Search tips
Search criteria

Results 1-25 (594985)

Clipboard (0)

Related Articles

1.  Calibration of diffuse correlation spectroscopy with a time-resolved near-infrared technique to yield absolute cerebral blood flow measurements 
Biomedical Optics Express  2011;2(7):2068-2081.
A primary focus of neurointensive care is the prevention of secondary brain injury, mainly caused by ischemia. A noninvasive bedside technique for continuous monitoring of cerebral blood flow (CBF) could improve patient management by detecting ischemia before brain injury occurs. A promising technique for this purpose is diffuse correlation spectroscopy (DCS) since it can continuously monitor relative perfusion changes in deep tissue. In this study, DCS was combined with a time-resolved near-infrared technique (TR-NIR) that can directly measure CBF using indocyanine green as a flow tracer. With this combination, the TR-NIR technique can be used to convert DCS data into absolute CBF measurements. The agreement between the two techniques was assessed by concurrent measurements of CBF changes in piglets. A strong correlation between CBF changes measured by TR-NIR and changes in the scaled diffusion coefficient measured by DCS was observed (R2 = 0.93) with a slope of 1.05 ± 0.06 and an intercept of 6.4 ± 4.3% (mean ± standard error).
PMCID: PMC3130590  PMID: 21750781
(170.1470) Blood or tissue constituent monitoring; (170.3660) Light propagation in tissues; (170.3890) Medical optics instrumentation
2.  Direct measurement of tissue blood flow and metabolism with diffuse optics 
Diffuse optics has proven useful for quantitative assessment of tissue oxy- and deoxyhaemoglobin concentrations and, more recently, for measurement of microvascular blood flow. In this paper, we focus on the flow monitoring technique: diffuse correlation spectroscopy (DCS). Representative clinical and pre-clinical studies from our laboratory illustrate the potential of DCS. Validation of DCS blood flow indices in human brain and muscle is presented. Comparison of DCS with arterial spin-labelled MRI, xenon-CT and Doppler ultrasound shows good agreement (0.50
PMCID: PMC3263785  PMID: 22006897
diffuse correlation spectroscopy; blood flow; cerebral blood flow; oxygen metabolism; brain; cancer
Delayed-cerebral ischemia is a major cause of morbidity and mortality in the setting of aneurysmal subarachnoid hemorrhage. Despite extensive research efforts and a breadth of collective clinical experience, accurate diagnosis of vasospasm remains difficult, and effective treatment options are limited. Classically, diagnosis has focused on imaging assessment of the cerebral vasculature. Recently, invasive and noninvasive bedside techniques designed to characterize relevant hemodynamic and metabolic alterations have gained substantial attention. Such modalities include microdialysis, brain tissue oxygenation, jugular bulb oximetry, thermal diffusion cerebral blood flow, and near-infrared spectroscopy. This paper reviews these modalities and examines data pertinent to the diagnosis and management of cerebral vasospasm.
PMCID: PMC3628660  PMID: 23606963
Physiological Reports  2013;1(7):e00179.
The potential interference of cutaneous circulation on muscle blood volume and oxygenation monitoring by near‐infrared spectroscopy (NIRS) remains an important limitation of this technique. Spatially resolved spectroscopy (SRS) was reported to minimize the contribution of superficial tissue layers in cerebral monitoring but this characteristic has never been documented in muscle tissue monitoring. This study aims to compare SRS with the standard Beer–Lambert (BL) technique in detecting blood volume changes selectively induced in muscle and skin. In 16 healthy subjects, the biceps brachii was investigated during isometric elbow flexion at 70% of the maximum voluntary contractions lasting 10 sec, performed before and after exposure of the upper arm to warm air flow. From probes applied over the muscle belly the following variables were recorded: total hemoglobin index (THI, SRS‐based), total hemoglobin concentration (tHb, BL‐based), tissue oxygenation index (TOI, SRS‐based), and skin blood flow (SBF), using laser Doppler flowmetry. Blood volume indices exhibited similar changes during muscle contraction but only tHb significantly increased during warming (+5.2 ± 0.7 μmol/L·cm, an effect comparable to the increase occurring in postcontraction hyperemia), accompanying a 10‐fold increase in SBF. Contraction‐induced changes in tHb and THI were not substantially affected by warming, although the tHb tracing was shifted upward by (5.2 ± 3.5 μmol/L·cm, P < 0.01). TOI was not affected by cutaneous warming. In conclusion, SRS appears to effectively reject interference by SBF in both muscle blood volume and oxygenation monitoring. Instead, BL‐based parameters should be interpreted with caution, whenever changes in cutaneous perfusion cannot be excluded.
The influence of cutaneous circulation on muscle NIRS monitoring has been seldom investigated and is often overlooked. This study shows that cutaneous dilatation induced by superficial warming produces consistent changes in standard Beer–Lambert (BL) parameters while leaving spatially resolved parameters unaffected. Thus, Beer–Lambert parameters should be interpreted with caution whenever changes in cutaneous perfusion are expected to occur.
PMCID: PMC3970749  PMID: 24744858
Muscle blood flow; muscle blood volume; muscle oxygenation; near‐infrared spectroscopy; skin blood flow
PLoS Computational Biology  2008;4(11):e1000212.
We construct a model of brain circulation and energy metabolism. The model is designed to explain experimental data and predict the response of the circulation and metabolism to a variety of stimuli, in particular, changes in arterial blood pressure, CO2 levels, O2 levels, and functional activation. Significant model outputs are predictions about blood flow, metabolic rate, and quantities measurable noninvasively using near-infrared spectroscopy (NIRS), including cerebral blood volume and oxygenation and the redox state of the CuA centre in cytochrome c oxidase. These quantities are now frequently measured in clinical settings; however the relationship between the measurements and the underlying physiological events is in general complex. We anticipate that the model will play an important role in helping to understand the NIRS signals, in particular, the cytochrome signal, which has been hard to interpret. A range of model simulations are presented, and model outputs are compared to published data obtained from both in vivo and in vitro settings. The comparisons are encouraging, showing that the model is able to reproduce observed behaviour in response to various stimuli.
Author Summary
Monitoring the brain noninvasively is key to solving various biological and clinical problems. Near-infrared spectroscopy (NIRS) is a technique that can measure changes in the colour of the brain. The brain has an absolute requirement for oxygen; the spectroscopically observed colour changes are due to the proteins that deliver (haemoglobin) and consume (mitochondrial cytochrome c oxidase) oxygen. Haemoglobin changes colour when it binds oxygen. The changes in cytochrome c oxidase are due to the electron occupancy (reduction) of a particular copper metal centre in the enzyme. The way that the state of this enzyme changes in various situations is poorly understood. Currently there is no theoretical model that can be used to decode simultaneously all of the spectroscopic changes in these proteins, and thus limited information about the underlying biochemistry and physiology can be extracted from the NIRS signals. We therefore constructed such a model, ensuring that it is consistent with the scientific literature, in vivo data, and the underlying thermodynamic principles. The model was able to predict the physiological and spectroscopic responses to a wide range of stimuli, including changes in brain activity and oxygen delivery. It is likely to be of significant value to a wide range of clinical and life science users.
