Full automation of total metabolic tumor volume from FDG-PET/CT in DLBCL for baseline risk assessments.

BACKGROUND: Current radiological assessments of 18fluorodeoxyglucose-positron emission tomography (FDG-PET) imaging data in diffuse large B-cell lymphoma (DLBCL) can be time consuming, do not yield real-time information regarding disease burden and organ involvement, and hinder the use of FDG-PET to potentially limit the reliance on invasive procedures (e.g. bone marrow biopsy) for risk assessment. METHODS: Our aim is to enable real-time assessment of imaging-based risk factors at a large scale and we propose a fully automatic artificial intelligence (AI)-based tool to rapidly extract FDG-PET imaging metrics in DLBCL. On availability of a scan, in combination with clinical data, our approach generates clinically informative risk scores with minimal resource requirements. Overall, 1268 patients with previously untreated DLBCL from the phase III GOYA trial (NCT01287741) were included in the analysis (training: n = 846; hold-out: n = 422). RESULTS: Our AI-based model comprising imaging and clinical variables yielded a tangible prognostic improvement compared to clinical models without imaging metrics. We observed a risk increase for progression-free survival (PFS) with hazard ratios [HR] of 1.87 (95% CI: 1.31-2.67) vs 1.38 (95% CI: 0.98-1.96) (C-index: 0.59 vs 0.55), and a risk increase for overall survival (OS) (HR: 2.16 (95% CI: 1.37-3.40) vs 1.40 (95% CI: 0.90-2.17); C-index: 0.59 vs 0.55). The combined model defined a high-risk population with 35% and 42% increased odds of a 4-year PFS and OS event, respectively, versus the International Prognostic Index components alone. The method also identified a subpopulation with a 2-year Central Nervous System (CNS)-relapse probability of 17.1%. CONCLUSION: Our tool enables an enhanced risk stratification compared with IPI, and the results indicate that imaging can be used to improve the prediction of central nervous system relapse in DLBCL. These findings support integration of clinically informative AI-generated imaging metrics into clinical workflows to improve identification of high-risk DLBCL patients. TRIAL REGISTRATION: Registered clinicaltrials.gov number: NCT01287741.

Magnetic resonance enterography is feasible and reliable in multicenter clinical trials in patients with Crohn's disease, and may help select subjects with active inflammation.

BACKGROUND: Reliable tools for patient selection are critical for clinical drug trials. AIM: To evaluate a consensus-based, standardised magnetic resonance enterography (MRE) protocol for selecting patients for inclusion in Crohn's disease (CD) multicenter clinical trials. METHODS: This study recruited 20 patients [Crohn's Disease Activity Index (CDAI) scores: <150 (n = 8); 150-220 (n = 4); 220-450 (n = 8)], to undergo ileocolonoscopy and two MREs (with and without colonic contrast) within a 14-day period. Procedures were scored centrally using, Magnetic Resonance Index of Activity (MaRIA), and both Crohn's Disease Endoscopic Index of Severity (CDEIS) and Simplified Endoscopic Score (SES-CD). RESULTS: 37 MREs were acquired. Both MREs were evaluable in 16 patients for calculation of test-retest and inter-reader reliability scores. The MaRIA scores for the terminal ileum had excellent test-retest and inter-reader reliability, with correlations >0.9. The proximal ileum showed strong within-reader agreement (0.90-0.96), and fair between-reader agreement (0.59-0.72). MRE procedures were tolerable. MaRIA scores correlated with CDEIS and SES-CD (0.63 and 0.71), but not with CDAI (0.34). MRE identified 3 patients with intra-abdominal complications, who would otherwise have been included in clinical trials. Furthermore, both MRE and ileocolonoscopy identified active bowel wall inflammation in 2 patients with CDAI <150, and none in 1 patient with CDAI > 220. Data quality was good/excellent in 85% of scans, and fair or better in 96%. CONCLUSIONS: Magnetic resonance enterography of high-quality and reproducibility was feasible in a global multi- centre setting, with evidence for improved selectivity over CDAI and ileocolonoscopy in identifying appropriate CD patients for inclusion in therapeutic intervention trials.

Diffusion spectrum magnetic resonance imaging (DSI) tractography of crossing fibers.

