ABSTRACT
We have investigated the application of computer-based methods to the detection of microcalcifications in digital mammograms. The computer detection system is based on a difference-image technique in which a signal-suppressed image is subtracted from a signal-enhanced image to remove the structured background in a mammogram. Signal-extraction techniques adapted to the known physical characteristics of microcalcifications are then used to isolate microcalcifications from the remaining noise background. We employ Monte Carlo methods to generate simulated clusters of microcalcifications that are superimposed on normal mammographic backgrounds. This allows quantitative evaluation of detection accuracy of the computer method and the dependence of this accuracy on the physical characteristics of the microcalcifications. Our present computer method can achieve a true-positive cluster detection rate of approximately 80% at a false-positive detection rate of one cluster per image. The potential application of such a computer-aided system to mammographic interpretation is demonstrated by its ability to detect microcalcifications in clinical mammograms.
Subject(s)
Breast Neoplasms/diagnostic imaging , Calcinosis/diagnostic imaging , Mammography/methods , Algorithms , Autoanalysis , Breast Neoplasms/complications , Calcinosis/complications , Computers , Female , Humans , Mammography/instrumentation , Monte Carlo MethodABSTRACT
Data from single 10-minute magnetic resonance scans were used to create three-dimensional (3-D) views of the surfaces of the brain and skin of 12 patients. In each case, these views were used to make a preoperative assessment of the relationship of lesions to brain surface structures associated with movement, sensation, hearing, and speech. Interactive software was written so that the user could "slice" through the 3-D computer model and inspect cross-sectional images at any level. A surgery simulation program was written so that surgeons were able to "rehearse" craniotomies on 3-D computer models before performing the actual operations. In each case, the qualitative accuracy of the 3-D views was confirmed by intraoperative inspection of the brain surface and by intraoperative electrophysiological mapping, when available.
Subject(s)
Brain/pathology , Computer Graphics , Magnetic Resonance Imaging , Neurosurgery/methods , Patient Care Planning , Adolescent , Adult , Aged , Brain/surgery , Brain Diseases/diagnosis , Brain Diseases/surgery , Brain Neoplasms/diagnosis , Brain Neoplasms/surgery , Child , Computer Simulation , Female , Humans , Middle AgedABSTRACT
OBJECTIVE: To study the incidence, patient and event characteristics, and outcome of rapid response system (RRS) activation on an in-hospital haemodialysis unit. DESIGN: Retrospective review of all RRS events on an in-hospital 10-bed haemodialysis unit over a 64-month period (November 2001 to February 2007). SETTING: University of Pittsburgh Medical Center Presbyterian Hospital, a 730-bed academic, urban, tertiary care adult hospital in the USA. Interventions None. RESULTS: Over a 64-month-period, 107 of 8928 patients undergoing haemodialysis on the dialysis unit required an RRS activation (12 events/1000 patients dialysed). The most common reasons for RRS activation were respiratory distress/hypoxaemia (27%) and mental status change (24%). Predictors of in-hospital mortality included old age (33% in-hospital mortality for patients aged 65 years or older vs 14% for patients aged less than 65 years; χ(2) = 5.66, df = 1, p = 0.017), and RRS activation due to a respiratory abnormality (37% mortality for respiratory codes vs 18% for all other codes; χ(2) = 4.12, df = 1, p = 0.042). Surprisingly, only 71% of the patients who had an RRS event had the event as dialysis was occurring. Twenty-four patients (22%) met one or more RRS activation criteria upon first vital sign check in the dialysis unit; RRS was activated on 12 (11%) of these patients before dialysis was started. Nineteen (18%) additional patients had an RRS event after their dialysis session had ended, while awaiting transport back to their unit. CONCLUSIONS: From our findings, it can be suggested that critical events often occur before and after dialysis treatment, during or awaiting transport. Careful assessment of these high-risk patients before and after transport, to and from the dialysis unit may be warranted.