PMCID: PMC2573000  PMID: 18989392
Little is known about cerebral blood flow, cerebral blood volume (CBV), oxygenation, and oxygen consumption in the premature newborn brain. We combined quantitative frequency-domain near-infrared spectroscopy measures of cerebral hemoglobin oxygenation (SO2) and CBV with diffusion correlation spectroscopy measures of cerebral blood flow index (BFix) to determine the relationship between these measures, gestational age at birth (GA), and chronological age. We followed 56 neonates of various GA once a week during their hospital stay. We provide absolute values of SO2 and CBV, relative values of BFix, and relative cerebral metabolic rate of oxygen (rCMRO2) as a function of postmenstrual age (PMA) and chronological age for four GA groups. SO2 correlates with chronological age (r=−0.54, P value ⩽0.001) but not with PMA (r=−0.07), whereas BFix and rCMRO2 correlate better with PMA (r=0.37 and 0.43, respectively, P value ⩽0.001). Relative CMRO2 during the first month of life is lower when GA is lower. Blood flow index and rCMRO2 are more accurate biomarkers of the brain development than SO2 in the premature newborns.
PMCID: PMC3293111  PMID: 22027937
brain hemodynamic development; cerebral oxygen consumption; diffuse correlation spectroscopy; frequency-domain near-infrared spectroscopy; premature neonates
Topics in stroke rehabilitation  2008;15(5):427-450.
Brain-mapping techniques have proven to be vital in understanding the molecular, cellular, and functional mechanisms of recovery after stroke. This article briefly summarizes the current molecular and functional concepts of stroke recovery and addresses how various neuroimaging techniques can be used to observe these changes. The authors provide an overview of various techniques including diffusion-tensor imaging (DTI), magnetic resonance spectroscopy (MRS), ligand-based positron emission tomography (PET), single-photon emission computed tomography (SPECT), regional cerebral blood flow (rCBF) and regional metabolic rate of glucose (rCMRglc) PET and SPECT, functional magnetic resonance imaging (fMRI), near infrared spectroscopy (NIRS), electroencephalography (EEG), magnetoencephalography (MEG), and transcranial magnetic stimulation (TMS). Discussion in the context of poststroke recovery research informs about the applications and limitations of the techniques in the area of rehabilitation research. The authors also provide suggestions on using these techniques in tandem to more thoroughly address the outstanding questions in the field.
PMCID: PMC2663338  PMID: 19008203
DTI; EEG; fMRI; MEG; NIRS; PET; poststroke recovery; rehabilitation; SPECT; stroke; TMS
Biomedical Optics Express  2010;1(5):1443-1459.
Acetazolamide (ACZ) was used to stimulate the cerebral vasculature on ten healthy volunteers to assess the cerebral vasomotor reactivity (CVR). We have combined near infrared spectroscopy (NIRS), diffuse correlation spectroscopy (DCS) and transcranial Doppler (TCD) technologies to non-invasively assess CVR in real-time by measuring oxy- and deoxy-hemoglobin concentrations, using NIRS, local cerebral blood flow (CBF), using DCS, and blood flow velocity (CBFV) in the middle cerebral artery, using TCD. Robust and persistent increases in oxy-hemoglobin concentration, CBF and CBFV were observed. A significant agreement was found between macro-vascular (TCD) and micro-vascular (DCS) hemodynamics, between the NIRS and TCD data, and also within NIRS and DCS results. The relative cerebral metabolic rate of oxygen, rCMRO2, was also determined, and no significant change was observed. Our results showed that the combined diffuse optics-ultrasound technique is viable to follow (CVR) and rCMRO2 changes in adults, continuously, at the bed-side and in real time.
PMCID: PMC3018112  PMID: 21258561
(170.3660) Light propagation in tissues; (170.3890) Medical optics instrumentation; (170.6480) Spectroscopy, speckle; (170.7170) Ultrasound; (290.4210) Multiple scattering
Near-infrared spectroscopy has primarily been used in monitoring changes in cerebral haemoglobin oxygenation and haemodynamics. However its use as a method for the assessment of tissue viability following transplantation has recently been explored experimentally in our laboratory. The ability to measure changes in oxygenation and perfusion during harvesting and following transplantation of organs or transfer of free and pedicled flaps potentially important in reconstructive surgery. We have found that near-infrared spectroscopy is extremely useful in detecting vaso-occlusive events and can accurately and reliably distinguish between arterial, venous or total occlusions. Venous congestion indicated by raised levels of deoxygenated haemoglobin with a concomitant increase in blood volume and the presence and magnitude of reactive hyperaemia are both easily recognizable features by near-infrared spectroscopy. We have shown that near-infrared spectroscopy measurements of venous congestion in kidneys (and other tissues) following prolonged storage correlate with medullary vascular congestion confirmed by angiographical and histological analysis of intrarenal perfusion. Clinically we have shown that flap perfusion can be improved by altering fluid replacement regimes and the addition of ionotropes. Cerebral near-infrared spectroscopy measurements in a liver transplant model showed statistically significant differences within minutes after the anhepatic phase in cerebral perfusion and oxygenation, between animals transplanted with ischaemically damaged livers compared to those isografted with minimally stored livers. Similarly we have found that near-infrared spectroscopy can be used as a monitor to assess the adequacy of fluid or blood replacement in haemorrhagic and hypovolaemic models. We believe that near-infrared spectroscopy provides a sensitive and reliable postoperative method for the assessment of tissue viability following the transfer of free and pedicled flaps and organs.
PMCID: PMC1691953  PMID: 9232856
Arthritis Research & Therapy  2012;14(6):R236.
Women with fibromyalgia (FM) have symptoms of increased muscular fatigue and reduced exercise tolerance, which may be associated with alterations in muscle microcirculation and oxygen metabolism. This study used near-infrared diffuse optical spectroscopies to noninvasively evaluate muscle blood flow, blood oxygenation and oxygen metabolism during leg fatiguing exercise and during arm arterial cuff occlusion in post-menopausal women with and without FM.
