Automated System Developed to Monitor Drug-Induced Comas

Emery Brown (Mass. Institute of Technology)

Engineers at Massachusetts Institute of Technology and medical researchers at Massachusetts General Hospital in Boston developed and tested in lab animals a system connecting the brain to a drug infusion device that automatically controls anesthesia drugs administered to patients in a drug-induced coma. The team led by MIT engineering professor Emery Brown, who is also on the faculty at Harvard Medical School, describe their system online in yesterday’s issue of the journal PLoS Computational Biology.

In some cases of traumatic brain injury, patient are placed in a drug-induced coma to allow swelling in the brain to subside and give time for healing. During that time, the patient must be continuously monitored to keep the brain at the proper level of sedation, which with current technology requires frequent checks and adjustments by trained staff, for hours or sometimes days at a time.

Emery and colleagues previously analyzed electrical waves generated by the brain and measured by an  electroencephalogram (EEG) during various states, such as when awake, asleep, or under anesthesia. When in a medically-induced coma, the patient’s brain is quiet for several seconds, followed by short bursts of energy, and staff monitoring the patient need to control the flow of anesthesia drugs in response to the EEG readings to minimize these energy bursts.

While anesthesiologists regularly use computerized equipment, there is no system yet approved by FDA to completely control the administration of anesthesia drugs based on the patient’s brain activity. The MIT/Mass. General team aimed to fill that gap by providing the ability to both monitor EEG activity and control the anesthesia.

The system is based on a connection between the brain, monitored by an EEG, and a drug infusion pump, controlled by algorithms on a computer that adjust the flow of the drug propofol at a precise target level of activity. The connection makes it possible to change drug flow second-by-second to keep it within those target levels.

The MIT/Mass. General team tested the system with rodents, first determining individual target levels for each tested rodent, then inducing and maintaining comas in each animal. The researchers report a performance error for the brain-machine interface of 3.6 percent.

The system can quickly increase the depth of the coma if needed, which the researchers say would be impossible for a human monitor to do accurately. An administrator can also program the controller to bring out a patient from a comatose state, if needed by doctors to perform tests.

Brown says a system of this kind could make it possible to maintain a medically-induced coma with lower doses of anesthetic drugs and reduce the need for constant monitoring by trained nurses, with benefits for overall staffing at hospital intensive care units. “We’re now in discussions with the FDA for approval to start testing this in patients,” says Brown in a Mass. General statement. Mass. General also applied for a patent on the technology.

Read more:

• Transparent Skull Implant Devised for Laser Brain Treatments
• Prototype Robotic Brain Blood-Clot Surgery Device Developed
• Injectable LEDs Developed to Study Brain Functions
• Wireless Brain Sensor Designed, Tested in Animals
• U.S. Alzheimer’s Patient Implanted with Deep Brain Pacemaker

*     *     *