HIPAA compliant AI Photo Organizer App

# AI Orthodontic Photo Organizer & Automatic AJODO-Style Layout Web Application

## Project Description

I am an orthodontist looking to develop a **simple AI-powered web application** that automatically organizes and formats a patient's orthodontic photographs into a standardized **AJODO-style clinical photo layout**.

The user will upload approximately **9 intraoral and extraoral patient photographs at the same time**. The application should use AI/computer vision to recognize each photograph, determine its correct orientation, and automatically place it into the appropriate position in a predefined orthodontic photography template.

## Basic Workflow

**Upload 9 photos → AI recognizes each view → Automatically rotate/crop → Arrange in predefined template → Preview/edit → Export final image**

## Main Requirements

### 1. Multiple Image Upload

* Drag-and-drop upload of approximately 9 photographs
* JPG/JPEG/PNG support
* Ability to upload all photographs simultaneously

### 2. AI Photo Recognition

The application should automatically identify common orthodontic views such as:

* Frontal facial photograph
* Smiling facial photograph
* Profile photograph
* Right intraoral photograph
* Frontal intraoral photograph
* Left intraoral photograph
* Maxillary occlusal photograph
* Mandibular occlusal photograph
* Additional facial/clinical photograph as defined by our template

The system should assign every uploaded image to the appropriate position automatically.

### 3. Automatic Image Orientation

The software should recognize whether an image needs to be:

* Rotated 90°
* Rotated 180°
* Rotated 270°
* Flipped/mirrored when necessary

This is particularly important for maxillary and mandibular occlusal photographs.

### 4. Automatic Cropping and Alignment

The application should attempt to:

* Center the patient's face in extraoral photographs
* Center the dental midline in intraoral photographs
* Standardize image size
* Maintain appropriate clinical proportions
* Avoid cutting off important teeth or facial structures

### 5. Predefined Orthodontic Photo Template

After identification, the images should automatically be placed into a standardized layout that I will provide.

The layout should have predefined:

* Image positions
* Image dimensions
* Aspect ratios
* Margins
* Spacing
* Background
* Overall canvas dimensions

### 6. Manual Adjustment

Before exporting, the user should be able to easily:

* Drag photos to another position
* Swap two photographs
* Rotate a photograph
* Zoom in/out
* Reposition/crop the photograph
* Reset the crop

AI should do most of the work, but the user must always have the ability to manually correct the result.

### 7. Missing/Duplicate Photo Detection

The application should warn the user if it detects:

* A missing required view
* Two photographs of the same view
* An image that cannot be confidently classified

Example:

**✓ Frontal**
**✓ Smile**
**✓ Profile**
**✓ Right intraoral**
**✓ Frontal intraoral**
**✓ Left intraoral**
**✓ Upper occlusal**
**⚠ Lower occlusal missing**

### 8. Export

The completed photo layout should be downloadable as a high-resolution:

* JPG/JPEG
* PNG

Ideally, image quality should be sufficient for:

* Clinical documentation
* Presentations
* Publications
* Before-and-after comparisons

## AI / Technical Approach

I am open to recommendations regarding the technology.

Possible technologies include:

* OpenAI Vision API or another vision model for photograph classification
* Python
* OpenCV
* Pillow
* React / Next.js or similar web frontend
* Face/dental landmark detection if helpful

**I am not looking for generative AI to modify or recreate the patient's teeth or face.**

AI should primarily be used to **recognize, classify, orient, and assist in positioning the original clinical photographs**.

## Privacy / Security

Because these are patient clinical photographs, privacy and security are important.

The developer should have experience with or understand:

* Secure image uploads
* Encryption
* Temporary image processing
* Automatic deletion of patient images when appropriate
* HIPAA considerations

The initial prototype can be developed and tested using de-identified/sample images.

## Phase 1 — MVP

The first version should focus on:

**Upload → Classify → Rotate → Arrange → Manual Adjust → Export**

I prefer to first build a reliable, simple application rather than an overly complicated system.

## Future Features

If Phase 1 is successful, I may continue working with the same developer to add:

* Pretreatment vs. posttreatment photo comparison
* Automatic before-and-after layouts
* Progress photo layouts
* Automatic brightness/exposure correction
* Detection of poor-quality photographs
* Facial/dental midline alignment
* Standardization of magnification between treatment stages
* Publication-specific templates
* PowerPoint/social-media export formats
* Patient record integration

## Developer Requirements

Preferred developer should have experience with:

* Computer vision / image recognition
* AI vision APIs
* Python / OpenCV
* Image cropping and manipulation
* Web application development
* Drag-and-drop interfaces
* High-resolution image export

Experience with **medical, dental, photographic, or healthcare applications** is a plus but not required.

## When Applying

Please include:

1. Examples of similar computer-vision or image-processing applications you have developed.
2. Your recommended technology stack.
3. How you would classify the 9 orthodontic photographs automatically.
4. Whether you recommend using an existing vision AI model or training a custom model.
5. How you would handle automatic rotation, cropping, and positioning.
6. Estimated cost for the **Phase 1 MVP**.

Please do not submit a generic proposal. I am specifically looking for someone experienced with **AI image recognition and image processing**, not simply website development.