Univer­sity’s hi-tech way to help beat can­cer

Glossop Advertiser - - News - Da­mon Wilkin­son

IT’S a space-age tech­nique nor­mally used to map the sur­face of Mars.

But now sci­en­tists at the Univer­sity of Manch­ester are putting it to another, more down to earth, use – help­ing to fight can­cer.

Spe­cial­ists from the univer­sity and the Manch­ester Can­cer Re­search Cen­tre hope the pre­cise num­ber-crunch­ing method – known as a ma­chine learn­ing ap­proach – will help as­sess the ef­fec­tive­ness of tu­mour treat­ments.

Ini­tial find­ings sug­gest it could be four times as ac­cu­rate as con­ven­tional tech­niques at mea­sur­ing changes in tu­mours.

And it could even­tu­ally mean doc­tors are able to pre­scribe can­cer pa­tients with the right treat­ments and drugs ear­lier, with less trial and er­ror.

Dr James O’Con­nor, a Can­cer Re­search UK ad­vanced clin­i­cian sci­en­tist, said: “Ev­ery per­son’s can­cer is unique, which can make treat­ing the dis­ease chal­leng­ing as a drug that works for one pa­tient might not work for some­one else.

“That’s why we’re in­creas­ingly look­ing at find­ing new ways to make treat­ment more per­sonal, and this in­no­va­tive work could be a step to­wards that goal.

“The next step will be fur­ther re­search to find out if that’s the case, and to help un­cover this method’s po­ten­tial.”

The tech­nique was de­vel­oped at Manch­ester to help plan­e­tary sci­en­tists map fea­tures on plan­ets such as Mars.

It was de­signed to bet­ter un­der­stand the er­rors and un­cer­tain­ties of ob­ser­va­tions, en­abling re­searchers to present their find­ings with con­fi­dence.

The Manch­ester team, from the Di­vi­sion of In­for­mat­ics, Imag­ing & Data Sciences, worked with Dr James O’Con­nor, head of imag­ing within the Manch­ester Can­cer Re­search Cen­tre on stud­ies of lab mice.

Dr Neil Thacker, from the Di­vi­sion of In­for­mat­ics, Imag­ing & Data Sciences, said: “The re­sults of this study show that we can present find­ings which re­searchers can be much more cer­tain of.

“This means you can get the same qual­ity of data from one sam­ple in­stead of 16.

“This has im­por­tant im­pli­ca­tions for re­search, mean­ing that in­stead of us­ing 16 mice, in some stud­ies only one is needed.

“This could help re­duce the use of lab mice in med­i­cal re­search.

“It also opens up the po­ten­tial for this tech­nique to be used in pa­tients by quickly and con­fi­dently iden­ti­fy­ing if drugs are hav­ing a spe­cific ef­fect on their tu­mours.”

Dr Paul Tar, who code­vel­oped the method dur­ing his PhD project, added: “This tech­nique is all about mak­ing the most of ‘ small data’, which is com­mon in med­i­cal stud­ies where it is dif­fi­cult to ob­tain large num­bers of sam­ples.

“Re­searchers use char­i­ta­ble or pub­lic money, so it is im­por­tant that they use it in the most ef­fi­cient way pos­si­ble, some­thing which this tech­nique al­lows.”

●● Re­searchers are us­ing a method cre­ated for space ex­plo­ration in their work on beat­ing can­cer

Newspapers in English

Newspapers from UK

© PressReader. All rights reserved.