LowDose Computed Tomography

1
Low-Dose Computed Tomography

• Jessica de Ryk

2
Computed Tomography (CT)

• CT image formation is based on the reconstruction
  of detected X-ray photons which are passed
  through the body.
• CT images are a reflection of the density
  variations of the tissues with in the body.
• CT produces high quality anatomical detail and is
  most widely used for diagnosis of diseases which
  affect structure.

3
Increasing use of CT

• Helical multi-detector computer tomography (MDCT)
• Decreased scan time
• Combined MDCT with PET or SPECT
• Structural and functional imaging
• Computer-Aided diagnosis (CAD)
• Removes subjectivity
• Increased use for planning and guiding other
  procedures

4
Risks of X-ray radiation

• Deterministic effects
• Deterministic effects involve cell death
• Occur above a particular radiation threshold
• Stochastic effects
• Stochastic effects include radiation-induced
  cancer
• Randomly initiated effects, independent of a
  threshold exposure level
• Severity of effect is independent of dose
• Increased probability of effect occurring with
  increased dosage

5
Determinants of radiation dose

• Milliampere-second (mAs)
• Corresponds to the rate at which electrons leave
  the cathode in the X-ray tube, related to time
• Linear relationship between tube current and
  number of photons produced
• mAs affects density apparent in images
• A decreased mAs leads to increased noise, mottle,
  in the image
• mAs is directly proportional to radiation dose

6
Determinants of radiation dose

• Kilovoltage-peak (kVp)
• Corresponds to the voltage applied across the
  anode and the cathode in the X-ray tube
• Controls the velocity at which the electrons
  collide with the anode.
• Hence, controls the energy of the resulting
  photons
• kVp controls the contrast in the resulting image
• kVp is not linearly proportional to dose
• kVp alters energy of photons and quantity

7
Determinants of radiation dose

• Pitch
• Relates to the speed of table motion per rotation
• Pitch is inversely related to dosage
• Slice thickness
• Thinner slices are more susceptible to noise
• To maintain image quality, mAs, and hence dosage,
  is increased
• Scan time
• Increased scan time decreases the dose
• However, faster scanners lead to temptation to
  increase scan area and hence increase dose

8
Low-dose computed tomography

• A reduced mAs is the most practical means of
  lowering dosage due to direct proportionality
  between mAs and dose
• Issues of focus
• Investigation of optimal low-dose imaging
  protocols which result in a minimal reduction in
  image quality
• The possibility that what would appear to be a
  poor quality image, contains the same degree of
  valuable information and thus the same diagnostic
  power as standard CT images

9
Low-dose CT in chest imaging

• Substantial reductions in mAs have been achieved
  with minimal reductions in image quality for
  chest CT
• This is largely possible due to the inherent
  contrast in the thorax along with the low
  attenuation coefficient of air

290 mAs
40 mAs
10
Dose reduction and patient size

• Kalra et al. found acceptable image quality could
  be achieved at a 50 reduced mAs in
• Chest CT for patients weighing less than 81.6 kg
• Abdominal CT with cross-sectional area less than
  800 cm2
• Abdominal CT with a circumference of less than
  105cm
• Haaga et al. also investigated low dosage CT,
  patient cross-sectional area and image noise in
  chest CT

11
Identification of a threshold mAs

• Zhu et al. recently attempted to determine a
  threshold mAs level for chest CT above which
  image quality was not significantly compromised
• An acceptable level of noise was defined by
  National Metrological Bureau
• 115 mAs was considered conventional and 40, 25,
  15 and 7.5 mAs were regarded
• A milliampere-second value of 25 mAs was
  determined as the acceptable threshold

12
Age and low-dose CT

• Lucaya et al. examined reduction in mAs in chest
  CT of pediatric patients
• Good image quality achieved with
• 34 mAs in cooperative patients
• 50 mAs in non-cooperative patients

13
Lung cancer screening

• There will be approximately 173,770 cases of lung
  cancer in the US in 2004
• For lung cancer the overall 5 year survival rate
  is 13 - 15, similar to survival rate from 30
  years ago
• At an early stage, cure by surgical resection is
  possible in 70 80 of cases
• Currently no accepted screening for lung cancer

14
Screening and Low-dose CT

• Low-dose CT versus chest radiography is currently
  being investigated as a screening method for lung
  cancer
• Early Lung Cancer Action Program (ELCAP)
• National Lung Screening Trial (NLST)
• Assessment is based not only the ability to
  detect cancer at an early stage but also include
  a measure of impact such as curability or
  mortality

15
Enabling technologies

• Tools to aid the development of low-dose CT
  protocols
• Low-dose simulation software
• Hardware developments to further reduce patient
  radiation exposure
• X-ray filters
• Milliampere-second modulation
• Image processing based
• Image filtering
• Computer-aided diagnosis (CAD)

16
Simulating low-dose CT

• Difficulty in developing studies investigating CT
  protocols
• Gaining a significant number of participants
• Morality issues subjecting patients to radiation
  risks
• Limitations due to legal issues and annual
  allowable patient radiation exposure
• Possible alternatives to scanning human patients
  which are also subject to limitations
• Cadavers
• Animals
• Phantoms

17
Simulating low-dose CT

• CT manufacturers and some researchers have
  developed algorithms to simulate low-dose CT
  scanning
• Scanning is performed at standard mAs
• Random noise is added to raw data to simulate a
  lower mAs acquisition
• Accuracy of the simulation varies with developer
• Syngo Explorer (Siemens) can reliably simulate
  low-dose scanning down to 40 mAs, below which the
  image noise is overestimated

18
Hardware developments

• Filtering
• Filtering to remove low energy photons which are
  commonly attenuated in the body and dont
  contribute to image
• Bow-tie or beam-shaping filters reduce dosage by
  minimizing radiation to thinner portions of the
  body
• Milliampere-second modulation
• CT manufacturers are making move to automatically
  adjust mAs based on patient size

19
Software developments

• Noise reduction filters
• Various algorithms have been developed to remove
  noise mottle from a reduction in dose
• Need to minimize the decreased image contrast and
  sharpness which is of great clinical significance
  to detection of small structures such as early
  lung cancer nodules
• Computer-aided diagnosis (CAD)
• Development needed to remove radiologist
  subjectivity in interpretation of low-dose CT
  images

20
Conclusion

• The conversion to low-dose CT protocols will be
  beneficial to patient health and a necessary
  development if the applications of CT imaging
  continue to grow.
• Low-dose CT is highly suitable for lung imaging
  and while preliminary results from the NLST and
  ELCAP studies are promising, the full impact of
  early lung cancer detection on curability and
  mortality must be determined before low-dose CT
  screening will be widely accepted.

21
Thank you

• Questions ?