Fourteen women with FM and twenty-three well-matched healthy controls participated in this study. For the fatiguing exercise protocol, the subject was instructed to perform 6 sets of 12 isometric contractions of knee extensor muscles with intensity steadily increasing from 20 to 70% maximal voluntary isometric contraction (MVIC). For the cuff occlusion protocol, forearm arterial blood flow was occluded via a tourniquet on the upper arm for 3 minutes. Leg or arm muscle hemodynamics, including relative blood flow (rBF), oxy- and deoxy-hemoglobin concentration ([HbO2] and [Hb]), total hemoglobin concentration (THC) and blood oxygen saturation (StO2), were continuously monitored throughout protocols using a custom-built hybrid diffuse optical instrument that combined a commercial near-infrared oximeter for tissue oxygenation measurements and a custom-designed diffuse correlation spectroscopy (DCS) flowmeter for tissue blood flow measurements. Relative oxygen extraction fraction (rOEF) and oxygen consumption rate (rVO2) were calculated from the measured blood flow and oxygenation data. Post-manipulation (fatiguing exercise or cuff occlusion) recovery in muscle hemodynamics was characterized by the recovery half-time, a time interval from the end of manipulation to the time that tissue hemodynamics reached a half-maximal value.
Subjects with FM had similar hemodynamic and metabolic response/recovery patterns as healthy controls during exercise and during arterial occlusion. However, tissue rOEF during exercise in subjects with FM was significantly lower than in healthy controls, and the half-times of oxygenation recovery (Δ[HbO2] and Δ[Hb]) were significantly longer following fatiguing exercise and cuff occlusion.
Our results suggest an alteration of muscle oxygen utilization in the FM population. This study demonstrates the potential of using combined diffuse optical spectroscopies (i.e., NIRS/DCS) to comprehensively evaluate tissue oxygen and flow kinetics in skeletal muscle.
PMCID: PMC3674608  PMID: 23116302
Seminars in perinatology  2010;34(1):46-56.
Advances in medical and surgical care of the high-risk neonate have led to increased survival. A significant number of these neonates suffer from neurodevelopmental delays and failure in school. The focus of clinical research has shifted to understanding events contributing to neurological morbidity in these patients. Assessing changes in cerebral oxygenation and regulation of cerebral blood flow (CBF) is important in evaluating the status of the central nervous system. Traditional CBF imaging methods fail for both ethical and logistical reasons. Optical near infrared spectroscopy (NIRS) is increasingly being used for bedside monitoring of cerebral oxygenation and blood volume in both very low birth weight infants and neonates with congenital heart disease. Although trends in CBF may be inferred from changes in cerebral oxygenation and/or blood volume, NIRS does not allow a direct measure of CBF in these populations. Two relatively new modalities, arterial spin-labeled perfusion magnetic resonance imaging and optical diffuse correlation spectroscopy, provide direct, noninvasive measures of cerebral perfusion suitable for the high-risk neonates. Herein we discuss the instrumentation, applications, and limitations of these noninvasive imaging techniques for measuring and/or monitoring CBF.
PMCID: PMC2829712  PMID: 20109972
infant cerebral blood flow; CBF; arterial spin labeled perfusion; MRI; PVL; optical spectroscopy
Anesthesia and analgesia  2012;116(3):10.1213/ANE.0b013e318277a255.
Near-infrared spectroscopy is used during cardiac surgery to monitor the adequacy of cerebral perfusion. In this systematic review, we evaluated available data for adult patients to determine (1) whether decrements in cerebral oximetry during cardiac surgery are associated with stroke, postoperative cognitive dysfunction (POCD), or delirium and (2) whether interventions aimed at correcting cerebral oximetry decrements improve neurologic outcomes.
We searched PubMed, Cochrane, and Embase databases from inception until January 31, 2012, without restriction on languages. Each article was examined for additional references. A publication was excluded if it did not include original data (e.g., review, commentary) or if it was not published as a full-length article in a peer-reviewed journal (e.g., abstract only). The identified abstracts were screened first, and full texts of eligible papers were reviewed independently by two investigators. For eligible publications, we recorded the number of subjects, type of surgery, and criteria for diagnosis of neurologic endpoints.
We identified 13 case reports, 27 observational studies, and two prospectively randomized intervention trials that met our inclusion criteria. Case reports and two observational studies contained anecdotal evidence suggesting that regional cerebral O2 saturation (rScO2) monitoring could be used to identify cardiopulmonary bypass (CPB) cannula malposition. Six of nine observational studies reported an association between acute rScO2 desaturation and POCD based on the Mini-Mental Status Examination (n=3 studies) or more detailed cognitive testing (n=6 studies). Two retrospective studies reported a relationship between rScO2 desaturation and stroke or type I and II neurologic injury after surgery. The observational studies had many limitations, including small sample size, assessments only during the immediate postoperative period, and failure to perform risk adjustments. Two randomized studies evaluated the efficacy of interventions for treating rScO2 desaturation during surgery, but adherence to the protocol was poor in one. In the other study, interventions for rScO2 desaturation were associated with less major organ injury and shorter intensive care unit hospitalization compared with nonintervention.
Reductions in rScO2 during cardiac surgery may identify CPB cannula malposition, particularly during aortic surgery. Only low-level evidence links low rScO2 during cardiac surgery to postoperative neurologic complications, and data are insufficient to conclude that interventions to improve rScO2 desaturation prevent stroke or POCD.
PMCID: PMC3863709  PMID: 23267000
Astrobiology  2011;11(5):393-408.
The EPOXI Discovery Mission of Opportunity reused the Deep Impact flyby spacecraft to obtain spatially and temporally resolved visible photometric and moderate resolution near-infrared (NIR) spectroscopic observations of Earth. These remote observations provide a rigorous validation of whole-disk Earth model simulations used to better understand remotely detectable extrasolar planet characteristics. We have used these data to upgrade, correct, and validate the NASA Astrobiology Institute's Virtual Planetary Laboratory three-dimensional line-by-line, multiple-scattering spectral Earth model. This comprehensive model now includes specular reflectance from the ocean and explicitly includes atmospheric effects such as Rayleigh scattering, gas absorption, and temperature structure. We have used this model to generate spatially and temporally resolved synthetic spectra and images of Earth for the dates of EPOXI observation. Model parameters were varied to yield an optimum fit to the data. We found that a minimum spatial resolution of ∼100 pixels on the visible disk, and four categories of water clouds, which were defined by using observed cloud positions and optical thicknesses, were needed to yield acceptable fits. The validated model provides a simultaneous fit to Earth's lightcurve, absolute brightness, and spectral data, with a root-mean-square (RMS) error of typically less than 3% for the multiwavelength lightcurves and residuals of ∼10% for the absolute brightness throughout the visible and NIR spectral range. We have extended our validation into the mid-infrared by comparing the model to high spectral resolution observations of Earth from the Atmospheric Infrared Sounder, obtaining a fit with residuals of ∼7% and brightness temperature errors of less than 1 K in the atmospheric window. For the purpose of understanding the observable characteristics of the distant Earth at arbitrary viewing geometry and observing cadence, our validated forward model can be used to simulate Earth's time-dependent brightness and spectral properties for wavelengths from the far ultraviolet to the far infrared. Key Words: Astrobiology—Extrasolar terrestrial planets—Habitability—Planetary science—Radiative transfer. Astrobiology 11, 393–408.
