Design of Health Technologies lecture 12

1
Design of Health Technologieslecture 12
John Canny10/17/05
2
Advanced Sensing Systems

• Biosensors
• Glucose monitoring
• Other systems

3
Magneto-elastic sensors (Grimes)

• The magneto-elastic material resonates at a
  characteristic frequency when excited by a
  magnetic field.

4
Magneto-elastic sensors

• The magneto-elastic ribbon is made of a
  commercial sheet called Metglas.
• The polymer is a custom co-polymer made by the
  Grimes group. It is believed to work because
  glucose bonds to sites on polymer chains that
  separate them from other chains. This allows the
  polymer to absorb water.

5
Magneto-elastic sensors

• Its frequency response (in air) shows a sharp
  peak which is determined by the density of the
  polymer layer

6
Magneto-elastic sensors

• Resonant frequency in a liquid is lower, and the
  peak is not as sharp.

7
Magneto-elastic sensors

• Frequency response in water varies with the
  glucose concentration, in an almost perfectly
  linear curve.

8
Sensor measurement

• The electronics are simple. A sharp spike is
  applied to a driving coil, and a response is
  measured in a sense coil.

9
Sensor measurement

• The magnetic spike is short, about 3 gauss for 16
  micro-seconds (earths magnetic field is about
  0.5 gauss, and a refrigerator magnet about 10
  gauss).
• The pickup coil measures sensor activity for a
  further 8 milli-seconds. The response is
  transformed with an FFT to determine the
  frequency peak.
• This should be easy to do with a small,
  battery-powered device. Because the sensors
  response is quite slow (tens of minutes to
  respond), it is enough to take readings every few
  minutes.

10
Biosensor status

• There are many promising systems on the horizon,
  but the only commercially-deployed biosensors are
  glucose monitors (4B). 3 main types
• Single Use Disposable sensing material, often
  static measurement. Cheap and portable, but low
  sensitivity and accuracy.
• Intermittent Use Often use hydrodynamics
  generally much better performance from sensing a
  moving fluid. Its still a challenge to move these
  out of the lab and onto a chip.

11
Biosensor status

• Continuous (In Vivo) Sensors Very economical,
  but very hard to calibrate and may suffer from
  unknown amount of drift.

12
Biosensor design

• We give a brief introduction to micro-fluidic
  sensor design.
• While these were originally fabricated in silicon
  using MEMS techniques, the trend is toward glass
  and plastic as the substrate.
• Both glass and many plastics allow optical
  measurements, but silicon is opaque to visible
  light.
• Glass and plastic are also more resistant to
  contamination from the chemicals used in the
  measurement.

13
Biosensor design

• Surface immobilization The first step is sensing
  is creating a selective surface to react to the
  sensed agent

14
Biosensor design

• Bead immobilization A variation that uses beads
  to increase relative surface area.

15
Biosensor design

• Detection Several methods, including resonant
  frequency of MEMS cantilevers. But amperometry
  (current measurement) is the most widely used
  approach. Typical mechanisms for current flow
  include redox cycles between the target group and
  variants.

16
Biosensor design

• Optical Detection A 2D array of agent/antigen
  reactions produces fluorescent traces

17
Biosensor design

• Magnetic Detection The antibodies are
  immobilized on a surface and magnetic beads bind
  to sites where the analyte is attached.

18
Enzyme-Linked Immunosorbent Assay
19
Reuse

• Most immunosensors use bound antibodies and
  immobilization. Removing the bound species can be
  difficult without destroying the sensors.
• Methods and results vary, but a recent detector
  for Chagas disease used glycine-HCl to wash the
  sensor, and reported efficacy for more than 30
  cycles.

20
Biosensor design

• Systems-on-a-chip are promising but coming
  slowly. Biosensing still seems a long way from
  commercial viability. But there are some
  promising prototypes

21
Discussion Questions

• It may be a while before we have highly
  integrated sensors for many pathogens (and
  economics dictates that they will come for
  first-world diseases first). Can you think of
  telemedicine/information tools to help facilitate
  traditional (but simple) lab methods?
• Sensors for medical diagnosis may always be a
  difficult economic proposition. Can you think of
  other models that might work? E.g. home testing?