MRI tractography is the mapping of neural fiber pathways based on diffusion MRI of tissue diffusion anisotropy. Tractography based on diffusion tensor imaging (DTI) cannot directly image multiple fiber orientations within a single voxel. To address this limitation, diffusion spectrum MRI (DSI) and related methods were developed to image complex distributions of intravoxel fiber orientation. Here we demonstrate that tractography based on DSI has the capacity to image crossing fibers in neural tissue. DSI was performed in formalin-fixed brains of adult macaque and in the brains of healthy human subjects. Fiber tract solutions were constructed by a streamline procedure, following directions of maximum diffusion at every point, and analyzed in an interactive visualization environment (TrackVis). We report that DSI tractography accurately shows the known anatomic fiber crossings in optic chiasm, centrum semiovale, and brainstem; fiber intersections in gray matter, including cerebellar folia and the caudate nucleus; and radial fiber architecture in cerebral cortex. In contrast, none of these examples of fiber crossing and complex structure was identified by DTI analysis of the same data sets. These findings indicate that DSI tractography is able to image crossing fibers in neural tissue, an essential step toward non-invasive imaging of connectional neuroanatomy.

Superselective intracerebral catheterization of a branch of the internal carotid artery coupled with magnetic resonance imaging.

SUMMARY: We used fluoroscopic guidance and over-thewire techniques to superselectively place a microcatheter into a branch of the MCA of three macaques and MRI bolus tracking techniques to measure perfusion within the selected brain region. Such techniques are likely to be useful in the assessment and treatment of ischemic infarction, cerebral vasospasm, and monitoring local delivery of drugs into the brain.

Long-term monitoring of post-stroke plasticity after transient cerebral ischemia in mice using in vivo and ex vivo diffusion tensor MRI.

WE USED A MURINE MODEL OF TRANSIENT FOCAL CEREBRAL ISCHEMIA TO STUDY: 1) in vivo DTI long-term temporal evolution of the apparent diffusion coefficient (ADC) and diffusion fractional anisotropy (FA) at days 4, 10, 15 and 21 after stroke 2) ex vivo distribution of a plasticity-related protein (GAP-43) and its relationship with the ex vivo DTI characteristics of the striato-thalamic pathway (21 days). All animals recovered motor function. In vivo ADC within the infarct was significantly increased after stroke. In the stroke group, GAP-43 expression and FA values were significantly higher in the ipsilateral (IL) striatum and contralateral (CL) hippocampus compared to the shams. DTI tractography showed fiber trajectories connecting the CL striatum to the stroke region, where increased GAP43 and FA were observed and fiber tracts from the CL striatum terminating in the IL hippocampus.Our data demonstrate that DTI changes parallel histological remodeling and recovery of function.

A comparison of CH3-DTPA-GD (NMS60) and GD-DTPA for evaluation of acute myocardial ischemia.

PURPOSE: Our objective was to evaluate the use of a new medium weight MRI contrast agent, NMS60 (a synthetic oligomeric Gd-complex containing three Gd(3+) atoms, molecular weight 2158 Da) compared to gadolinium-diethylene triamine pentaacetic acid (Gd-DTPA) in a pig myocardial ischemia model. MATERIALS: We used 13 male white hybrid pigs. Animals were scanned in the acute phase 2-3 h after the onset of myocardial ischemia. Scans were acquired on a 1.5T GE Signa with dynamic T1-weighted imaging during a bolus injection of 0.1 mmol(gd)/kg of either NMS60 or Gd-DTPA, 2D CINE at 5 min after injection, and T1-weighted spin-echo imaging up to 60 min. RESULTS: The postcontrast CINE scans showed improved contrast-to-noise ratio after NMS60 injection, compared to Gd-DTPA. There was significantly greater enhancement with NMS60 in both normal myocardium and in the ischemic lesion on T1-weighted spin-echo scans up to 60 min after injection. The dose ranging study shows a 24% greater enhancement with NMS60 compared to Gd-DTPA. DISCUSSION: This new medium weighted contrast agent offers improved enhancement for cardiac MRI, compared to Gd-DTPA, with similar washout kinetics and lower toxicity, and may prove useful for better detection of myocardial ischemia as well as delayed or hyperenhancement after reperfusion.

Near-infrared frequency-domain optical spectroscopy and magnetic resonance imaging: a combined approach to studying cerebral maturation in neonatal rabbits.

