DSM–Neuro–Pharm Explorer

Inspiration

My intensive training across psychology, psychotherapy and neuroscience inspired me to bring clinical, neurobiological and pharmacological knowledge together in one modern educational reference.

Clinical education often separates symptom recognition, diagnostic frameworks, brain mechanisms and pharmacology into different resources. I wanted to explore how these layers could be connected through a transparent workflow that remains informative, current and clinically responsible without pretending to provide a diagnosis or treatment recommendation.

I have also begun working with MATLAB and Python to simulate artificial neural networks for future research and potential applications in psychotherapy. DSM–Neuro–Pharm Explorer is the transparent, rule-based foundation for that broader research direction; the current prototype does not use an ANN or automatically interpret free text.

What it does

Users select standardized symptom descriptions and add clinical context such as duration, functional impact, activated mood periods, trauma history, substance or medication exposure and medical or neurological factors.

How we built it

The prototype uses static HTML, CSS and JavaScript with explicit, human-readable rule tables. This makes the routing logic transparent and allows symptoms, contextual weights and safeguards to be inspected.

Codex with GPT-5.6 supported the iterative design and implementation process, including interface development, rule organization, debugging, safety language, responsive design, edge-case analysis and final browser-flow validation.

The finished application is deployed through GitHub Pages and requires no backend, installation or API key.

Challenges we ran into

The main challenge was balancing educational usefulness with clinical safety. Contextual information needed to support relevant routes without being counted as an additional checkbox symptom.

Mixed patterns and weak signals required safeguards so that the application would not narrow arbitrarily. It was also important to distinguish conditional educational considerations from clinical diagnoses.

Another challenge was connecting symptom dimensions with neuroscience and pharmacology while clearly presenting these sections as broad educational summaries rather than individualized treatment recommendations.

Accomplishments that we're proud of

• 45 standardized symptom options across 9 visible groups.
• Seven broad dimensional routes.
• A focused multi-stage interface with separate symptom, follow-up and results views.
• Transparent context support for duration, activation, trauma and functional impact.
• Conditional DSM chapter considerations and differential flags.
• Persistent urgent-safety messaging.
• Browser-only privacy with no server-side form-data collection.
• Responsive desktop and mobile layouts.
• Syntax, routing and browser-flow validation.
• A public live deployment, documented source repository and short demonstration video.

What we learned

Transparent rules make clinical-education logic easier to inspect, test and challenge. Duration and functional impact are essential clinical-context variables. In this educational prototype, they refine the symptom-based route without independently determining an outcome, and their contribution is displayed transparently. Free-text symptom interpretation should require a privacy-preserving mapping layer and explicit user confirmation before affecting rule-based results.

What's next for DSM–Neuro–Pharm Explorer

This pilot will gradually expand its curated symptom bank, DSM chapter coverage and focused follow-up questions. Future releases may include neurodevelopmental and additional disorder families, broader expert-reviewed neuroscience and pharmacology profiles, accessibility testing and structured evaluation cases.

A later research stage could explore MATLAB- and Python-based artificial neural networks for pattern analysis, together with a privacy-preserving NLP service that suggests standardized symptoms from optional free text. Any suggested mapping would require explicit user confirmation and careful expert evaluation.

The long-term goal is an expanding, current educational knowledge explorer that connects clinical psychiatry, neuroscience and pharmacology while remaining transparent, privacy-conscious and non-diagnostic.

Built With

• codex
• css
• github
• gpt-5.6
• html
• javascript