Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Relating formulation variables to in vitro dissolution using an artificial neural network

N K Ebube1, T McCall, Y Chen

  • 1College of Pharmacy and Pharmaceutical Sciences, Florida Agricultural and Mechanical University, Tallahassee 32307-3800, USA.

Pharmaceutical Development and Technology
|August 1, 1997
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Response to Comment on "Permanent human occupation of the central Tibetan Plateau in the early Holocene".

Science (New York, N.Y.)·2017
Same author

Response to Comment on "Permanent human occupation of the central Tibetan Plateau in the early Holocene".

Science (New York, N.Y.)·2017
Same author

Permanent human occupation of the central Tibetan Plateau in the early Holocene.

Science (New York, N.Y.)·2017
Same author

RORγt expression in T<sub>regs</sub> promotes systemic lupus erythematosus via IL-17 secretion, alteration of T<sub>reg</sub> phenotype and suppression of Th2 responses.

Clinical and experimental immunology·2016
Same author

Critical questions concerning the validity of the Bayer study report of "Differences in bioavailability between 60 mg of nifedipine osmotic push-pull systems after fasted and fed administration".

International journal of clinical pharmacology and therapeutics·2010
Same author

Further notes on the leeches (Piscicolidae) living on fresh-water fishes of North America.

Transactions of the American Microscopical Society·2010

Optimizing neural network variables and carefully selecting training data significantly improves predictions of in vitro drug dissolution rates. Accurate dissolution predictions are achievable even with limited formulation data when variables are optimized.

Area of Science:

  • Pharmaceutical Sciences
  • Computational Chemistry
  • Drug Delivery Systems

Background:

  • Predicting in vitro drug dissolution rates is crucial for pharmaceutical development.
  • Neural networks offer a powerful tool for modeling complex relationships between formulation and dissolution.
  • Understanding the impact of experimental variables on neural network performance is key to reliable predictions.

Purpose of the Study:

  • To investigate how experimental variables affect neural network accuracy in predicting in vitro drug dissolution rates.
  • To determine the optimal conditions for training and validating neural network models for dissolution prediction.
  • To assess the influence of training set size, data replication, and validation strategy on prediction accuracy.

Main Methods:

Related Experiment Videos

  • Training neural network software with hypothetical and experimental formulation data.
  • Validating trained models against unseen formulations.
  • Investigating the impact of hidden-layer nodes, iterations, and data types (mean vs. replicate) on prediction accuracy.
  • Analyzing prediction errors based on the range of training and validation data.

    • Main Results:

    • Optimizing the number of hidden-layer nodes and iterations was critical for accurate predictions.
    • Prediction error increased when validation data fell outside the training data range.
    • Accurate predictions were achieved with as few as four carefully selected formulations in the training set.
    • Using replicate data and increasing training set size improved prediction accuracy.
    • A single formulation was insufficient for validating the neural network model.

    Conclusions:

    • Neural network models can accurately predict in vitro drug release.
    • Careful optimization of neural network variables and appropriate selection of training/validation datasets are essential for reliable dissolution rate predictions.
    • The study highlights the importance of data quality and representativeness in developing robust predictive models for pharmaceutical formulations.