The neonatal rabbit brain shows prolonged postnatal development both structurally and physiologically. We use noninvasive near-IR frequency-domain optical spectroscopy (NIRS) and magnetic resonance imaging (MRI) to follow early developmental changes in cerebral oxygenation and anatomy, respectively. Four groups of animals are measured: NIRS in normals, MRI in normals, and both NIRS and MRI with hypoxia-ischemia (HI) (diffusion MRI staging). NIRS and/or MRI are performed from P3 (postnatal day=P) up to P76. NIRS is performed on awake animals with a frequency-domain tissue photometer. Absolute values of oxyhemoglobin concentration ([HbO2]), deoxyhemoglobin concentration ([HbR]), total hemoglobin concentration (HbT), and hemoglobin saturation (StO2) are calculated. The brains of all animals appeared to be maturing as shown in the diffusion tensor MRI. Mean optical coefficients (reduced scattering) remained unchanged in all animals throughout. StO2 increased in all animals (40% at P9 to 65% at P43) and there are no differences between normal, HI controls, and HI brains. The measured increase in StO2 is in agreement with the reported increase in blood flow during the first 2 months of life in rabbits. HbT, which reflects blood volume, peaked at postnatal day P17, as expected since the capillary density increases up to P17 when the microvasculature matures.

A comparative evaluation of CH3-DTPA-Gd (NMS60) for contrast-enhanced magnetic resonance angiography.

In a canine model the signal dynamics of a new oligomer-based MR contrast agent (NMS60, 2158 Da) were compared to Gd-DTPA to investigate the agents' potential for magnetic resonance angiography (MRA). Twelve male mongrel dogs were imaged sequentially under anesthesia with two different MRA sequences (Tlw 3DSPGR). Initial enhancement was measured every 9 s for eight points in time. Thereafter, spatial highly resolved MRAs were obtained at 5, 10, 15, 20, 30, 45, and 60 min post-injection of two different dosages. Over the first 20 s following bolus administration the average arterial enhancement of 0.1 mmol(Gd)kg NMS60 was 44% greater than Gd-DTPA. Twenty minutes post-injection the relative signal intensity of NMS60 was as high as the peak signal intensity with Gd-DTPA at the same dosage level (0.1 mmol(Gd)/kg). In the animals that received NMS60 injections the vascular conspicuity was overly superior to those who received Gd-DTPA. No significant toxicity effects were noted for either dosage level. The intermediate weight contrast agent NMS60 offers greater vascular enhancement and retention time than Gd-DTPA. For a given set of optimized imaging parameters this offers improved spatial details, less arterial/venous overlap, and better vascular contrast.

Multilocal magnetic resonance perfusion mapping comparing the cerebral hemodynamic effects of decompressive craniectomy versus reperfusion in experimental acute hemispheric stroke in rats.

This study examined the hemodynamic effects of craniectomy compared to reperfusion on the temporal evolution of cerebral perfusion in different brain regions in a rat model of focal cerebral ischemia. Three groups were investigated: no treatment, reperfusion or craniectomy at 1 h. Perfusion-weighted magnetic resonance imaging (PWI) was performed serially from 0.5 to 6 h. Relative regional cerebral blood flow was calculated for different regions and infarct volume was assessed by histology at 24 h. As conclusion, both, craniectomy and reperfusion increased cerebral perfusion in the acute phase of cerebral ischemia. While reperfusion resulted in a homogeneous improvement of perfusion in the cortex and basal ganglia, craniectomy improved only cortical perfusion in areas directly under the craniectomy site. PWI is well suited to non-invasively monitor perfusion alterations after aggressive therapeutical approaches in stroke.

Evolution of apparent diffusion coefficient, diffusion-weighted, and T2-weighted signal intensity of acute stroke.

BACKGROUND AND PURPOSE: Serial study of such MR parameters as diffusion-weighted imaging (DWI), apparent diffusion coefficient (ADC), ADC with fluid-attenuated inversion recovery (ADC(FLAIR)), and T2-weighted imaging may provide information on the pathophysiological mechanisms of acute ischemic stroke. Our goals were to establish the natural evolution of MR signal intensity characteristics of acute ischemic lesions and to assess the potential of using specific MR parameters to estimate lesion age. METHODS: Five serial echo-planar DWI studies with and without an inversion recovery pulse were performed in 27 patients with acute stroke. The following lesion characteristics were studied: 1) conventional ADC (ADC(CONV)); 2) ADC(FLAIR); 3) DWI signal intensity (SI(DWI)); 4) T2-weighted signal intensity (SI(T2)), and 5) FLAIR signal intensity (SI(FLAIR)). RESULTS: The lesion ADC(CONV) gradually increased from low values during the first week to pseudonormal during the second week to supranormal thereafter. The lesion ADC(FLAIR) showed the same pattern of evolution but with lower absolute values. A low ADC value indicated, with good sensitivity (88%) and specificity (90%), that a lesion was less than 10 days old. All signal intensities remained high throughout follow-up. SI(DWI) showed no significant change during the first week but decreased thereafter. SI(T2) initially increased, decreased slightly during week 2, and again increased after 14 days. SI(FLAIR) showed the same initial increase as the SI(T2) but remained relatively stable thereafter. CONCLUSION: Our findings further clarify the time course of stroke evolution on MR parameters and indicate that the ADC map may be useful for estimating lesion age. Application of an inversion recovery pulse results in lower, potentially more accurate, absolute ADC values.

