SubscribeBiological agents pose one of the greatest risks to personnel at the current time.
Biological agents are extremely lethal to organisms that can be infected by the agent and pose a very high risk due to the very small amount of agent which constitutes a lethal dose. For example, a lethal dose of inhaled anthrax is a very small fraction of a gram.
Such agents are typically many orders of magnitudes more effective on a personnel level than chemical agents. As a consequence of this high degree of lethality, biological agent detectors need to be exceptionally sensitive.
They must be capable of detecting low levels or even single particle agents against a constantly changing background aerosol distribution, which may represent background to agent ratios of hundreds to one or lower. A further problem arises from the fact that biological agents are also intrinsically similar to naturally occurring biological backgrounds, including pollens, spores and benign bacteria.
Given these challenges, biological agent detectors need to be sensitive enough to detect and differentiate on a level where the agent poses a threat to human health and also to be able to warn about an impending attack in sufficient time for appropriate action to be taken.
The field of biological agent detection has received considerable interest recently due to the relative ease with which an agent can be manufactured and disseminated. The threat is high on the list of possible attack scenarios due to the extremely small amounts of agent required to cause widespread disruption and devastation.
While there are considerable problems in attempting a bio-agent attack, such as ensuring that the agent is still viable when human contact occurs and the fact that the production and transportation of agents is rigorously controlled, the high impact of such an assault makes bio-agents a potential method for terrorist activity. An additional problem with bio-agents is the similarity between very lethal pathogens and seemingly benign organisms.
The effect of an agent combining properties of both could result in extreme lethality. For example a highly worrying scenario would be the combination of elements from a lethal agent with a benign but highly infectious organism, such as a flu or common cold virus or similarly infectious bacteria.
In such a situation the effect of the combined bio-agent would result in considerably more damage than the original due to higher infection rates and possibly increased longevity in the environment. The genetic engineering of such a pathogen is not currently possible although new forms of naturally occurring diseases and methods available for the artificial manipulation of bacteria are increasing.
Different detection approaches
The ideal solution for a real-time detector is a biological organism specific response that results in almost instantaneous, specific and repeatable identification. However, there are considerable technological and practical difficulties in the development of sensors that provide a real-time response for all three of these criteria.
As a consequence, the need for real-time information on which to make an informed decision concerning a potential threat results in a relaxation of the need for specific identification to one of a more general nature.
In the most likely scenario where specific identification is not possible, it is likely that very similar organisms would result in the 'detection' of a bio-agent, even though some of these 'agents' do not actually pose a threat.
While such a situation is not ideal, the extreme lethality of bio-agents means that such a semi specific detection would be acceptable as a warning device. In situations where there is a high level of disruption it is likely that a generic detection coupled with a small probability of false alarms is required.
However, in a dangerous environment or if there will be little disruption, a much-reduced specificity would allow for a high degree of protection to personnel who could be potentially exposed.
There are two main detection strategies involved in the protection of personnel - "detect to warn" and "detect to treat". The former is used in situations where there is sufficient time for personnel to be protected or even evacuated, and the latter in situations where exposure has already occurred, or is imminent and treatment is the only remaining cause of action.
There are various strategies currently employed to detect bio-agents - all of which result in a compromise between the specificity, speed and the cost of ownership to the user.
Biotechnology offers the most specific detection approach and the Polymerase Chain Reaction (PCR) technique is capable of the amplification and detection of a DNA sample from a single bio-agent cell within 30 minutes. This technique is too slow to be used as a 'detect to warn' sensor, but could be used as a specific sensor once an agent attack is in progress and used to identify appropriate medical treatment.
Immuno-assay techniques also give a similar specific analysis. However, an additional drawback, on top of the long response time, is the requirement for specialist chemical consumables that add considerably to the logistic burden and running costs that can add hundreds of dollars per hour to the operational cost.
At the other extreme, optical technologies intrinsically result in real-time bio-detection and devices based on these technologies have been available to military and civil defence organisations for a number of years. The response time for these systems arises from the computational processing of the data and as a consequence can be considered as genuinely real-time - Figure 1.
Figure 1. Advantages of different detection strategies.
The figure graphically describes the relative abilities of the optical sensors and the biotechnological approaches of PCR and Immuno-assay. For example the optical technique produces a faster response than either the biotechnologies and is more cost effective but is not as specific.
These devices have little or no consumables and hence a low logistics burden on the deployment, can be run continuously to provide 24 hour detection capability over periods of months or longer and are therefore extremely cost effective.
However the common drawback of this type of sensor is the lack of specificity, with sensors mostly offering a generic detection capability at best, since the optical similarity of the target particles with benign, naturally occurring backgrounds makes them difficult to distinguish. As a consequence of this lack of specificity this type of sensor has been used to monitor changes in the ambient aerosol concentration and distribution and thus provide an indication of the onset of an attack.
One major advantage of these systems is that they can be used as an early warning device and used to trigger more specific detector technologies, such as PCR, thus significantly reducing running costs. Also, networked sensors could be implemented where the response to an 'attack' could be looked at over the network and responses due to localised background changes or localised and benign man-made aerosols ignored.
This type of technology represents the principle of 'detect to warn' and is generally used as a trigger for other technologies or as networked point detection sensors. Common applications include the protection of an area of importance such as a military battlefield, airbase or port, or a civilian area such as a country's major population centre or major airport.
Historical optical detection philosophies
The practical design of optical systems to characterise aerosol distributions uses lasers to determine independent parameters of the particle distribution as a whole. In this manner a map of the background can be constructed using the independent parameters and any changes to this map can then be considered as a change in the aerosol background.
