#computer algorithm

Counting Copies

<

You have almost as many cells in your brain as there are stars in the Milky Way – roughly 200 billion – all of which came from a small pool of fast-dividing stem cells. In order for the brain to develop correctly as a foetus grows in the womb, not only is the number of cells important – they must also be the right types of cells in the right places. To find out more about how this happens, researchers developed a computer algorithm that can identify individual cells in living tissue (each cell is highlighted by a white spot in this image) and track what happens as they divide and specialise. Using fruit fly larvae as a simple model for humans, the researchers were able to continuously monitor developing larval brains for up to 24 hours. The results provide unprecedented insights into how complex organs like brains are built from single cells.

Written by Kat Arney

Image from work by Martin Hailstone and colleagues

Department of Biochemistry, University of Oxford, Oxford

Image originally published with a Creative Commons Attribution 4.0 International (CC BY 4.0)

Published in eLife, May 2020

Pinpointing Breast Cancer

Surgery is the single most effective treatment for breast cancer diagnosed at an early stage, before it has spread through the body. If the tumour is small enough, patients will be offered a ‘lumpectomy’ (breast conserving surgery) rather than removing the whole breast. In order to cut out the tumour and a safe margin of healthy tissue around it, which may be harbouring rogue cancer cells, it’s essential that the surgeon knows exactly where it is within the breast. This is usually done through an MRI scan, with the patient lying face down inside the scanner. However, surgery obviously has to be done with the patient lying face up. Researchers are using computer algorithms to analyse and compare MRI scans taken in the face down (top row) or face up position (bottom row), enabling surgeons to predict the location of a tumour more accurately and ensure they’ve removed it all.

Written by Kat Arney

Image from work by Chuan-Bing Wang, Sangwook Lee, Namkug Kim and BeomSeok Ko, and colleagues

University of Ulsan College of Medicine, Asan Medical Center, Seoul, South Korea

Image originally published with a Creative Commons Attribution 4.0 International (CC BY 4.0)

Published in Scientific Reports, March 2020

Unstill Life

Wanting to know more about the dark world inside our cells, microscopists are usually torn by a choice – do they aim for a detailed 3D picture or capture a quick, but less-detailed video? Super-resolved structured illumination microscopy (SR-SIM), for example, captures high-resolution pictures of the tiniest aspects of a living cell, but these images require processing which usually limits how fast pictures can be snapped. Here though, a new technique uses algorithms to offload the processing work to a separate graphics processing unit – an example of GPU-acceleration captures this image of a living bone cancer cell in a fraction of a second, with its nucleus highlighted in blue, mitochondria (green) and cytoskeleton (pink). Using SR-SIM to take videos has huge potential – from recording the speedy movements of tiny microbes in living cells to allowing researchers to quickly 'screen' samples based on tell-tale signs of health and disease.

Written by John Ankers

Image by W. Hübner, Bielefeld University

Biomolecular Photonics, Faculty of Physics, Bielefeld University, Bielefeld, Germany

Image copyright held by the original authors

Research published in Nature Communications, September 2019

New Resolution

Microscopy is enjoying huge leaps in recent technology and technique, bringing tiny life into sharp focus and allowing its processes to be studied like never before. Yet for the best results, techniques like super-resolution microscopy require optimisation – juggling mechanical settings of lenses and lasers with software for capturing images. Even then, scientists struggle with a question – how do you know if the image you’ve taken is the best it can be? Maybe just one more tweak in the right direction would reveal even more detail. In this artificially-coloured human cell, we see the results of a new computer algorithm. Designed to estimate an image’s resolution, it helps to spot when microscope settings could be improved further (left), guiding scientists towards a clearer, optimised image (right). Released as open-source software, such approaches will help researchers squeeze more information from their experiments, and value from their expensive microscopes.

Written by John Ankers

Image from work by A. Descloux, K. S. Grußmayer and A. Radenovic, ÉPFL

École Polytechnique Fédérale de Lausanne, Laboratory of Nanoscale Biology, Lausanne, Switzerland

Image copyright held by the original authors

Research published in Nature Methods, September 2019

Stitched Up

The brain’s processing power easily surpasses the most advanced supercomputer, yet information technology can still help us to understand our nervous systems. Mapping out connections of this developing fruit fly’s nervous system, the sophisticated computer algorithms of a new software called BigStitcher work alongside two microscopy techniques: expansion microscopy, to swell and inflate the tiny tissues under the lens, and light sheet microscopy to take thousands of virtual 'slices' through the cells. Each coloured area is a huge 3D image (carrying many gigabytes of data) virtually reassembled into a mosaic ‘map’ of the whole nervous system. The next challenge is interpreting the information – perhaps calling on algorithms again to help follow the snaking paths of each individual nerve cell or neuron – and deciphering clues about common neuronal patterns found in ‘higher’ organisms like humans.

Written by John Ankers

Image from Janelia / MDC

Berlin Institute for Medical Systems Biology, Max Delbrück Center for Molecular Medicine in the Helmholtz Association, Berlin, Germany

Image copyright held by the original authors

Research published in Nature Methods, August 2019

by Ana Santos Rutschman

Is this face just an assembly of computer bits? PHOTOCREO Michal Bednarek/Shutterstock.com

When artificial intelligence systems start getting creative, they can create great things – and scary ones. Take, for instance, an AI program that let web users compose music along with a virtual Johann Sebastian Bach by entering notes into a program that generates Bach-like harmonies to match them.

Run by Google, the app drew great praise for being groundbreaking and fun to play with. It also attracted criticism, and raised concerns about AI’s dangers.

My study of how emerging technologies affect people’s lives has taught me that the problems go beyond the admittedly large concern about whether algorithms can really create music or art in general. Some complaints seemed small, but really weren’t, like observations that Google’s AI was breaking basic rules of music composition.

In fact, efforts to have computers mimic the behavior of actual people can be confusing and potentially harmful.