Wireless BCI for Disables

1
Wireless Brain Monitoring in the Emergency
Department
Authors Aveh Bastani, MD1 Hani Kayyali, MS,
MBA2 Robert N. Schmidt, MS, MBA, JD, PE2
Rizwan Qadir, MD1 Prasanth Manthena,
MD3 Institutions 1 Troy Beaumont Hospital, Troy,
MI 2 Cleveland Medical Devices, Inc.,
Cleveland, OH 3 Northwestern University,
Chicago, IL
2
Objective

• To evaluate the feasibility, quality and utility
  of four-channels of electroencephalogram (EEG)
  telemetered from patients presenting with Altered
  Metal Status in the Emergency Department.

3
Background

• Many modalities exist to evaluate neurologic
  status of patients such as
• Bispectral Index (BIS),
• Cerebral Oximetry, or
• Positron Emission Tomography (PET).
• Despite some optimistic trials in the Emergency
  Department (ED) to evaluate the neurologic status
  of patients, none has supplanted the need for the
  Electroencephalogram (EEG).
• EEG is the gold-standard for objectively
  evaluating the functional neurologic status of
  patients.

4
Background - Continued

• Only a handful of hospitals in the US routinely
  perform EEGs in the ED. Subsequently, patients
  in whom an EEG is required are admitted to the
  hospital with their potential disorder
  undiagnosed and untreated for many hours or days.
• This Altered Mental Status group currently makes
  up 10 or 14 million of the 140 million yearly ED
  visits in the U.S.

5
Background - Continued

• Most emergency departments (ED) do not perform
  electroencephalogram (EEG) studies. This is due
  to several inhibiting factors including
• The bulk of the equipment makes it inconvenient
  to be permanently located in an ED setting,
• The cost of equipment at 20,000-40,000 per unit
  is very expensive for most ED budgets,
• The time and expertise required to set up and
  monitor an EEG is typically lacking in the ED,
  and
• ED personnel are not trained to read EEGs and a
  neurologist may not be immediately available to
  read the EEG.

6
New EEG Equipment is Needed for the Emergency
Department

• The Crystal Monitor was developed under US
  National Institutes of Health Grants to have an
  ED EEG device
• Wireless, small, unobtrusive
• Low Cost, 1/2-1/3 of most EEG machines
• Fast, easy to use, limited number of channels
• Easy to learn how to use

7
The Study

• Guided by NIH recommendations and support,
  Cleveland Medical Devices Inc. (CleveMed) has
  created a portable telemetry multi-channel EEG
  monitor. A four-channel montage (Fp1-C3, Fp2-C4,
  C3-O1, C4-O2, Gnd FpZ) was used to maximize EEG
  coverage while minimizing electrode set-up time.
• Telemetry allows the patient to be un-tethered
  and moved about freely while still being
  monitored, an important requirement for any
  patient being assessed in the ED. An internet
  connection allows a neurologist to interpret the
  EEG from anywhere.

8
Patient with Crystal Monitor 16 (older version)
9
Inclusion Criteria

• The following patients will be eligible for study
  inclusion
• Patients with known seizure disorder of any type,
  but with prolonged (gt 1 hr.) post-ictal mental
  status change.
• Patients with status epilepticus who have
  received a muscle relaxant for intubation to
  determine the presence of subclinical seizures.
• Patients with brief alteration of mental status
  of unknown origin. (This group includes new
  onset seizure disorder, syncope, spells,
  blackouts, etc.)
• Patients with behavioral changes that may
  indicate nonconvulsive seizures (impaired
  consciousness, violent outbursts, unusual
  behaviors, etc.)
• Acute head injury patients with mental status
  changes that may indicate nonconvulsive seizures.
  
• Patients with a history of previous head injury
  presenting with new onset mental status changes.
  (Head injured patients are at risk for
  post-traumatic seizures).
• Patients with neurological exams that may be
  consistent with focal or partial nonconvulsive
  seizure. (Eg aphasia, Todds paralysis, etc.)

10
Exclusion Criteria

• The following patients will be excluded from the
  above groups.
• Patients who are convulsing.
• Medically or surgically unstable patients.
• Family member, other authorized representative
  unable to give informed consent.
• Patients with a head injury incompatible with the
  use of EEG (eg gunshots, severe scalp abrasions,
  etc.)

