The problem is that we still don't understand in much detail the identities of the many circulating molecules that can give early warning of disease, decline and decay, and they are extremely difficult to pin down.

There is one way to amplify molecules to easily detectable amounts. The technique is called PCR, but it is limited to genetic material (DNA or RNA). Now a highly precise way of identifying trace quantities of any kind of molecule has been developed by Professor Chad Mirkin, director of Northwestern's Institute for Nanotechnology in Evanston, Illinois, and colleagues.

They work in ``bio-nanotechnology'', in which they have combined two seemingly incompatible materials - proteins, DNA and other building blocks of life plus synthetic structures made of gold. The test is classified as nanotechnology because it relies on minuscule objects in the order of one to 100 nanometres. A nanometre is one billionth of a metre.

The new method combines nanoscopic particles of gold with DNA labels, called bio-barcodes, to hunt down as few as 10 molecules in 50 millionths of a litre of blood, saliva, cerebral spinal fluid or urine in a matter of minutes. That makes the new test as sensitive as PCR but much more flexible.

Mirkin's team has found a way to label molecules that herald the onset of disease with any DNA sequence he wants. These unique bio-barcodes can then be scanned to identify the associated diseases, ranging from prostate cancer to Alzheimer's. They can also help screen blood for HIV or bovine spongiform encephalopathy (mad cow disease).

Details of ``bio-barcode amplification'' were discussed by Mirkin last month at a meeting of the Royal Society of Chemistry in London. ``This test has the potential to revolutionise medical diagnostics,'' he said. ``What we have is a way to amplify a signal. This system is a million times more sensitive than conventional tools used for screening.''

The test relies on sticking two chemical probes on a target molecule. One probe enables Mirkin to remove the molecule from a sample. The second comes along for the ride to announce that the right molecule has been found. The first probe consists of a magnetic nanoparticle that sticks to a molecule of interest in a sample of blood or urine. Once attached to the `nanomagnet', the molecule can be dragged out of a sample.

The second probe is labelled with hundreds to thousands of identical strands of DNA, which identify the target molecule. The heart of the second probe is a gold nanoparticle. By adding both probes to a sample of blood, scientists can label target molecules with the DNA bio-barcodes and the magnetic particle. Thus the target molecule, along with gold particles bristling with thousands of the barcode molecules, can be separated magnetically, scanned and identified by standard methods used to read the ``letters'' of the DNA in the code.

``This a powerful way of amplifying the signal for a DNA target of interest, such as anthrax,'' said Mirkin, who has been funded by the military, the United States National Institutes of Health and the National Science Foundation. ``There is power in its simplicity.''

The test has proved accurate in detecting ``anthrax lethal factor'' - a marker for anthrax exposure. Overall, it offers 10,000 times more sensitivity and 100,000 times more specificity than current methods used to scan for anthrax DNA.

Mirkin told the London conference that he and his colleague Professor William Klein have also taken the first step towards developing a sensitive test for Alzheimer's. They use the technique to screen cerebrospinal fluid for ADDL proteins that have been linked with the common cause of dementia. No other test used so far has been able to do this.

Klein has shown how ADDLs interfere with one of the nerve cell processes that is essential to memory. This dysfunction occurred well in advance of the cellular degeneration considered by many researchers to be the cause of Alzheimer's, showing it could offer a powerful way to diagnose who will succumb to the disease.

The new method has been used to detect prostate-specific antigen (PSA), a marker for prostate cancer, at low levels. It can also detect the PSA found in low concentrations in breast cancer patients. ``Current tests can't pick it up,'' he said.

The technique is potentially 100 times more sensitive than conventional tests for variant Creutzfeldt-Jakob disease, opening up the possibility of screening blood for the agent or even detecting the disease before symptoms have emerged.

Mirkin is working with Professor Randy Lewis of the University of Wyoming to evaluate its sensitivity. Preliminary data is promising and shows that the assay may be able to detect 10-20 abnormal prion molecules in a cubic centimetre of blood.

A company, Nanosphere, in Northbrook, Illinois, was founded by Mirkin four years ago to develop nanoparticle-based technology for medicine. In a year or two he expects to deliver the first tests he hopes will shift diagnostics out of laboratories and into clinics. ``We are the first to demonstrate technology that can compete with - and in some cases beat - PCR in many categories,'' said Mirkin.

``Nanoscience has made this possible. Our alternative method promises to bring diagnostics to places PCR is unlikely to go - the battlefield, the post office, a Third World village, the hospital and perhaps, ultimately, the home.''