Autologous Ear Reconstruction: A Semi-Automatic Procedure for Patient-Specific Surgical Guides

Autologous ear reconstruction is the preferred treatment to repair the deformed or missing ear anatomy due to a trauma, a tumor intervention, or a congenital malformation (e.g. microtia). Such a procedure involves: 1) removing a portion of the patient's costal cartilage, 2) manually cutting, carving and suturing the ear “framework”, and 3) placing the framework inside a skin pocket in the auricular region. The framework consists of three segments (i.e. helix, antihelix, tragus-antitragus), which replicate anatomical ear elements, and a base which act as a support. Developing a suitable framework can be very challenging, so the aesthetic outcomes are highly dependent on the capacity of the surgeon to carve and shape the cartilaginous tissue. The traditional procedure requires the use of a 2D template to support the ear model replication by the surgeon. The 2D template is obtained by placing a 2D X-ray film above the healthy ear and by drawing its contours. Unsurprisingly, this template misses of some relevant information on the ear morphology, such as thickness and height. For this reason, the result strongly dependents on the surgeon artistic and technical skills and on the capacity to draw out each anatomical element. To overcome these limitations, this study aims to develop a new approach based on patient-specific surgical guides capable to help the surgeon to execute a guided surgery. In detail, a semi-automatic procedure to generate the 3D CAD models of each anatomical element is proposed. The physical replicas of these elements (using, for instance, additive manufacturing techniques) can be used in the operating room to guide the tracing of contours on the harvested costal cartilage and its subsequent carving.

The procedure starts from the 3D model of a reference ear obtained by scanning and mirroring the healthy ear, in case of mono-lateral defect, or by scanning the parent’s ear. The mesh obtained by the scanning is then automatically oriented to maximize the visible portion of each anatomical element involved in the ear reconstruction surgery. The definition of this plane is an essential step of the procedure since all the segments, involved in the reconstruction phases, are built from the 2D ear projection on this plane. The procedure requires the manual selection of six easily identifiable reference points, namely: three points on the helix in correspondence of the extremities of this element and other three points located on the tragus-antitragus element. Such points are then used to define a new set of (automatically generated) points from which the profile of each segment is built.  The so generated profiles are extruded according to the ear morphology and, depending on the segment, some typical CAD modeling operations, such as chamfer, fillet etc., are automatically performed.

The CAD modelling process is designed to be handled by non-expert users (i.e. the medical staff); this means that the non-trivial CAD-based reconstruction is totally hidden to them. In fact, the above-mentioned operations are controlled via software by means of an appositely devised graphical user interface (GUI). The system was tested in the generation of ear templates for a panel of ten people, five females and five males. The panel group allowed the modeling method to be validated on significantly different ear geometries. Obtained results prove the effectiveness of the proposed method in better designing and reconstructing the deformed or missing ear of patients.