The objective of this study is to design a library of biomechanical musculoskeletal models of patients before and after surgery. These models will be evaluated to analyze the support they can provide for virtual surgical planning in orthopedics.

While there are multiple surgical options for the treatment of a specific orthopaedic pathology, the patient is most often operated on according to the preferred method used within a team and not necessarily according to the method that would optimize the functional result. In addition, the variability of morphologies means that some patients, due to their anatomical characteristics, cannot benefit from an intervention that is necessary (incorrectly sized implant, unsuitable equipment). As the orthopaedic surgical procedure is definitive, the investigators believe that it may be useful to perform surgical simulations on a digital clone before performing on the real patient. This would be relevant for the patient, the caregiver, but also the community, as many surgical reopenings or treatment failures can be attributed to the difficulty of properly planning surgery for a functional objective in the long term.

This is a Retrospective study based on patient's record and medical imaging files.

50 patients For 3D medical imaging, the data will be obtained from the examinations performed by the patient before and after his surgery, they will be anonymized. The investigators will use the so-called "DICOM" sequences allowing the reconstruction of 3D models The personal data collected will be taken from the Surgical Report and the pre-surgical check-up (age, sex, height, weight) The patient identifier will be assigned for each subject with the centre number, initials (Surname / First name of the patient) and its inclusion number per centre.

Personal data will be obtained from the department's computerized or paper patient file Medical imaging data will be anonymized by the medical imaging department staff who will be assigned the patient identifier.

The database will be centralized and hosted on a laboratory computer on an encrypted and password-protected session.

The usable patient records will be recruited according to the database of surgical procedures performed in the various departments. If the file meets the inclusion and non-inclusion criteria, patients will then be contacted (telephone) to discuss their participation in the study, and verbally obtain their non-opposition.

Patients will be informed in a complete and fair manner, in understandable terms, of the objectives of the study, their right to refuse to participate in the study or the possibility of withdrawing at any time. All this information will be included on an information form given to the patient by email or post.

Once the authorization has been obtained, the data will be anonymized and downloaded to an encrypted hard disk for processing for model reconstruction.

Clinical data will be retrieved from the operative report and the pre-surgical file and then integrated into the database.

The models resulting from this procedure will bear the name of the surgical procedure associated with a patient identification number.

Morphological comparison of individual bone structures and the outer envelope including internal soft tissues (measurement of distances and volumes) Comparison of radiological parameters used in routine pre-surgery routine (measurement of angles, segment lengths, tissue thickness) Between February 2019 and June 2021, the three surgical teams will search for patients who have benefited from the surgeries and who meet the inclusion criteria.

The distribution will be as follows:

Grenoble University Hospital (12 cases):

• 5 patients Knee
• 5 patients Rachis
• 2 patients Hip

Raymond Poincaré University Hospital (19 cases):

2 patients for tendon transfer procedures

• 1 patient for muscle stretching
• 2 patients for lower limb amputations
• 5 patients Rachis
• 6 patients Foot / Ankle
• 1 patient Shoulder
• 2 patients Elbow / Hand

Groupe Hospitalier Croix saint Simon (19 cases):

• 3 patients Knee
• 2 Patients Hip
• 9 patients Foot / Ankle
• 2 patients Shoulder
• 3 patients Elbow / Hand

From April 2019, the models will begin to be developed by the TIMC IMAG laboratory team.

Once the biomechanical models are developed, the investigators will virtually apply the surgery performed on the pre-surgical models. Then the investigators will compare the virtual post surgery models with the post surgical models. A description of the limitations of the virtual procedures will be established in order to analyze the scientific locks to be removed and move towards biomechanical simulation-based surgical planning.

Inclusion criteria:

• Men or women aged 20 to 65 years
• Patients who have had orthopaedic surgery
• Presence in the file of a preoperative CT scan and/or magnetic resonance imaging (MRI)
• Presence in the file of a CT scan and/or MRI postoperatively at a distance from the operation (at least 4 weeks)
• Presence in the file of an operating report describing the procedure in detail with access routes, operating times, type of equipment and possible complications.
• Patient who has authorized the anonymous use of data from his pre and post surgical examinations
• Patient who is a member of a social security organisation

Criteria for non-inclusion:

• Patient who has expressed his refusal to use his medical data
• Minor, patient under guardianship, inability to receive information and express consent
• Pregnant women
• Non-affiliation to a social security scheme (beneficiary or beneficiary)
• Examinations of images that do not allow 3D reconstruction of structures (poor resolution or type of sequence)
• Incomplete operating report not allowing the understanding of the steps of the surgery.
• Surgical report describing surgery that does not represent the usual procedures performed.

Patients aged 20 to 65 years who require orthopaedic surgery as described in the literature and for which 3D medical imaging will be performed before and after the surgical procedure.