11
Transmission and Reception

• After EEG was completed, the data was password
  encrypted and transmitted to one of two study
  neurologists.
• The neurologist would then provide a
  real-time read for the EEG via telephone
  conversation or email.
• The neurologist also subjectively evaluated
  the quality of the EEG utilizing the following
  four point scale
• ? 4 Excellent quality/Acceptable
• ? 3 Good quality/Acceptable
• ? 2 Fair quality/Acceptable
• ? 1 Poor quality/Unacceptable
• Patients were followed to either attain their
  discharge diagnosis from the ED or the hospital
  in the case of admission.

12
48.6 of the Subjects had EEGs that were
Abnormal

• EEG interpretations (5 unusable EEGs were not
  included)
• 37/72 (51.4) EEGs were interpreted as
  normal
• 2/37 were diagnosed as pseudoseizure by the
  ED physician
• 28/72 (38.9) EEGs were interpreted as
  slowing
• 11/28 were patients who clinically appeared
  post-ictal
• 7/72 (9.7) EEGs identified a sub clinical
  epileptogenic foci
• Correlation with Standard Inpatient EEG
• 24/77 (31.1) patients with EEG had an
  inpatient EEG
• 18/24 (75) were equivalent to the study EEG
• The six dissimilar results are described below

13
75 of the EEGs Corresponded with the 32 Channel
Clinical EEG
Six ED EEGs were different than the EEG Lab
results that were performed later
14
Discussion and Future Considerations

• Understanding that EEG is a time-sensitive
  modality, it is important that we perform EEGs
  when they can be most useful, i.e. in the acute
  setting.
• No enrolled patient failed to complete an EEG.
  Only 5 of 77 patients (6.5) had unusable EEGs
  primarily due to combination of muscular artifact
  and gaps in the data for interference during
  wireless transmission.
• An improved radio that can re-transmit lost
  packets has been developed (Crystal Monitor Model
  20) and will be used for the second half of this
  study.
• Based on this data we believe that ED EEG
  provides valuable information to the ED
  physician, which can expedite safe medical care.
  We do not assert that a four-channel EEG is
  superior or equivalent to the standard EEG.
• We do believe its use as a screening tool in the
  ED provides the ED physician with the additional
  information necessary to make a more appropriate
  disposition from the ED.

15
Crystal Monitor 20 Specifications

• Dimensions 135 mm x 63 mm x 25 mm (5.3 x 2.5
  x 1) (not including antenna)
• Weight 210 grams (6.4 oz.) with batteries
• Antenna 76 mm (3.0) flexible
• Number of Input Channels
• 8 configurable channels (external sensors) plus
  
• 1 internal position sensor,
• 1 pulse oximeter,
• 1 airflow sensor,
• 1 DC channel Input Range 750µV to 2V
  (configurable)
• Resolution 8, 12, 16 bits, configurable
• Sampling Rate 960 Samples per second per channel
  
• Filter Input bandwidth 0.5 Hz - 250 Hz (-3dB
  attenuation) CMRR 100 dB
• Noise lt 2 µV peak-to-peak (0.5 Hz 100 Hz)
• Input Impedance gt 20 M? _at_ 10 Hz
• Input Interface Standard no-touch 1.5 mm
  connectors
• Power Supply 2 AA alkaline batteries, Battery
  Life 12 hours continuous use

16
Conclusions

• Four-channel telemetry EEG used in the ED is
  feasible, provides good quality screening EEGs
  and was able to diagnose underlying seizure in a
  significant number of patients.

17
ACKNOWLEDGMENT
This work was supported by NIH Phase II SBIR
Grant No. 5 R44 NS042977-03 National Institute of
Neurological Disorders and Stroke US National
Institutes of Health
18
For Questions Contact

• Robert N. Schmidt
• Cleveland Medical Devices Inc.
• 4415 Euclid Ave., Suite 400
• Cleveland, Ohio 44103 USA
• rschmidt_at_CleveMed.com
• Phone 1-877-CleveMed (253-8363) (US Toll Free)
• Phone Direct 01-216-619-5925
• Fax [phone]

19
Smaller Hardware

• Non-Programmable (factory settings)
• 2 channels
• 960 sps
• Up to 12 bits
• Input Selections
• EEG /- 1 mV, 0.1-70 Hz.
• EKG /- 5 mV, 0.1-150 Hz.
• EMG /- 50 mV, 0.1-500 Hz.
• Range, 50 ft.
• Low Noise , lt 1 uV RMS
• Low weight, 191 grams (0.42 oz.)
• Battery options, 14 hrs to days

BioRadio Jr.