Subject(s)
Hemodialysis Units, Hospital , Hospital Rapid Response Team , Outcome Assessment, Health Care , Academic Medical Centers , Adult , Databases, Factual , Female , Hospital Bed Capacity, 500 and over , Humans , Male , Middle Aged , Outcome Assessment, Health Care/methods , Pennsylvania , Retrospective Studies , Young AdultABSTRACT
INTRODUCTION: Arterial catheterisation is used for continuous haemodynamic monitoring in patients undergoing surgery and in critical care units. Although it is considered a safe procedure, a major complication such as arterial occlusion and limb gangrene can occur. OBJECTIVE: To determine the incidence, outcome and potential to avoid complications associated with arterial catheterisation. METHODS: The number of arterial catheterisation was determined using an anaesthesiology and critical care medicine billing database over a period of 4 years (1 January 2003 to 31 December 2006). Possible major complications were identified from two hospital databases; all identified charts were screened and then reviewed by an expert panel that determined causation. A major complication was defined as requiring operative intervention and/or resulting in permanent harm. RESULTS: 15 (0.084%) major complications were identified among 17 840 instances of arterial catheterisation insertions. Of 15 arterial catheterisations, nine were performed in the operating room and six in the intensive care unit. Nine patients suffered ischaemic injury, which progressed to gangrene in three patients. Three patients developed haematoma that required surgical evacuation; two of these required vascular repair. One patient had compartment syndrome requiring fasciotomy and two patients had sheared catheter fragments that needed to be removed. All 15 patients had multiple comorbidities, and those in the operating room had an American Society of Anesthesiologists score of >or=3. Seven (46.6%) had arterial catheterisation done under emergent circumstances. Six (40%) died during hospitalisation because of complications unrelated to arterial catheterisation. CONCLUSION: Arterial catheterisation had a very low rate of major complications. They seem associated with high severity of illness and emergency surgery.
Subject(s)
Catheterization/adverse effects , Hospitals, University/standards , Patient Safety/standards , Adult , Aged , Aged, 80 and over , Arteries , Critical Care , Extremities/blood supply , Female , Gangrene/etiology , Gangrene/prevention & control , Hematoma/etiology , Hematoma/prevention & control , Humans , Ischemia/etiology , Ischemia/prevention & control , Male , Middle Aged , Monitoring, Intraoperative/adverse effects , Monitoring, Intraoperative/methods , Monitoring, Physiologic/adverse effects , Monitoring, Physiologic/methods , Pennsylvania , Retrospective Studies , Risk Factors , Young AdultABSTRACT
Image processing and volume rendering have been applied to magnetic resonance (MR) images to produce three-dimensional views of the surface of the brain. Four healthy volunteers and 17 patients with a variety of disorders underwent MR imaging of the head, and these images were processed and subjected to volume rendering. The resulting three-dimensional views depict abnormalities of the brain surface, as well as important landmarks of normal brain surface anatomy such as precentral, postcentral, superior temporal, and inferior frontal gyri, which are difficult to identify on cross-sectional images.
Subject(s)
Brain/anatomy & histology , Computer Graphics , Image Processing, Computer-Assisted , Magnetic Resonance Imaging/methods , Adult , Aged , Child , Female , Humans , Male , Meningeal Neoplasms/diagnosis , Meningioma/diagnosisABSTRACT
Three patients with intractable epilepsy, two with brain tumors, and one with encephalitis were imaged with magnetic resonance (MR) and positron emission tomography (PET). MR data were used to construct a three-dimensional (3D) computer model of the brain surface depicting the precentral (movement), postcentral (sensation), left inferior frontal (speech), and left superior temporal (hearing) gyri. PET-derived measurements of average surface metabolism were encoded as colors and mapped onto the 3D model by means of a retrospective technique for registering the two scans. The integrated 3D model depicted the location of PET-detected metabolic abnormalities with respect to the gyral anatomy visualized with MR. In each case, the predicted relationships were confirmed intraoperatively by means of inspection of the brain and electroencephalography. Multimodality 3D displays are likely to be particularly valuable for interpreting PET studies of epileptic patients and others with normal MR anatomy.