The combination of artificial intelligence (AI) and medical image processing is revolutionizing maxillofacial surgery. Automated modeling methods enable the efficient and precise creation of personalized patient models. This facilitates improved diagnoses, tailored treatments, and faster surgical preparations.
Patient-specific modeling holds significant potential for diagnosis and treatment, marking a crucial step toward personalized medicine. In maxillofacial surgery, 3D models of the anatomy under examination not only offer intuitive visualization but also form the foundation for surgical planning and the design of patient-specific implants. In practice, medical image data is modeled by experts using software tools—a time-consuming and expensive process. Thanks to advances in AI-based annotation, comprehensive solutions are now technically feasible. However, their application outside of research remains limited due to the time-intensive preparation of suitable datasets for training these methods. Through a collaboration between RISC Software GmbH and CADS GmbH, the development of automatic modeling methods was efficiently implemented using an AI-supported pipeline.
As part of its research activities, the Medical Informatics Unit at RISC Software GmbH is developing an AI-supported pipeline for modeling medical datasets. Unlike comparable technologies, such as the AI platform MONAI [1], this pipeline also supports aspects of method development, beginning with the review, selection, and stratification of the data basis. It allows the chaining of any AI-based and classical image processing methods into workflows that can be applied to individual patient data as well as entire datasets. Partial results, such as training AI methods or calculating atlas datasets, are standalone workflows as well as final modeling methods for integration into end applications. This approach leads to maximum reusability of workflows.
Figure 1: The pipeline’s flexible modular concept allows processing steps to be linked into workflows. These workflows can, in turn, be embedded as sub-steps into overarching workflows.
The pipeline has already been successfully applied in various research projects, such as the automatic modeling of aneurysm patients for the MEDUSA surgical simulator [2]. In these projects, the pipeline enables the automated creation of complex patient models, supporting further anatomy visualization, automated feature calculation, phantom printing, and hybrid surgical simulation. By integrating modern AI methods like the nnU-Net framework [3], the pipeline achieves state-of-the-art results.
Figure 2: Modeling of aneurysm patients in the MEDUSA research project.
Figure 3: Modeling of patients in maxillofacial surgery from the collaboration with CADS GmbH.
Within the collaboration, the AI-supported pipeline was used for the annotation process, as well as for training and validating AI methods. Specifically, the annotation process benefited from the automated preparation of image data for external providers and subsequent validation and integration of the annotations into the dataset. Training, validation, and statistical evaluation of the AI methods were also fully automated and can easily be applied to the growing dataset. The results were presented at AAPR 2023 [4], convincing the medical experts of the project partner. Finally, the pipeline was embedded into an annotation platform by CADS GmbH, enabling medical experts to automatically generate patient-specific models from image data in the future. The pipeline also supports the continuous expansion of the dataset with new annotations and the subsequent optimization of AI methods, allowing the AI methods to be continuously improved and additional anatomical structures to be incorporated into the dataset. A particular focus lies on the automatic evaluation and selection of image data based on quality criteria regarding its suitability for modeling, enhancing the efficiency of the annotation platform for its users.
Figure 4: Examples of automatically modeled datasets from patients in maxillofacial surgery.
Researcher & Developer
