Development of a YOLOv9 model with angle loss for automatic detection of landmarks and cephalometric analysis

Development of a YOLOv9 model with angle loss for automatic detection of landmarks and cephalometric analysis. Download PDF Abstract Artificial intelligence has been widely applied to identify anatomical landmarks on lateral cephalometric radiographs, reducing localization errors and improving the efficiency of cephalometric analysis. This study aimed to develop and evaluate an automated cephalometric landmark detection system based on You Only Look Once version 9 (YOLOv9) with an angle-based loss function and automated Steiner cephalometric analysis using radiographs from an Indonesian population.

The proposed angle-based loss was incorporated into YOLOv9 to enforce geometric consistency among anatomically related landmarks. Model performance was evaluated using mean radial error (MRE), successful detection rate (SDR), and mean average precision (mAP). Automated Steiner measurements derived from the predicted landmarks were compared with expert annotations using the mean absolute error (MAE).

The proposed model achieved an overall MRE of 0.99 mm, an SDR of 86.8% at a 2-mm threshold, and a mAP of 0.754. Automated Steiner analysis yielded a mean absolute error of 1.38° compared with expert measurements. Compared with the baseline YOLOv9 model, the proposed angle-based loss modestly improved overall landmark localization performance while preserving anatomical relationships among predicted landmarks.

Posted by Matthew McCullough, VP, Product Management, Android Developer When we first launched Android Bench, we built a rigorous foundation for evaluating how large language models (LLMs) assist developers with real-world Android tasks. As AI models and agents rapidly evolve, we’ve been updating our methodology, such as aligning our benchmark framework with the Harbor framework.

Today we’re releasing the first set of long-horizon tasks (LHT), which are tasks of great complexity that take an engineer multiple days or even a week to complete. We are also introducing agentic evaluation, starting with agents from corresponding model providers.

This addition brings us to Android Bench 2.0 —a major upgrade designed to evaluate AI models and agents against the scale, ambiguity, and complex multi-step problem solving that you tackle every day. The Android Bench 2.0 leaderboard From incremental fixes to long-horizon tasks The first iteration of Android Bench, along with similar early AI coding benchmarks, focused on incremental changes to existing repositories, in many cases limited to bug fixes or smaller feature requests.

This was a reflection of the capabilities of AI assistance at the time, as well as how you were using it. To continue helping you find the models and coding agents best suited to your development workflow, we have raised the bar of our evaluations to match the work you delegate to AI.

Android Bench 2.0 mirrors these ambitious challenges with LHTs that include upgrading dependencies, adding new features, building apps from scratch, or converting a cross-platform app to Android. Complex tasks require a more nuanced evaluation and scoring On multi-day engineering tasks, binary pass or fail grading doesn’t capture the full picture.

For example, an agent might refactor 40 screens to Jetpack Compose, set up database tables, and pass 90% of requirements, but fail a single edge-case assertion. Binary scoring rates this run as 0%, obscuring the model’s architectural capabilities.

We are moving to continuous scoring to provide a more meaningful signal, both for model development and for your understanding of how AI can help you. We calculate this completion rate through a combination of factors like functionality, visual fidelity, and avoiding regressions.

We also apply objective scoring penalties for deviations from evaluation instructions or structural constraints. Check out the updated leaderboard and click into each model’s card view to see additional elements such as the pass rate, completion rate, and average costs per model and per task.

The highest pass rate for LHTs is around 28%, much lower than the ~91% for the original tasks in the benchmark. The model card view allows you to explore the strengths and pitfalls of each model Long-horizon tasks uncover helpful insights for AI assistance Beyond measuring how well AI handles long-running tasks, the LHT dataset helps us learn more about the strengths and weaknesses of tested models, and we offer you more practical guidance.

Across model tiers, AI does a better job at writing new code rather than refactoring existing code. Refactors and migrations get trickier because success depends on architectural complexity rather than code volume.

Models show strong capabilities on well-established, deterministic transformations, such as converting Java to Kotlin, swapping Retrofit for Ktor, or introducing a ViewModel layer. They apply these patterns consistently, even across 125+ files and 8,000+ lines of code.

Medical Disclaimer: This article is for informational purposes only and does not constitute medical advice. Always consult with qualified healthcare professionals for medical decisions and treatment options.