MRI of subacute hemorrhagic transformation in the rat suture occlusion model.

In this study we investigated the utility of different MRI techniques for the detection and predictability of hemorrhagic transformation (HT) in a rat model of transient focal cerebral ischemia. Hemorrhagic infarction was reliably identified with gradient-echo sequences and developed between 2 and 7 days following the insult. None of the investigated early MRI features of the ischemic lesions (including the apparent diffusion coefficient and post-reperfusion blood-brain barrier damage) was a good predictor of HT severity at 7 days. This indicates that subacute HT at 2-7 days occurs independently of the severity of acute tissue and BBB damage.

Comparison of diffusion, blood oxygenation, and blood volume changes during global ischemia in rats.

Rapid diffusion, blood oxygenation, and blood volume weighted echo planar imaging was used to monitor global cerebral ischemia by cardiac arrest in rats. Serial CBV measurements used intravascular iron oxide contrast media (iron dextran). ADC dropped by 5% within 20 sec of cardiac arrest, then continued to decay slowly until a larger rapid drop after 2 min. After iron oxide injection, the initial 5% drop was not observed. The transverse relaxation rate (R(2), R(*)(2) no iron injection) increased rapidly after cardiac arrest, peaking at about 30 sec, then declining towards baseline. The CBV dropped by about 50% within 20 sec. The initial 5% ADC drop may be a vascular artifact. The rapidity of the CBV-weighted signal drop suggests a flow-mediated contribution to the iron oxide contrast mechanism. Magn Reson Med 45:10-16, 2001.

High speed diffusion magnetic resonance imaging of ischemia and spontaneous periinfarct spreading depression after thromboembolic stroke in the rat.

Spontaneous episodes of transient cell membrane depolarization (spreading depression [SD]) occur in the surroundings of experimental stroke lesions and are believed to contribute to infarct growth. Diffusion-weighted imaging (DWI) is capable of detecting the water shifts from extracellular to intracellular space associated with SD waves and ischemia, and can make in vivo measurements of these two features on a pixel-by-pixel basis with good temporal resolution. Using continuous high speed DWI with a temporal resolution of 12 seconds over a period of 3 hours, the in vivo contribution of spontaneous SDs to the development of ischemic tissue injury was examined in 8 rats using a thromboembolic stroke model. During the observation period, the initial lesion volume increased in 4 animals, remained unchanged in 1 animal, and decreased in 3 animals (most likely because of spontaneous clot lysis). Irrespective of the lesion evolution patterns, animals demonstrated 6.5 +/- 2.1 spontaneous SDs outside of the ischemic core. A time-to-peak analysis of apparent diffusion coefficient (ADC) changes for each SD wave demonstrated multidirectional propagation patterns from variable initiation sites. Maps of the time constants of ADC recovery, reflecting the local energy supply and cerebral blood flow, revealed prolonged recovery times in areas close to the ischemic core. However, repetitive SD episodes in the periinfarct tissue did not eventually lead to permanent ADC reductions. These results suggest that spontaneous SD waves do not necessarily contribute to the expansion of the ischemic lesion volume in this model.

99mTc annexin V imaging of neonatal hypoxic brain injury.