PMCID: PMC3133830  PMID: 21631250
Critical care medicine  2011;39(10):2337-2345.
Knowledge remains limited regarding cerebral blood flow autoregulation after cardiac arrest and during post-resuscitation hypothermia. We determined the relationship of cerebral blood flow to cerebral perfusion pressure in a swine model of pediatric hypoxic-asphyxic cardiac arrest during normothermia and hypothermia and tested novel measures of autoregulation derived from near-infrared spectroscopy.
Prospective, balanced animal study.
Basic physiology laboratory at an academic institution.
Eighty-four neonatal swine.
Piglets underwent hypoxic-asphyhxic cardiac arrest or sham surgery and recovered for 2 hours with normothermia followed by 4 hours of either moderate hypothermia or normothermia. In half of the groups, blood pressure was slowly decreased through inflation of a balloon catheter in the inferior vena cava to identify the lower limit of cerebral autoregulation at 6 hours post-resuscitation. In the remaining groups, blood pressure was gradually increased by inflation of a balloon catheter in the aorta to determine the autoregulatory response to hypertension. Measures of autoregulation obtained from standard laser-Doppler flowmetry and indices derived from near-infrared spectroscopy were compared.
Measurements and Main Results
Laser-Doppler flux was lower in post-arrest animals compared to sham-operated controls during the 2-hour normothermic period after resuscitation. During the subsequent 4-hour recovery, hypothermia decreased laser-Doppler flux in both the sham surgery and post-arrest groups. Autoregulation was intact during hypertension in all groups. With arterial hypotension, post-arrest, hypothermic piglets had a significant decrease in the perfusion pressure lower limit of autoregulation compared to post-arrest, normothermic piglets. The near-infrared spectroscopy-derived measures of autoregulation accurately detected loss of autoregulation during hypotension.
In a pediatric model of cardiac arrest and resuscitation, delayed induction of hypothermia decreased cerebral perfusion and decreased the lower limit of autoregulation. Metrics derived from non-invasive near-infrared spectroscopy accurately identified the lower limit of autoregulation during normothermia and hypothermia in piglets resuscitated from arrest.
PMCID: PMC3178742  PMID: 21705904
pediatrics; heart arrest; cerebrovascular circulation; ischemia; blood pressure; hypothermia
AIMS—To investigate the relation between cerebral blood flow on the first day of postnatal life and the severity of any subsequent germinal matrix haemorrhage-intraventricular haemorrhage (GMH-IVH).
METHODS—Cerebral blood flow was measured in 24 babies during the first 24 hours of life using near infrared spectroscopy. Repeated cerebral ultrasound examination was performed to define the maximum extent of GMH-IVH. Infants were classified as: normal scan, minor periventricular haemorrhage (haemorrhage that resolved), or severe GMH-IVH (haemorrhage distending the ventricles, that progressed to either post haemorrhagic dilatation or porencephalic cyst formation).
RESULTS—Cerebral blood flow was significantly lower in the infants with GMH-IVH (median 7.0 ml/100 g/min) than those without haemorrhage (median 12.2 ml/100 g/min), despite no difference in carbon dioxide tension and a higher mean arterial blood pressure. On subgroup analysis, those infants with severe GMH-IVH had the lowest cerebral blood flow.
CONCLUSION—A low cerebral blood flow on the first day of life is associated with the subsequent development of severe intraventricular haemorrhage.

PMCID: PMC1720962  PMID: 10375356
Biomedical Optics Express  2010;1(2):553-565.
Cerebral blood flow (CBF) during stepped hypercapnia was measured simultaneously in the rat brain using near-infrared diffuse correlation spectroscopy (DCS) and arterial spin labeling MRI (ASL). DCS and ASL CBF values agree very well, with high correlation (R=0.86, p< 10-9), even when physiological instability perturbed the vascular response. A partial volume effect was evident in the smaller magnitude of the optical CBF response compared to the MRI values (averaged over the cortical area), primarily due to the inclusion of white matter in the optically sampled volume. The 8.2 and 11.7 mm mid-separation channels of the multi-distance optical probe had the lowest partial volume impact, reflecting ~75 % of the MR signal change. Using a multiplicative correction factor, the ASL CBF could be predicted with no more than 10% relative error, affording an opportunity for real-time relative cerebral metabolism monitoring in conjunction with MR measurement of cerebral blood volume using super paramagnetic contrast agents.
PMCID: PMC3017992  PMID: 21258489
(170.2655) Functional monitoring and imaging; (170.1470) Blood or tissue constituent monitoring; (170.0170) Medical optics and biotechnology; (170.3340) Laser Doppler velocimetry
Executive Summary
For cases of acute respiratory distress syndrome (ARDS) and progressive chronic respiratory failure, the first choice or treatment is mechanical ventilation. For decades, this method has been used to support critically ill patients in respiratory failure. Despite its life-saving potential, however, several experimental and clinical studies have suggested that ventilator-induced lung injury can adversely affect the lungs and patient outcomes. Current opinion is that by reducing the pressure and volume of gas delivered to the lungs during mechanical ventilation, the stress applied to the lungs is eased, enabling them to rest and recover. In addition, mechanical ventilation may fail to provide adequate gas exchange, thus patients may suffer from severe hypoxia and hypercapnea. For these reasons, extracorporeal lung support technologies may play an important role in the clinical management of patients with lung failure, allowing not only the transfer of oxygen and carbon dioxide (CO2) but also buying the lungs the time needed to rest and heal.
The objective of this analysis was to assess the effectiveness, safety, and cost-effectiveness of extracorporeal lung support technologies in the improvement of pulmonary gas exchange and the survival of adult patients with acute pulmonary failure and those with end-stage chronic progressive lung disease as a bridge to lung transplantation (LTx). The application of these technologies in primary graft dysfunction (PGD) after LTx is beyond the scope of this review and is not discussed.
Clinical Applications of Extracorporeal Lung Support
Extracorporeal lung support technologies [i.e., Interventional Lung Assist (ILA) and extracorporeal membrane oxygenation (ECMO)] have been advocated for use in the treatment of patients with respiratory failure. These techniques do not treat the underlying lung condition; rather, they improve gas exchange while enabling the implantation of a protective ventilation strategy to prevent further damage to the lung tissues imposed by the ventilator. As such, extracorporeal lung support technologies have been used in three major lung failure case types:
As a bridge to recovery in acute lung failure – for patients with injured or diseased lungs to give their lungs time to heal and regain normal physiologic function.
