Fast-track diagnosis

3 min read

Diagnosis of cancer could be speeded up with software that uses all a computer's processors to analyse the results of cell testing with flow cytometry.

In these tests cells are extracted from a blood sample, diluted and dyed, then put onto a machine that pushes the cells in single file past a laser beam. A detector finds the type of pulse created by the cell as it goes past the beam and sends the data about the cell, such as its colour intensity, to a computer for processing.

Applied Cytometry, a medical device company based in Dinnington, Sheffield, developed VenturiOne for the increasingly large amounts of data produced by cytometry testing.

'Most computers you buy now have two processors in them —99.99 per cent of all software packages run on one processor because it is very difficult to make a piece of software work on more than one processor,' said Tony Burpee, Applied Cytometry chief executive.

'We have spent the last three and a bit years creating a technology that allows the software to scale evenly across all processors. We have a piece of software called VenturiOne which is based on the core technology, called Venturi. If you have one processor in your box, it uses that one, if you have 16, it uses all 16 and makes it run really, really fast.'

Burpee claims that, apart from Venturi, there is no existing technology that will scale software evenly. The only option would be to employ a group of engineers to rewrite an entire software package so that it will run evenly and efficiently on two or more processors.

The generic nature of Venturi means Applied Cytometry will be able to license the technology to third-party software companies to embed into their products. But it is the company's background in life sciences that has led to the development of the VenturiOne software.

Each time a cell is pushed past the laser beam, it is referred to as an event. The largest data file on which the team has successfully tested VenturiOne measured 780MB, the equivalent of 10 million events. Burpee claimed one advantage of VenturiOne over other software was that it would take only 10 seconds to upload a 780MB file, compared with minutes in existing technology.

Dr Tracey Long, Applied Cytometry's marketing manager, added: 'When you open VenturiOne, you can see all the plots in thumbnail images of them all [the events], whereas in other products you have to ask for each one to be created, and if you want to see them all side by side you have to open another window and cut and paste into it.

'The only reason we can do it is because we have the speed to do it. So when you want to do the gating [filtering], you just say to all your data: "I just want to look at this population of cells" then our software does it very quickly in comparison to other software.

'You just double-click on it and it automatically gates it on all the data, where competitor software will only do it on the ones you have selected and they will take a lot of time.'

Applied Cytometry worked with the University of Pittsburgh Cancer Institute to test the software, and found the cell analysis ran more than 10 times faster than with other software.

To show how much time could be saved, a typical analysis in cancer stem cell investigation of six samples, each containing 11 parameters (characteristics to be detected) with 250,000 to five million events, would usually require 50 hours of analysis time.

By speeding up the analysis, the new software will also be able to carry out faster rare-event analysis (detecting abnormal cells that occur infrequently in a sample).

'Let us say if someone has cancer and you have been treated so that the cancer's gone away. You go back to hospital on a quarterly basis and they would look for cancer cells in your blood. The cancer cells that occur in your blood are pretty infrequent.

'Or if someone is pregnant, approximately one cell in 50,000 in her blood supply will actually be from the baby. With VenturiOne we could detect down to one in 100,000, one in half a million. It means that with this technology you can actually push back the time when you can detect the onset of cancer, so hopefully the therapies will be more effective,' said Burpee.