BACKGROUND AND PURPOSE: Delayed cell loss in neonates after cerebral hypoxic-ischemic injury (HII) is believed to be a major cause of cerebral palsy. In this study, we used radiolabeled annexin V, a marker of delayed cell loss (apoptosis), to image neonatal rabbits suffering from HII. METHODS: Twenty-two neonatal New Zealand White rabbits had ligation of the right common carotid artery with reduction of inspired oxygen concentration to induce HII. Experimental animals (n=17) were exposed to hypoxia until an ipsilateral hemispheric decrease in the average diffusion coefficient occurred. After reversal of hypoxia and normalization of average diffusion coefficient values, experimental animals were injected with (99m)Tc annexin V. Radionuclide images were recorded 2 hours later. RESULTS: Experimental animals showed no MR evidence of blood-brain barrier breakdown or perfusion abnormalities after hypoxia. Annexin images demonstrated multifocal brain uptake in both hemispheres of experimental but not control animals. Histology of the brains from experimental animals demonstrated scattered pyknotic cortical and hippocampal neurons with cytoplasmic vacuolization of glial cells without evidence of apoptotic nuclei by terminal deoxynucleotidyl transferase-mediated dUTP nick end-labeling (TUNEL) staining. Double staining with markers of cell type and exogenous annexin V revealed that annexin V was localized in the cytoplasm of scattered neurons and astrocytes in experimental and, less commonly, control brains in the presence of an intact blood-brain barrier. CONCLUSIONS: Apoptosis may develop after HII even in brains that appear normal on diffusion-weighted and perfusion MR. These data suggest a role of radiolabeled annexin V screening of neonates at risk for the development of cerebral palsy.

Spreading depression-induced expression of c-fos and cyclooxygenase-2 in transgenic mice that overexpress human copper/zinc-superoxide dismutase.

Spreading depression (SD) is a wave of sustained depolarization challenging the energy metabolism of cells without causing irreversible damage. SD is a major mechanism of gene induction that takes place in cortical injury, including ischemia. We studied the role of oxygen radicals in SD-induced c-fos and cyclooxygenase-2 (COX-2) induction using transgenic (Tg) mice that overexpress copper/zinc-superoxide dismutase (SOD1). The frequency, amplitude and duration of SD waves were similar in the Tg mice and wild-type littermates. c-fos and COX-2 mRNAs were strongly induced 1 and 4 h after SD. The induction of both genes was slightly but significantly less at 4 h in the Tg mice. The results indicate that even a mild, noninjurious metabolic stimulation increases the concentration of oxygen radicals to the level that contributes to gene expression.

Serial MRI after transient focal cerebral ischemia in rats: dynamics of tissue injury, blood-brain barrier damage, and edema formation.

BACKGROUND AND PURPOSE: With the advent of thrombolytic therapy for acute stroke, reperfusion-associated mechanisms of tissue injury have assumed greater importance. In this experimental study, we used several MRI techniques to monitor the dynamics of secondary ischemic damage, blood-brain barrier (BBB) disturbances, and the development of vasogenic edema during the reperfusion phase after focal cerebral ischemia in rats. METHODS: Nineteen Sprague-Dawley rats were subjected to transient middle cerebral artery occlusion of 30 minutes, 60 minutes, or 2.5 hours with the suture occlusion model. MRI, including diffusion-weighted imaging (DWI), T2-weighted imaging, perfusion-weighted imaging, and T1-weighted imaging, was performed 5 to 15 minutes before reperfusion, as well as 0.5, 1.5, and 2.5 hours and 1, 2, and 7 days after withdrawal of the suture. Final infarct size was determined histologically at 7 days. RESULTS: In the 30-minute ischemia group (and partially also after 60 minutes), DWI abnormalities reversed transiently during the early reperfusion period but recurred after 1 day, probably due to secondary ischemic damage. After 2.5 hours of ischemia, DWI abnormalities no longer reversed, and signal intensity on both DWI and T2-weighted images increased rapidly in the previously ischemic region due to BBB damage (enhancement on postcontrast T1-weighted images) and edema formation. Early BBB damage during reperfusion was found to be predictive of relatively pronounced edema at subacute time points and was probably related to the increased mortality rates in this experimental group (3 of 7). CONCLUSIONS: Reperfusion after short periods of ischemia (30 to 60 minutes) appears to be mainly complicated by secondary ischemic damage as shown by the delayed recurrence of the DWI lesions, whereas BBB damage associated with vasogenic edema becomes a dominant factor with longer occlusion times (2.5 hours).

Spreading waves of transient and prolonged decreases in water diffusion after subarachnoid hemorrhage in rats.

