EHRs improve clinical quality by providing ready access to all relevant clinical information at the time of the patient encounter or phone call, receipt of clinical alerts at the point of care, the ability to easily monitor and analyze patient outcomes. The EHR software's currently available in the markets are very expensive and require extensive training before physicians can use these systems. Also all these systems require the physician or nurse to enter the data in the record manually.

We have developed a smart context-aware system to semi-automatically build an EHR that records the medically significant events of a surgery. The system analyzes the data streams obtained from various sensors deployed in an Operating Room (OR) in real-time to detect events. We refer to this electronic record as the Electronic Medical Encounter Record (EMR). This record then becomes a part of the patient's medical history and will provide the next physician an accurate account of the medical treatment given to the patient. Sensors in the OR include the blood oxygen monitor, the heart rate monitor etc.

Data from these sensor streams is analyzed using a stream processing engine to extract the low level events, such as high blood pressure etc, occurring during the surgery. These events are correlated using techniques such as multi-variable analysis, trend based analysis etc to identify events that become a part of the electronic medical record. Radio Frequency Identification (RFID) is used to acquire contextual information such as presence of medical staff in the operating room and identification of medicines used during the surgery.

The prototype of the system will be used to create an EMR in Traumapod, robotic system that provides trauma care to wounded soldiers on the battlefield, and Context-Aware Surgical Training System (CAST), where the EMR is used to evaluate the performance of the trainee during the surgery.