Episode 49

Machine Learning (Epilepsy)

Dr. Gerold Baier

Jul 16, 2021 | Science

Dr. Baier joins me to talk about his research in algorithmic approaches to disease.

Listen on Spotify
I think the current scientific approach of being the isolated scientist who watches the brain, and infers things from it, will not be successful in the long run.  You can try to interact with the brain, and learn, not via concluding as a physicist, but as becoming part of that rhythmic interaction.  It's like saying, "I know nothing, tell me something."

About This Episode

Dr. Gerold Baier joins me from the UK to discuss machine learning, and his research into algorithmic approaches in disease. He’s been researching rhythms and patterns in epilepsy with the hope that one day, we can predict seizures before they happen!

Topics and Timestamps

  • (0:00:55) Biochemistry and decoding the rhythms of the human brain.
  • (0:02:13) The MEG and the study of epilepsy.
  • (0:03:45) Using rhythm and audio recordings.
  • (0:05:10) Epilepsy and the rhythms of seizures.
  • (0:06:07) The goals: Being able to stop a seizure.
  • (0:07:36) Implantation and the future. 
  • (0:08:36) Surgery vs the needle: Lessening the destruction of brain tissue.
  • (0:10:04) The current solution: Cutting and burning the tissue.
  • (0:11:13) Machine learning and data science.
  • (0:11:58) What is machine learning?
  • (0:14:22) Isolating the rhythms.
  • (0:15:03) What causes a seizure?
  • (0:16:10) Where we are now with research: Identifying and classifying.
  • (0:18:03) The upcoming study in Beijing.
  • (0:19:20) The current scientific approach: A new approach is needed.
  • (0:21:09) Algorithms: The mathematical vs the physiological.
  • (0:21:53) Brain computer interfaces.
  • (0:23:10) Are there any limits to understanding the human brain?
  • (0:24:05) The auditory background in your brain.
  • (0:27:13) Psychedelics and the brain.
  • (0:30:31) Meditation, the brain and your inner self.
  • (0:31:49) Creativity and the collaboration of networks.
  • (0:33:38) Why use Python in machine learning?
  • (0:35:20) Where Python is used: Teaching and research.
  • (0:36:11) Python: Not the only language used.
  • (0:37:10) The future and artificial intelligence: A good idea?
  • (0:39:36) Seeing the danger and vulnerability.
  • (0:42:03) Decoding the diseases: How close are we?
  • (0:43:50) The consuming process of data points.
  • (0:45:06) Studying nature. 
  • (0:46:17) Mexico and settling in the United Kingdom.
  • (0:47:40) Other collaborations: Working with the Children’s Hospital.
  • (0:49:01) Teaching in China.

Follow Planet B612 on Twitter: https://twitter.com/PlanetB612fm

Support Planet B612 on Patreon: https://patreon.com/juliesworld

Related Episodes

Ep. 52: Reproductive Immunology with Dr. Melanie Conrad

Ep. 52: Reproductive Immunology with Dr. Melanie Conrad

Episode 52 Reproductive Immunology Dr. Melanie ConradIn this episode, we talk all about the impact of the environment on asthma, antibiotics in pregnancy, and all of the cool research being done by Dr. Conrad and her team! About This Episode What role might the...

Ep. 51: Tardigrades with Dr. Frank Smith

Ep. 51: Tardigrades with Dr. Frank Smith

Episode 51 Tardigrades Dr. Frank SmithIn this episode, we talk all about Tardigrades, where to find them, how they've evolved, and more! About This Episode How did Tardigrades evolve? Where can we find them? Are there Tardigrades in Space? All this, and more, with...