Diffusion-weighted MRI (DWI), which can detect cortical spreading depressions (SDs) as propagating waves of reduced apparent diffusion coefficient (ADC) of water, was used to investigate whether spreading depression occurs after subarachnoid hemorrhage (SAH) induced by endovascular perforation in the rat. Eleven rats underwent SAH while positioned in the magnet. The ADC measurements had a temporal resolution of 12 sec. Transient decreases in ADC to 74 +/- 5% of pre-SAH values were observed in three rats after SAH, which propagated over the cortex with an average speed of 4.2 +/- 0. 6 mm/min, consistent with an SD wave. Furthermore, in all 11 rats, a wavefront of reduced ADC, which did not resolve within the 12 min observation period, spread at a speed of 3.2 +/- 1.7 mm/min in the ipsilateral cortex, and again is consistent with the speed of SD propagation. Therefore, spreading depression-like cellular depolarization is a consequence of acute subarachnoid hemorrhage in rats. Magn Reson Med 44:110-116, 2000.

MRI of focal cerebral ischemia using (17)O-labeled water.

This work presents a novel approach for quantifying low concentrations of H(2)(17)O in vivo and explores its utility for assessing cerebral ischemia. Oxygen-17 enriched water acts as a T(2) shortening contrast agent whose effect can be suppressed by decoupling at the (17)O frequency during TE interval in a spin-echo MR image. Serial T(2)-weighted echo planar images were acquired in phantoms and rat brain with decoupler power alternated every eight images. The resulting periodic signal change (proportional to H(2)(17)O concentration) was detected by cross-correlating the square-wave decoupler power timecourse with the signal intensity in each voxel. Natural abundance (0.037 atom%) images of H(2)(17)O in rat brain were generated. The transverse relaxivity of H(2)(17)O in brain was estimated, R(2) = 2.4+/-0.5 s(-1)(atom%)(-1). After bolus injection of 1 ml of 10 atom% H(2)(17)O, brain H(2)(17)O concentration was estimated at 0.06+/-0.01 atom%. In the rat focal ischemia model, (17)O cross-correlation maps compared well with diffusion and Gd-DTPA perfusion images to indicate infarct location. Magn Reson Med 43:876-883, 2000.

Dynamic contrast-enhanced MRI of Implanted VX2 tumors in rabbit muscle: comparison of Gd-DTPA and NMS60.

We studied the dynamics of injected contrast enhancement in implanted VX2 tumors in rabbit thigh muscle. We compared two contrast agents Gd-DTPA and NMS60, a novel gadolinium containing trimer of molecular weight 2.1 kd. T1-weighted spin echo images were acquired preinjection and at 5-60 min after i.v. injection of 0.1 mmol/kg of agent. Dynamic T1-weighted SPGR images (1.9 s/image) were acquired during the bolus injection. Male NZW rabbits (n = 13) were implanted with approximately 2 x 10(6) VX2 tumor cells and grew tumors of 28+/-27 mL over 12 to 21 days. NMS60 showed significantly greater peak enhancement in muscle, tumor rim, and core compared to DTPA in both T1-weighted and SPGR images. NMS60 also showed delayed peak enhancement in the dynamic scans (compared to Gd-DTPA) and significantly reduced leakage rate constant into the extravascular space for tumor rim (K21 = 5.1 min(-1) vs. 11.5 min(-1) based on a 2 compartment kinetic model). The intermediate weight contrast agent NMS60 offers greater tumor enhancement than Gd-DTPA and may offer improved regional differentiation on the basis of vascular permeability in tumors.

Longitudinal magnetic resonance imaging study of perfusion and diffusion in stroke: evolution of lesion volume and correlation with clinical outcome.

A prospective longitudinal diffusion-weighted and perfusion-weighted magnetic resonance imaging (DWI/PWI) study of stroke patients (n = 21) at five distinct time points was performed to evaluate lesion evolution and to assess whether DWI and PWI can accurately and objectively demonstrate the degree of ischemia-induced deficits within hours after stroke onset. Patients were scanned first within 7 hours of symptom onset and then subsequently at 3 to 6 hours, 24 to 36 hours, 5 to 7 days, and 30 days after the initial scan. Lesion evolution was dynamic during the first month after stroke. Most patients (18 of 19, 95%) showed increased lesion volume over the first week and then decreased at 1 month relative to 1 week (12 of 14, 86%). Overall, lesion growth appeared to depend on the degree of mismatch between diffusion and perfusion at the initial scan. Abnormal volumes on the acute DWI and PWI (<7 hours) correlated well with initial National Institutes of Health (NIH) stroke scale scores, outcome NIH stroke scale scores, and final lesion volume. DWI and PWI can provide an early measure of metabolic and hemodynamic insufficiency, and thus can improve our understanding of the evolution and outcome after acute ischemic stroke.