As a bridge to LTx – for patients with irreversible end stage lung disease requiring LTx.
As a bridge to recovery after LTx – used as lung support for patients with PGD or severe hypoxemia.
Ex-Vivo Lung Perfusion and Assessment
Recently, the evaluation and reconditioning of donor lungs ex-vivo has been introduced into clinical practice as a method of improving the rate of donor lung utilization. Generally, about 15% to 20% of donor lungs are suitable for LTx, but these figures may increase with the use of ex-vivo lung perfusion. The ex-vivo evaluation and reconditioning of donor lungs is currently performed at the Toronto General Hospital (TGH) and preliminary results have been encouraging (Personal communication, clinical expert, December 17, 2009). If its effectiveness is confirmed, the use of the technique could lead to further expansion of donor organ pools and improvements in post-LTx outcomes.
Extracorporeal Lung support Technologies
The ECMO system consists of a centrifugal pump, a membrane oxygenator, inlet and outlet cannulas, and tubing. The exchange of oxygen and CO2 then takes place in the oxygenator, which delivers the reoxygenated blood back into one of the patient’s veins or arteries. Additional ports may be added for haemodialysis or ultrafiltration.
Two different techniques may be used to introduce ECMO: venoarterial and venovenous. In the venoarterial technique, cannulation is through either the femoral artery and the femoral vein, or through the carotid artery and the internal jugular vein. In the venovenous technique cannulation is through both femoral veins or a femoral vein and internal jugular vein; one cannula acts as inflow or arterial line, and the other as an outflow or venous line. Venovenous ECMO will not provide adequate support if a patient has pulmonary hypertension or right heart failure. Problems associated with cannulation during the procedure include bleeding around the cannulation site and limb ischemia distal to the cannulation site.
Interventional Lung Assist (ILA) is used to remove excess CO2 from the blood of patients in respiratory failure. The system is characterized by a novel, low-resistance gas exchange device with a diffusion membrane composed of polymethylpentene (PMP) fibres. These fibres are woven into a complex configuration that maximizes the exchange of oxygen and CO2 by simple diffusion. The system is also designed to operate without the help of an external pump, though one can be added if higher blood flow is required. The device is then applied across an arteriovenous shunt between the femoral artery and femoral vein. Depending on the size of the arterial cannula used and the mean systemic arterial pressure, a blood flow of up to 2.5 L/min can be achieved (up to 5.5 L/min with an external pump). The cannulation is performed after intravenous administration of heparin.
Recently, the first commercially available extracorporeal membrane ventilator (NovaLung GmbH, Hechingen, Germany) was approved for clinical use by Health Canada for patients in respiratory failure. The system has been used in more than 2,000 patients with various indications in Europe, and was used for the first time in North America at the Toronto General Hospital in 2006.
Evidence-Based Analysis
The research questions addressed in this report are:
Does ILA/ECMO facilitate gas exchange in the lungs of patients with severe respiratory failure?
Does ILA/ECMO improve the survival rate of patients with respiratory failure caused by a range of underlying conditions including patients awaiting LTx?
What are the possible serious adverse events associated with ILA/ECMO therapy?
To address these questions, a systematic literature search was performed on September 28, 2009 using OVID MEDLINE, MEDLINE In-Process and Other Non-Indexed Citations, EMBASE, the Cumulative Index to Nursing & Allied Health Literature (CINAHL), the Cochrane Library, and the International Agency for Health Technology Assessment (INAHTA) for studies published from January 1, 2005 to September 28, 2008. Abstracts were reviewed by a single reviewer and, for those studies meeting the eligibility criteria, full-text articles were obtained. Reference lists were also examined for any additional relevant studies not identified through the search. Articles with an unknown eligibility were reviewed with a second clinical epidemiologist and then a group of epidemiologists until consensus was established.
Inclusion Criteria
Studies in which ILA/ECMO was used as a bridge to recovery or bridge to LTx
Studies containing information relevant to the effectiveness and safety of the procedure
Studies including at least five patients
Exclusion Criteria
Studies reporting the use of ILA/ECMO for inter-hospital transfers of critically ill patients
Studies reporting the use of ILA/ECMO in patients during or after LTx
Animal or laboratory studies
Case reports
Outcomes of Interest
Reduction in partial pressure of CO2
Correction of respiratory acidosis
Improvement in partial pressure of oxygen
Improvement in patient survival
Frequency and severity of adverse events
The search yielded 107 citations in Medline and 107 citations in EMBASE. After reviewing the information provided in the titles and abstracts, eight citations were found to meet the study inclusion criteria. One study was then excluded because of an overlap in the study population with a previous study. Reference checking did not produce any additional studies for inclusion. Seven case series studies, all conducted in Germany, were thus included in this review (see Table 1).
Also included is the recently published CESAR trial, a multicentre RCT in the UK in which ECMO was compared with conventional intensive care management. The results of the CESAR trial were published when this review was initiated. In the absence of any other recent RCT on ECMO, the results of this trial were considered for this assessment and no further searches were conducted. A literature search was then conducted for application of ECMO as bridge to LTx patients (January, 1, 2005 to current). A total of 127 citations on this topic were identified and reviewed but none were found to have examined the use of ECMO as bridge to LTx.
Quality of Evidence
To grade the quality of evidence, the grading system formulated by the GRADE working group and adopted by MAS was applied. The GRADE system classifies the quality of a body of evidence as high, moderate, low, or very low according to four key elements: study design, study quality, consistency across studies, and directness.
Trials on ILA
Of the seven studies identified, six involved patients with ARDS caused by a range of underlying conditions; the seventh included only patients awaiting LTx. All studies reported the rate of gas exchange and respiratory mechanics before ILA and for up to 7 days of ILA therapy. Four studies reported the means and standard deviations of blood gas transfer and arterial blood pH, which were used for meta-analysis.
Fischer et al. reported their first experience on the use of ILA as a bridge to LTx. In their study, 12 patients at high urgency status for LTx, who also had severe ventilation refractory hypercapnea and respiratory acidosis, were connected to ILA prior to LTx. Seven patients had a systemic infection or sepsis prior to ILA insertion. Six hours after initiation of ILA, the partial pressure of CO2 in arterial blood significantly decreased (P < .05) and arterial blood pH significantly improved (P < .05) and remained stable for one week (last time point reported). The partial pressure of oxygen in arterial blood improved from 71 mmHg to 83 mmHg 6 hours after insertion of ILA. The ratio of PaO2/FiO2 improved from 135 at baseline to 168 at 24 hours after insertion of ILA but returned to baseline values in the following week.