However, the situation is complicated by the fact that the aerosol background is constantly changing and any sampling of the aerosol distribution needs to be sufficiently large so as to reduce statistical variations in the sample with respect to the background. As a consequence of this sampling strategy, sensors that characterise the aerosol in as many different independent parameters as possible have a higher degree of differentiation between the background and potential agents.
There have been two separate historical strategies that have dominated in the US and UK. The US approach has been to analyse particle distributions in terms of the count and size. In contrast the UK's MoD has used a strategy which is based on the philosophy of count, size and shape in the form of the Aerosol Size And Shape (ASAS™) technology manufactured by BIRAL, UK.
More recently the US has adopted the use of fluorescence as well as the size and particle count, with established commercial sensors including the BAWS and BARTS devices from General Dynamics and FLAPS manufactured by TSI.
In the US strategy an ultra-violet (UV) laser is commonly used to illuminate a stream of particles to excite fluorophores contained within bio-agent cells. In this manner only those particles that contain the fluorophores of interest can be differentiated from the background by the fluorescence signature of the particles.
When this information is linked with the size and count differences between the background, the bio-agents can be distinguished and alarms triggered. While this type of sensor has been used in major point detection systems, such as the US military's JBPDS, it has limitations such as susceptibility to false alarms and long term reliability and cost issues.
There have been two historic approaches to the wavelength of laser system used in this form of detector, which are based on common solid state bulk laser systems to produce 266 and 355 nm radiation. These laser wavelengths stimulate different molecules associated with biological cell activity, with the 266 radiation stimulating the amino acid tryptophan, a component of proteins and therefore present in agents of interest (e.g. bacteria, viruses and toxins). A sample fluorescence profile is presented in Figure 2.
Figure 2. Emission spectra for a sample bacteria for different excitation wavelengths.
From the figure it can be seen that the fluorescence response is much stronger for excitation wavelengths around 280 nm. At this wavelength there is a strong response over the 300 - 400+ nm range. For excitation around 350 the fluorescence emission is significantly reduced and is shifted to higher wavelengths.
Furthermore lots of biological and non-biologically active materials fluoresce in the emission band for 350 nm excitation and thus this wavelength produces a large number of false alarms. 280 nm excitation is the preferred source for optical fluorescence detection of biological materials.
(Results kindly supplied by DSTL, UK).
The 355 nm wavelength excites NADH, (Nicotinamide Adenine Dinucleotide), an aerobic respiration-involving molecule and therefore present in biologically active bio-agents, e.g. bacteria. The limitations of the different wavelengths are that the 266 nm source is harder to control and suffers lifetime damage problems, while the 355 nm sources are not as efficient exciters of the bacteria's NADH with respect to 266 and tryptophan. One specific advantage of the 355 nm excitation strategy is the large fluorescence from bacterial growth media -however this can be removed from the bacteria by cleaning.
Furthermore it is questionable whether viruses and toxins (no NADH) can be detected using the 355 method. Additionally, these laser sources are expensive and require large amounts of electrical power due to their inefficient 'wall-plug' operation. Also these systems require cooling and in some cases this is achieved using water and heat exchangers and the systems also require regular maintenance.
Another of the limitations of the fluorescence technique is that commonly occurring aerosol particles also emit similar fluorescence to the bio-agents of interest, and thus these particles are likely to confuse the sensor and result in too many false alarms when a challenging environment is encountered.
Perhaps one of the most worrying of these interferants is fuel oil, which would be prevalent in any situation where such a sensor is likely to be placed, whether military or civilian.
One way to reduce the risk of false alarms from fluorescing interferants is to combine the data with other properties of the particle. Even so false alarms can occur many tens of time a day with this type of sensor.
By contrast the UK has traditionally adopted the use of size, shape and count information for the aerosol distribution to distinguish the particle distribution. This technique uses the elastically scattered light from a stream of particles to calculate the size and shape of the particles and uses the information to produce a "finger print" for the aerosol distribution.
In this way changes to the aerosol profile can be easily detected in the size and shape information and the particle count used to set alarm thresholds when a change in the size-shape distribution has been confirmed.
This device, or variants on it, have been used by the UK Armed Forces for over six years in PBDS (Prototype Biological Detection System) and will be deployed in the upgraded IBDS (Integrated Biological Detection System). It has been also used by companies such as Smiths Detection in bio-detection systems such as NBCerberus. The device itself is not capable of specific identification but is very effective at detecting changes in the particle distribution.
Algorithms and neural networks can then be utilised to pick out tell-tale changes in the distribution indicating the onset of a significant new aerosol source. Fundamentally this technology is extremely sensitive due to the capability to determine minor changes in aerosol backgrounds.
Additionally, a crucial advantage of this sensor is the ability to differentiate between the many sources of bio-agent interference and potential agent releases due to the distinctive size-shape distributions. Common interferants such as fuel oils (diesel, aviation fuel and petrol), smokes, combustion products and natural bio-aerosols such as some pollens can be discriminated and hence this ability greatly reduces the risk of false alarms over situations where this level of discrimination is not available - see Figure 3.
Figure 3. Example of ASAS™ size & shape discrimination.
Use of the size and shape information can be used to determine the background particle profile (Figure 3a). While this profile is constantly changing, it is possible to track the changes and to detect relatively small aerosol events that do not match the background profile.
Once a sufficient number of particles representing a different aerosol profile have been recorded a warning will be produced (Figure 3b). In this figure the change in the lower left-hand side represents a potential threat and the change in profile in the middle right represents an obscurant or other non bio-agent aerosol.
(These images are for illustration purposes only and are not necessarily representative.)
