AI and biology: deep learning for microscopy image analysis with Prof. Constantin Pape

In this episode, we talk to Prof. Constantin Pape, leader of the Computational Cell Analytics group at the Institute of Computer Science and a Junior Professor at the Georg-August University Goettingen. We explore the intersection of machine learning and biology, specifically focusing on the analysis of microscopy images. Prof. Pape explains the principles of the artificial neural networks and sheds light on the challenges accompanying their use in scientific research. Our conversation extends to the latest AI developments from big companies like OpenAI and Meta and how they can help scientists. Towards the end, Prof. Pape shares with us his thoughts on the issues with the academic publishing system, drawing comparisons between the fields of biology and machine learning.
Our channel is supported by International Max Planck Research School (IMPRS) for Neurosciences as well as Cluster of Excellence Multiscale Bioimaging (MBExC) in Göttingen, Germany. The views presented on this channel are not representative of IMPRS or MBExC.

Timestamps:
00:00:00 Intro
00:02:00 Early career and motivation for doing science
00:04:46 Working culture in the USA (Janelia Research Campus) and Germany 00:11:21 Research interest: automating analysis of microscopy images
00:13:55 Differences between light microscopes and electron microscopes
00:20:11 Image analysis in biomedical research and algorithms used for it
00:27:38 BigData in neuroscience: working with thousands of images to reconstruct the neuronal connections in the brain
00:35:39 Challenges in developing generalised image analysis algorithms
00:45:10 Segment Anything: general algorithm for identifying objects in images 00:49:24 Programming languages for image analysis: advice for beginners
00:53:14 Next big thing in machine learning: multimodality, combining different types of data
00:59:01 Problems in academia: publishing system
01:08:50 Problems in academia: postdoc crisis and funding

References: ImageJ/Fiji; Schindelin, J., Arganda-Carreras, I., Frise, E. et al. Fiji: an open-source platform for biological-image analysis. Nat Methods 9, 676–682 (2012). doi.org/10.1038/nmeth.2019

Read more about: Segment Anything by Meta segment-anything.com ChatGPT by OpenAI openai.com/blog/chatgpt

Intro music by penguinmusic from Pixabay Following footage downloaded from videvo.net: Storage of samples in a hospital setting. Ungraded Prores file by MaartenLCF A scientist adjusting the magnification on a microscope by Videvo Stock footage provided by Pressmaster Like and subscribe animation downloaded from SweetPhilMedia

Follow us on social media for the latest updates! Twitter: twitter.com/neuro_beyond Instagram: instagram.com/neurosciencebeyond TikTok: tiktok.com/@neurosciencebeyond Facebook: facebook.com/neurosciencebeyond

Neuroscience and Beyond team: Svilen Georgiev Kristina Jevdokimenko Sabina Nowakowska #neuroscience #brain #science #podcast #göttingen #IMPRS #MBExC #AcademicChatter #SciComm #PhD