Trials on ECMO
The UK-based CESAR trial was conducted to assess the effectiveness and cost of ECMO therapy for severe, acute respiratory failure. The trial protocol were published in 2006 and details of the methods used for the economic evaluation were published in 2008. The study itself was a pragmatic trial (similar to a UK trial of neonatal ECMO), in which best standard practice was compared with an ECMO protocol. The trial involved 180 patients with acute but potentially reversible respiratory failure, with each also having a Murray score of ≥ 3.0 or uncompensated hypercapnea at a pH of < 7.2. Enrolled patients were randomized in a 1:1 ratio to receive either conventional ventilation treatment or ECMO while on ventilator. Conventional management included intermittent positive pressure ventilation, high frequency oscillatory ventilation, or both. As a pragmatic trial, a specific management protocol was not followed; rather the treatment centres were advised to follow a low volume low pressure ventilation strategy. A tidal volume of 4 to 8 mL/kg body weight and a plateau pressure of < 30 cm H2O were recommended.
Bridge to recovery
No RCTs or observational studies compared ILA to other treatment modalities.
Case series have shown that ILA therapy results in significant CO2 removal from arterial blood and correction of respiratory acidosis, as well as an improvement in oxygen transfer.
ILA therapy enabled a lowering of respiratory settings to protect the lungs without causing a negative impact on arterial blood CO2 and arterial blood pH.
The impact of ILA on patient long-term survival cannot be determined through the studies reviewed.
In-hospital mortality across studies ranged from 20% to 65%.
Ischemic complications were the most frequent adverse events following ILA therapy.
Leg amputation is a rare but possible outcome of ILA therapy, having occurred in about 0.9% of patients in these case series. New techniques involving the insertion of additional cannula into the femoral artery to perfuse the leg may lower this rate.
Bridge to LTx
The results of one case series (n=12) showed that ILA effectively removes CO2 from arterial blood and corrects respiratory acidosis in patients with ventilation refractory hypercapnea awaiting a LTx
Eight of the 12 patients (67%) awaiting a LTx were successfully transplanted and one-year survival for those transplanted was 80%
Since all studies are case series, the grade of the evidence for these observations is classified as “LOW”.
Bridge to recovery
Based on the results of a pragmatic trial and an intention to treat analysis, referral of patient to an ECMO based centre significantly improves patient survival without disability compared to conventional ventilation. The results of CESAR trial showed that:
For patients with information about disability, survival without severe disability was significantly higher in ECMO arm
Assuming that the three patients in the conventional ventilation arm who did not have information about severe disability were all disabled, the results were also significant.
Assuming that none of these patients were disabled, the results were at borderline significance
A greater, though not statistically significant, proportion of patients in ECMO arm survived.
The rate of serious adverse events was higher among patients in ECMO group
The grade of evidence for the above observations is classified as “HIGH”.
Bridge to LTx
No studies fitting the inclusion criteria were identified.
There is no accurate data on the use of ECMO in patients awaiting LTx.
Economic Analysis
The objective of the economic analysis was to determine the costs associated with extracorporeal lung support technologies for bridge to LTx in adults. A literature search was conducted for which the target population was adults eligible for extracorporeal lung support. The primary analytic perspective was that of the Ministry of Health and Long-Term Care (MOHLTC). Articles published in English and fitting the following inclusion criteria were reviewed:
Full economic evaluations including cost-effectiveness analyses (CEA), cost-utility analyses (CUA), cost-benefit analyses (CBA);
Economic evaluations reporting incremental cost-effectiveness ratios (ICER) i.e. cost per quality adjusted life year (QALY), life years gained (LYG), or cost per event avoided; and
Studies in patients eligible for lung support technologies for to lung transplantation.
The search yielded no articles reporting comparative economic analyses.
Resource Use and Costs
Costs associated with both ILA and ECMO (outlined in Table ES-1) were obtained from the University Health Network (UHN) case costing initiative (personal communication, UHN, January 2010). Consultation with a clinical expert in the field was also conducted to verify resource utilization. The consultant was situated at the UHN in Toronto. The UHN has one ECMO machine, which cost approximately $100,000. The system is 18 years old and is used an average of 3 to 4 times a year with 35 procedures being performed over the last 9 years. The disposable cost per patient associated with ECMO is, on average, $2,200. There is a maintenance cost associated with the machine (not reported by the UHN), which is currently absorbed by the hospital’s biomedical engineering department.
The average capital cost of an ILA device is $7,100 per device, per patient, while the average cost of the reusable pump $65,000. The UHN has performed 16 of these procedures over the last 2.5 years. Similarly, there is a maintenance cost not that was reported by UHN but is absorbed by the hospital’s biomedical engineering department.
Resources Associated with Extracorporeal Lung Support Technologies
Hospital costs associated with ILA were based on the average cost incurred by the hospital for 11 cases performed in the FY 07/08 (personal communication, UHN, January 2010). The resources incurred with this hospital procedure included:
Device and disposables
OR transplant
Surgical ICU
Laboratory work
Medical imaging
Clinical nutrition
Occupational therapy
Speech and language pathology
Social work
The average length of stay in hospital was 61 days for ILA (range: 5 to 164 days) and the average direct cost was $186,000 per case (range: $19,000 to $552,000). This procedure has a high staffing requirement to monitor patients in hospital, driving up the average cost per case.
PMCID: PMC3415698  PMID: 23074408
Medical Image Analysis  2013;17(8):1025-1036.
Graphical abstract
•We propose a method for dynamic angiographic blood volume flow rate quantification.•This technique is compatible with most types of dynamic angiographic data.•We applied this method to vessel-encoded arterial spin labeling angiography.•Blood volume flow rate estimates were validated using a flow phantom.•Results in the cerebral arteries of healthy volunteers agree with previous studies.
Angiographic methods can provide valuable information on vessel morphology and hemodynamics, but are often qualitative in nature, somewhat limiting their ability for comparison across arteries and subjects. In this work we present a method for quantifying absolute blood volume flow rates within large vessels using dynamic angiographic data. First, a kinetic model incorporating relative blood volume, bolus dispersion and signal attenuation is fitted to the data. A self-calibration method is also described for both 2D and 3D data sets to convert the relative blood volume parameter into absolute units. The parameter values are then used to simulate the signal arising from a very short bolus, in the absence of signal attenuation, which can be readily encompassed within a vessel mask of interest. The volume flow rate can then be determined by calculating the resultant blood volume within the vessel mask divided by the simulated bolus duration. This method is applied to non-contrast magnetic resonance imaging data from a flow phantom and also to the cerebral arteries of healthy volunteers acquired using a 2D vessel-encoded pseudocontinuous arterial spin labeling pulse sequence. This allows the quantitative flow contribution in downstream vessels to be determined from each major brain-feeding artery. Excellent agreement was found between the actual and estimated flow rates in the phantom, particularly below 4.5 ml/s, typical of the cerebral vasculature. Flow rates measured in healthy volunteers were generally consistent with values found in the literature. This method is likely to be of use in patients with a variety of cerebrovascular diseases, such as the assessment of collateral flow in patients with steno-occlusive disease or the evaluation of arteriovenous malformations.
