications?
I worked for a number of years on the first cameras in mobile phones through the University of Edinburgh startup VLSI Vision Ltd. I have always been very motivated by biomedical applications of electronics and my first job after my PhD was on low power hearing aids. I could see exciting opportunities to bring the sophisticated, but very low-cost, light sensing technologies, now deployed in billions of mobile phones to healthcare.
Tell us about SPAD. How is single photon detection possible?
SPAD stands for Single Photon Avalanche Diode, and is a semiconductor device which operates as a kind of Geiger detector for single quanta of light (photons). Single photons striking the surface of the tiny circular device (measuring around a tenth the diameter of a human hair) cause an avalanche of electrons and a large, extremely fast voltage pulse. The SPAD device therefore acts as an ultra-sensitive optical detector, capable of resolving photon arrival times to picosecond timing precision.
In Proteus, you and your team are responsible for designing the light sensors. What can these sensors do and how are they exploited in the project?
Our sensors will pick up the light returning from the tip of an optical fibre, which is inserted into deepest recesses of the human lung (alveolar space). This light, an optical signature related to the interaction of ultra short laser pulses with specially designed dye molecules, gives infomation about the disease state of the lung. The SPADs pick up the extremely weak signals of the few photons which return and allow the precise timing of the photon’s arrival to be interpreted diagnostically.
Your sensors don’t just register photons, do they? You add lots of functionality to them. What are the key building blocks? We hear a lot, for example, about time-to-digital converters – what are they, and how do they help diagnose lung conditions?
Timing is everything, so we need micro-scale stop-watch electronics capable of measuring the time of arrival of millions of single photons and delivering those to a computer which displays and interprets the time-stamp information. These time-to-digital converters measure at picosecond timescales to reveal biochemical information encoded in the nanosecond decay time of fluorescence from specially designed smart molecules, which probe the disease state. Tens of thousands of timing channels operating in parallel are required on a single microchip to give a complete picture of the underlying chemical interactions.
You have a lot of work going on in wide-field sensors. Will that replace confocal imaging at some point?
Confocal imaging requires scanning of a laser spot across a scene to build up an image. To image faster, it is desirable to flash the laser across the scene and gather the returning light in a series of snapshots. Fast imaging is required to avoid blurring of the image due to movements of the patient or the instrument. This poses the greatest challenge for our detectors: to deliver camera array comprising a 2D array of pixels poses considerable difficulties in capturing light efficiently, as well as timing photons. Our present technology uses lines of detectors or small arrays; nevertheless, we believe efficient SPAD cameras are around the corner due to progress in semiconductor manufacturing in recent years.
What are the big breakthroughs you’ve achieved so far in Proteus? Tell us about the different sensors you’ve been working on.
In Proteus, we have designed the first SPAD line sensor capable of delivering time and spectral information at rates sufficient for scanning of video images from the lung. The sensor produces prodigious amounts of information at 5Gb/s, and we expect will be of considerable value as a scientific tool as well as its primary biomedical function. In addition, we have designed the first ultra-miniature SPAD array suited to endoscopic imaging based on a new technology allowing the SPAD array to be stacked above the electronic layer.
What can we expect to see next in terms of technical breakthroughs and imaging innovation? Are there ways to diagnose health conditions without probing inside the body?
SPAD technology is being driven by adoption in high volume consumer applications, such as autofocus in mobile phones and laser ranging. The medical field can benefit greatly from the advanced technology which is being developed. Other medical imaging fields are set to benefit, such as positron emission tomography where SPADs are set to provide digital photon counting, or more portable forms of near infrared tomography.