Through the Spectroscopic Glass

Optical spectroscopy (excluding fluorescence) has not, until recently, been a major player in the clinical field. It generates complex spectra that are hard to interpret, and the instruments are bulky and expensive. However, portable spectrometers and spectroscopic sensors are poised to change that paradigm, by providing rapid answers at the point of need – significantly enhancing efficiency, quality, safety, and reduced cycle time. 

A typical spectrometer operating in the visible spectrum (~400–700 nm) can use a linear array detector with 128–512 elements – providing a high resolution of a few nm. However, in condensed phases, absorption bands in this region may be 50-100 nm wide, so that resolution is overkill. Therefore, a spectroscopic sensor with a much smaller number (for example, 16) of discrete resolution elements may work well. In the near-infrared region, array detectors are much more expensive and the bands are still broad, so using a multispectral sensor is beneficial for reducing size, cost, and complexity. It’s worth noting that multispectral sensors have been effectively used for over 50 years in satellite (for example, Landsat) and aerial surveys, as well as in early near-infrared analyzers for food and agriculture – so this concept is already well-established.

Unfortunately, the spectroscopic market is much smaller than sectors like optical telecommunications, consumer electronics, and automotive, so it doesn’t attract major investment for new clinical components, instead relying on advancements in these other fields. With light detection and ranging (LiDAR), I believe that the potential for high-volume commercial production of small, rugged optical components can significantly reduce costs. For example, VCSELs (solid-state lasers used in smartphone facial recognition) now cost just a few dollars to manufacture. LiDAR technology, such as fast swept-source lasers, might also be used in near-infrared spectroscopy in the future.

It will take time for these technologies to mature and be applied to miniature spectroscopy. The challenge lies in developing clinical applications, gaining approval, and convincing the medical field of their effectiveness in improving patient outcomes.

Looking ahead to newer technologies, such as photonic integrated circuits (PICs) and planar optics, we have the potential to enable much smaller spectrometers, manufactured on a large scale using semiconductor techniques. Of course, reducing size and cost provides opportunities across various applications. In the short term, this includes “wearables” and “ingestibles.”

Miniature and lightweight wearable sensors have the ability to monitor heart rate, blood pressure, and blood oxygen levels – transmitting data to a monitoring station. Ingestibles are able to monitor “from within” – equipped with a camera, a spectroscopic sensor, and an ablation device for treating digestive conditions. Using a small multispectral or hyperspectral camera could also provide support in surgery, determining tumor margins. These are just a few of the possibilities this technology could bring to the clinic.

Health-related smartwatches and rings have already reached popularity, with some professional sports teams taking advantage of this technology. Manufacturers are hoping to add more features, including a non-invasive optical blood glucose monitor. Unfortunately, this application has been incredibly difficult to achieve, despite 30–40 years of research from well-funded groups.

However, development continues across large publicized projects – and secret groups. Notable examples include a smartwatch with quantum cascade lasers and photonic integrated circuits (1) and mid-infrared quantum cascade lasers with photothermal detection (2).

We’ve even seen studies showcasing the capabilities of smartphones as thermometers using microbolometers. Some healthcare systems are already taking advantage of the telehealth applications – allowing for easy photo taking and sharing of skin conditions among clinicians. Microbolometer cameras are available today as additions to smartphones for detecting heat leaks in houses, and though clinical use is still developing, they could eventually be used to monitor wound healing at home via telehealth consultations.

Another concept, which is slightly bizarre, is a “smart toilet” for analyzing stool and urine using imaging and spectroscopy to provide health insights and diet suggestions. Scientific papers continue to delve into this topic and a product has been developed from a start-up – leading to discussions by established plumbing companies at events like the Consumer Electronics Show.

Technologies are advancing rapidly, as described above; however, a large amount of mathematical work (chemometrics) is required to turn the volume of data recorded by a spectroscopic sensor into usable information. This isn’t required for the more specific tests already in use across the clinical community; for example, based on antibodies, which give immediate results. In pathology, visual inspection of stained tissue sections is the established method. But spectroscopic researchers are using infrared and Raman imaging to replicate the results without stains. The challenge is to produce results that clinicians recognize and trust – and then obtain regulatory approval.

Looking towards the future of diagnostics, it’s likely that optical spectroscopic techniques will be available in cost-effective, miniature packages with an emphasis on the consumer rather than the clinical space – such as smart watches and other wearables. Of course, clinical applications are guaranteed to follow, but I believe we’re at least 10 years from this integration. 

Looking even further into the future, I expect spectroscopic sensors will become a key part of the home without us even realizing it – in smoke detectors, vacuum cleaners, refrigerators, and washing machines. I implore anyone interested in keeping up to date with this topic to look at SPIE’s BiOS meeting (3) – a part of Photonics West that focuses on biophotonics, biomedical optics, and imaging.