PMCID: PMC3898265  PMID: 23871963
Blood volume flow rate quantification; Magnetic resonance imaging; Dynamic angiography; Vessel-encoded pseudocontinuous arterial spin labeling; Cerebral arteries
Anesthesia and analgesia  2009;109(3):906-913.
A variety of near-infrared spectroscopy devices can be used to make noninvasive measurements of cerebral tissue oxygen saturation (ScO2). The ScO2 measured by the NIRO 300 spectrometer (Hamamatsu Photonics, Japan) is called the cerebral tissue oxygenation index (TOI) and is an assessment of the balance between cerebral oxygen delivery and utilization. We designed this study to investigate the effect of systemic and intracranial physiological changes on TOI.
Fifteen healthy volunteers were studied during isocapneic hyperoxia and hypoxemia, and normoxic hypercapnea and hypocapnea. Absolute cerebral TOI and changes in oxy- and deoxy-hemoglobin concentrations were measured using a NIRO 300. Changes in arterial oxygen saturation (SaO2), end-tidal carbon dioxide tension (EtCO2), heart rate, mean arterial blood pressure (MBP) and middle cerebral artery blood flow velocity (Vmca) were also measured during these physiological challenges. Changes in cerebral blood volume (CBV) were subsequently calculated from changes in total cerebral hemoglobin concentration.
Baseline TOI was 67.3% with an interquartile range (IQR) of 65.2% - 71.9%. Hypoxemia was associated with a median decrease in TOI of 7.1% (IQR -9.1% to -5.4%) from baseline (p<0.0001) and hyperoxia with a median increase of 2.3% (IQR 2.0% to 2.5%) (p<0.0001). Hypocapnea caused a reduction in TOI of 2.1% (IQR -3.3% to -1.3%) from baseline (p<0.0001) and hypercapnea an increase of 2.6% (IQR 1.4% to 3.7%) (p<0.0001). Changes in SaO2 (p<0.0001), EtCO2 (p<0.0001), CBV (p=0.0003) and MBP (p=0.03) were significant variables affecting TOI. Changes in Vmca (p=0.7) and heart rate (p=0.2) were not significant factors.
TOI is an easy-to-monitor variable that provides real-time, multi-site and noninvasive assessment of the balance between cerebral oxygen delivery and utilization. However, TOI is a complex variable that is affected by SaO2 and EtCO2, and, to a lesser extent, by MBP and CBV. Clinicians need to be aware of the systemic and cerebral physiological changes that can affect TOI in order to interpret changes in this variable during clinical monitoring.
PMCID: PMC2742623  PMID: 19690266
Cerebral Cortex (New York, NY)  2012;23(2):339-348.
Understanding the evolution of regional and hemispheric asymmetries in the early stages of life is essential to the advancement of developmental neuroscience. By using 2 noninvasive optical methods, frequency-domain near-infrared spectroscopy and diffuse correlation spectroscopy, we measured cerebral hemoglobin oxygenation (SO2), blood volume (CBV), an index of cerebral blood flow (CBFi), and the metabolic rate of oxygen (CMRO2i) in the frontal, temporal, and parietal regions of 70 premature and term newborns. In concordance with results obtained using more invasive imaging modalities, we verified both hemodynamic (CBV, CBFi, and SO2) and metabolic (CMRO2i) parameters were greater in the temporal and parietal regions than in the frontal region and that these differences increased with age. In addition, we found that most parameters were significantly greater in the right hemisphere than in the left. Finally, in comparing age-matched males and females, we found that males had higher CBFi in most cortical regions, higher CMRO2i in the frontal region, and more prominent right–left CBFi asymmetry. These results reveal, for the first time, that we can detect regional and hemispheric asymmetries in newborns using noninvasive optical techniques. Such a bedside screening tool may facilitate early detection of abnormalities and delays in maturation of specific cortical areas.
PMCID: PMC3584954  PMID: 22328446
brain development; cerebral asymmetry; diffuse correlation spectroscopy; near-infrared spectroscopy; newborns
Changes in the arterial partial pressure of CO2 (PaCO2) has a direct though transient effect on the cerebral vasculature and cerebral circulation. Decreased PaCO2 levels lead to vasoconstriction and can result in dangerously low levels of cerebral perfusion that resolve in 4–6 h. It is currently believed that perfusion abnormalities contribute to intraventricular hemorrhage (IVH) and periventricular leukomalacia (PVL) in the neonate. PaCO2-induced vasoconstriction may contribute to the pathology of IVH and PVL.
Near-infrared spectroscopy [NIRS; (INVOS cerebral/somatic oximeter; Somanetics Corporation, Troy, MI, USA)] was utilized to determine changes in regional oxygenation (rSO2) of the brain in response to changes in ventilation in isoflurane anesthetized newborn piglets.
Changes in cerebral rSO2 correlated significantly with end-tidal CO2 levels and to blood flow in the common carotid artery. This correlation was significant during baseline conditions, after periods of CO2 loading and during periods of hypothermia.
The results of the study demonstrate the utility of NIRS to accurately reflect changes in cerebral oxygenation and flow to the brain in response to changes in CO2 levels in anesthetized, ventilated neonatal piglets. The use of NIRS may provide an early alert of low levels of cerebral blood flow and brain oxygenation, potentially helping in preventing the progression of IVH or PVL in the neonate.
PMCID: PMC3115161  PMID: 21697979
Common carotid artery blood flow; INVOS; neonatal piglets; PaCO2; regional oxygen saturation (rSO2)
Human brain mapping  2010;31(3):341-352.
With the causes of perinatal brain injuries still unclear and the probable role of hemodynamic instability in their etiology, bedside monitoring of neonatal cerebral hemodynamics with standard values as a function of age are needed. In this study, we combined quantitative frequency domain near infrared spectroscopy (FD-NIRS) measures of cerebral tissue oxygenation (StO2) and cerebral blood volume (CBV) with diffusion correlation spectroscopy (DCS) measures of a cerebral blood flow index (CBFix) to test the validity of the CBV-CBF relationship in premature neonates and to estimate cerebral metabolic rate of oxygen (rCMRO2) with or without the CBFix measurement. We measured 11 premature neonates (28–34 weeks gestational age) without known neurological issues, once a week from one to six weeks of age. In nine patients, cerebral blood velocities from the middle cerebral artery were collected by transcranial Doppler (TCD) and compared with DCS values. Results show a steady decrease in StO2 during the first six weeks of life while CBV remains stable, and a steady increase in CBFix. rCMRO2 estimated from FD-NIRS remains constant but shows wide interindividual variability. rCMRO2 calculated from FD-NIRS and DCS combined increased by 40% during the first six weeks of life with reduced interindividual variability. TCD and DCS values are positively correlated. In conclusion, FD-NIRS combined with DCS offers a safe and quantitative bedside method to assess CBV, StO2, CBF, and rCMRO2 in the premature brain, facilitating individual follow-up and comparison among patients. A stable CBV-CBF relationship may not be valid for premature neonates.
