Displaced, multi-part intracapsular, proximal humerus fractures represent a major challenge for patients and orthopedic surgeons. Proximal humerus fractures represent the third most common fracture after hip and distal radius fractures, and more than 20% of these fractures meet operative indications. Unfortunately, one of the major complications of these fractures is the development of avascular necrosis (AVN), or death of the bone as a result of the loss of blood supply to it. Currently, the ability to predict AVN is limited. The purpose of this study is to determine if computed tomography imaging can identify a quantifiable predictor of AVN following this type of humeral fracture.

Current surgical treatment of humeral head fracture does not follow a universally accepted treatment algorithm. Treatment guidelines are lacking sufficient prospective controlled evidence to support any one option as doctrine. However, the generally accepted dogma regarding the surgical treatment of fractures considered high-risk for Avascular Necrosis (AVN), and also symptomatic avascular necrosis of the humeral head, is shoulder replacement (arthroplasty). This is a difficult scenario for younger patients where humeral head preservation is the priority, because of a expected lifespan that will require one, if not multiple, arthroplasty revisions. There is data suggesting no difference in functional outcome for complx humeral head fractures treated with open reduction and internal fixation (ORIF), that late progress to AVN, and patients who undergo hemi-arthroplasty, because of risk of AVN. However, failure of surgery related to AVN remains a challenging problem often leading to secondary surgery, which is not without negative outcomes. This study is aimed at trying to come up with a predictor that can be applied to patients prior to initial humeral head fracture treatment.

The specific aims of the study to help determine this predictor are:

1. In patients who have humeral head fracture, evaluate the difference in fracture fragment enhancement in multi-part fractures, by comparing the fracture fragment attenuation on iodine map overlay to the non-traumatized, contralateral humeral head using Dual Energy Computed Tomography (DECT).
2. In patients undergoing ORIF or joint replacement for humeral head fracture without known risk factors for AVN, identify the relationship between the attenuation due to iodine in the fractured humeral head, as measured preoperatively by DECT, with the presence of pulsatile blood flow in the humeral head intra-operatively, as measured with laser doppler.
3. In patients undergoing joint replacement for humeral head fracture without other known risk factors for AVN, identify the relationship between the attenuation due to iodine in the fractured humeral head, as measured preoperatively by DECT, with the presence of AVN in humeral head pathology specimen.
4. In patients who underwent ORIF for humeral head fracture without other known risk factors for AVN, identify the relationship between the attenuation due to iodine within the fractured humeral head, as measured preoperatively by DECT, with the incidence of radiographically apparent AVN after 2 years of clinical follow-up.

Inclusion Criteria:

• Patients with multi-part fracture of proximal humerus.

Exclusion Criteria:

• Patients with a humeral head prosthesis, hardware or avascular necrosis in the ipsilateral shoulder.

Consenting adult patients with proximal humerus fractures.