Playback speed
×
Share post
Share post at current time
0:00
/
0:00
Transcript

Diagnosing Psychiatric Conditions with ~70%+ Accuracy?

June 2024 Neurobane Newsletter

In 4 minutes catch up with what we’ve been up to during the month of June : )

Key Points:

  • Completed the first iteration of our proof of concept algorithm.

    • The model uses 1300 EEG brain scans and has relatively high accuracy with f1-scores of 81%, 85%, and 70% for diagnosing Addiction, Mood Disorders, and Trauma-related disorders.

  • We are raising a $16,000 Angel Round to develop our proof-of-concept headset and collect pilot data.

    • By the end of this week, we’ll have a presentation deck to send to angel investors. The deck is picture-heavy and focuses on the problem and solution rather than the technical bits.

Looking Ahead:

  • Building a deep learning model with more data during July and testing if the model is better than our current model(Overfitting may be an issue).

    • Integrating the model into a live web app, connecting it to our hardware, and getting feedback from psychiatrists in the months after.

  • Settling on a business model(B2B or B2C). Having some trouble with this bit.

  • Raising our pre-seed round once we have the POC ready and its accompanying data to start our clinical trial and completely change the state of psychiatric care, from notoriously inconsistent to incredibly reliable.

Neurobane Newsletter
Neurobane Newsletter
Authors
Ahad Hassan