When forensic investigators collect evidence from a crime scene — say a knife, or a doorknob the suspect probably touched — they generally don’t know how much, if any, of the suspect’s DNA is present until the sample is processed in the lab.
“You have to process the whole sample and it could turn out that you have nothing in there,” said Christopher Ehrhardt, Ph.D., an associate professor in the Department of Forensic Science in the College of Humanities and Sciences. “It’s a big problem, especially with touch samples, which is a huge proportion of evidence collected now. We’ve heard from some law enforcement agencies that anywhere between 30 percent and 70 percent of touch samples [that police] submit produce no DNA profiles.”
Ehrhardt has invented a new way to know whether a forensic sample contains biological evidence before it is processed. His process, which takes just a few minutes and does not damage the sample, involves technology called imaging flow cytometry that provides a picture of each individual cell in a sample across multiple fluorescent channels. Ehrhardt’s method then uses big data algorithms to let the analysts know what evidence is present, such as saliva, skin, vaginal cells or blood.
“Crime labs could do this [process], not consume any of the sample and get the information they need in a matter of minutes,” he said. “You could say, ‘Oh, this sample has saliva cells, this cotton swab doesn’t.’ Instead of processing 20 samples, you now only process three because you now know which samples are going to have the evidence.”