PMCID: PMC2826558  PMID: 19650140
premature neonates; brain hemodynamics; near-infrared spectroscopy; diffuse correlation spectroscopy; cerebral blood flow; cerebral oxygen consumption; brain development
Texas Heart Institute Journal  2012;39(5):744-746.
We describe the performance, in one surgical session, of bilateral pulmonary endarterectomy and a button-technique Bentall operation in a 68-year-old man. The patient had chronic thromboembolic pulmonary hypertension and an ascending aortic aneurysm with moderate aortic regurgitation. The procedures were concurrently completed during short periods of systemic circulatory arrest, with antegrade cerebral perfusion maintained through the brachiocephalic artery at a flow rate of 10 mL/min/kg. The patient's cerebral perfusion was monitored with use of near-infrared spectroscopy, to prevent symmetric bilateral values from falling below 20% of the base value. The patient experienced no multiorgan failure or neurologic sequelae and, by the 6th postoperative day, improved from New York Heart Association functional class IV to class I.
The reliable maintenance of continuous antegrade cerebral perfusion made the lengthy combined operation feasible, with low risk. The use of near-infrared spectroscopy enabled real-time monitoring of the patient's cerebral blood flow. Our experience shows the possibility of safely performing lengthy or multiple procedures in one surgical session.
PMCID: PMC3461666  PMID: 23109784
Brain ischemia/prevention & control; cardiac surgical procedures; endarterectomy/adverse effects/methods; hypertension, pulmonary/physiopathology/surgery; perfusion/methods; pulmonary artery/surgery; recovery of function; treatment outcome
Quantitative techniques have been derived for the measurement of global cerebral blood flow, cerebral blood volume, its response to changing arterial carbon dioxide tension and mixed cerebral venous saturation in the human newborn undergoing intensive care. Normal ranges have been established and significant disturbances of cerebral oxygenation and perfusion have been demonstrated in a variety of pathological conditions. Recently, absolute cerebral deoxyhaemoglobin concentration has been obtained in the newborn using second differential spectroscopy. When combined with the measurement of total cerebral haemoglobin concentration, the mean saturation of cerebral blood (SmcO2) may be obtained, allowing global cerebral oxygenation to be determined continuously in the intensive care unit. Marked changes in the concentrations of cerebral oxy- and deoxyhaemoglobin have been observed in foetuses undergoing labour. Measurements of SmcO2 from the foetal brain prior to delivery have shown the expected close correlation with acid-base status at birth. Although movement artefact remains a theoretical risk during uterine contractions, preliminary measurements of optical path length by intensity-modulated spectroscopy have demonstrated only small fluctuations. In future the clinical application of time, phase and spatially resolved spectroscopy is likely to improve both the quantitative accuracy and the regional specificity of physiological measurements in the foetal and neonatal brain.
PMCID: PMC1691957  PMID: 9232857
This paper constructs an authorship-linked collection or corpus of anonymous, sex-seeking ads found on the classifieds website Craigslist. This corpus is then used to validate an authorship attribution approach based on identifying near duplicate text in ad clusters, providing insight into how often anonymous individuals post sex-seeking ads and where they meet for encounters.
The increasing use of the Internet to arrange sexual encounters presents challenges to public health agencies formulating STD interventions, particularly in the context of anonymous encounters. These encounters complicate or break traditional interventions. In previous work [1], we examined a corpus of anonymous personal ads seeking sexual encounters from the classifieds website Craigslist and presented a way of linking multiple ads posted across time to a single author. The key observation of our approach is that some ads are simply reposts of older ads, often updated with only minor textual changes. Under the presumption that these ads, when not spam, originate from the same author, we can use efficient near-duplicate detection techniques to cluster ads within some threshold similarity. Linking ads in this way allows us to preserve the anonymity of authors while still extracting useful information on the frequency with which authors post ads, as well as the geographic regions in which they seek encounters.
While this process detects many clusters, the lack of a true corpus of authorship-linked ads makes it difficult to validate and tune the parameters of our system. Fortunately, many ad authors provide an obfuscated telephone number in ad text (e.g., 867–5309 becomes 8sixseven5three oh nine) to bypass Craigslist filters, which prohibit including phone numbers in personal ads. By matching phone numbers of this type across all ads, we can create a corpus of ad clusters known to be written by a single author. This authorship corpus can then be used to evaluate and tune our existing near-duplicate detection system, and in the future identify features for more robust authorship attribution techniques.
From 7-1-2009 until 7-1-2011, RSS feeds were collected daily for 8 personal ad categories from 414 sites across the United States, for a total of 67 million ads. To create an anonymous, author-linked corpus, we used a regular expression to identify obfuscated phone numbers in ad text. We measure the ability of near-duplicate detection to link clusters in two ways: 1) detecting all ads in a cluster; and 2) correctly detecting a subset of ads within a single cluster. Ads incorrectly assigned to more than 1 cluster are considered false positives. All results are reported in terms of precision, recall, and F-scores (common information retrieval metrics) across cluster size, expressed as number of ads.
652,014 ads contained phone numbers, producing a total of 46,079 authorship-linked ad clusters. For detecting all ads within a cluster, precision ranged from 0.05 to 0.0 and recall from 0.02 to 0.0 for all cluster sizes. For detecting partial clusters, see Figure 1.
We find that near-duplicate detection alone is insufficient to detect all ads within a cluster. However, we do find that the process can, with high precision and low recall, detect a subset of ads associated with a single author. This follows the intuition that an author’s total set of ads is itself comprised of multiple self-similar subsets. While a near-duplicate detection approach can correctly identify subsets of ads linked to a single author, this process alone cannot attribute multiple clusters to a single author. Future work will explore leveraging additional linguistic features to improve author attribution.
(Top) Evaluations for partial cluster detection using the near-duplicate identification approach to linking anonymous authorship in Craigslist ads and (bottom) the distribution of ad cluster sizes.
PMCID: PMC3692814
Surveillance; Public Health; STDs; Authorship Attribution; Computer Science

Results 1-25 (594985)