Multi-atlas based segmentation using probabilistic label fusion with adaptive weighting of image similarity measures.

Label fusion multi-atlas approaches for image segmentation can give better segmentation results than single atlas methods. We present a multi-atlas label fusion strategy based on probabilistic weighting of distance maps. Relationships between image similarities and segmentation similarities are estimated in a learning phase and used to derive fusion weights that are proportional to the probability for each atlas to improve the segmentation result. The method was tested using a leave-one-out strategy on a database of 21 pre-segmented prostate patients for different image registrations combined with different image similarity scorings. The probabilistic weighting yields results that are equal or better compared to both fusion with equal weights and results using the STAPLE algorithm. Results from the experiments demonstrate that label fusion by weighted distance maps is feasible, and that probabilistic weighted fusion improves segmentation quality more the stronger the individual atlas segmentation quality depends on the corresponding registered image similarity. The regions used for evaluation of the image similarity measures were found to be more important than the choice of similarity measure.

Oxygen cost and energy expenditure of racquetball.

AIM: Many sports are played intermittently in bursts of high, moderate, and low intensity activity. The pattern of exercise intensity has been assessed from heart rate (HR) and time motion analysis and few data are available based on assessment of exercise intensity by means of direct measurement of oxygen uptake. The aim of this study was to directly assess oxygen uptake (VO(2)) using a portable metabolic measuring device to describe the aerobic demand and to determine the associated energy expenditure (EE). METHODS: Fourteen recreational racquetball players (3 females and 11 males; mean age, height, and mass of 23.1+/-2.8 years, 178.1+/-7.1 cm, and 81.1+/-19.6 kg, respectively) played for 40 min while wearing a portable metabolic system to assess VO(2) and a Polar watch to measure HR. Rating of perceived exertion (RPE) (Borg 6-20 scale) was assessed at 5 min intervals during breaks in play. RESULTS: The mean VO(2), HR, and RPE over 40 min of play were 27.3+/-1.8 mL/kg/min, 155.3+/-2.8 bpm, and 12.9+/-0.6, respectively. The associated EE was a total of 1 844 kj (440 kcal) or 465 kj/min (11.1 kcal/min). These data were examined across 13.3 min time periods of play, in order to determine if they changed during play because of fatigue. VO(2) decreased significantly (P<0.05), while HR increased (P<0.05). RPE rose across periods of play, but not significantly. CONCLUSION: Recreational racquetball appears to elicit a VO(2) that would allow aerobic conditioning in many persons. It does so with an RPE that is nearly ''somewhat hard''.

Participatory design of information systems in health care.

AIM: To study the objectives, processes, and ideologies expressed during participatory design of information systems (PDIS) in health care. DESIGN: Longitudinal documentation of project meetings by video recording. Grounded theory development in three steps. SETTING: Systems development project in primary care. RESULTS: The developing system was discussed mainly from a clinical and practical, as opposed to a technical, viewpoint. The design decisions were related to societal-level participants and institutions. This external influence on design decisions was mediated by design voices in discussions, each having its own scope. CONCLUSIONS: The social environment has to be considered when applying PDIS in health care. Instructions for nondesigners can be used to introduce them to the design objectives, processes, and ideologies on which PDIS is based and to support them when relating clinical and practical design questions to the existing social constraints and norms.

Development of systems for support of collaboration in health care: the design arenas.

To explore the design of computer-supported collaborative work in health care, a case study is described addressing the social contexts and conditions influencing the development process. The data set covers 13 consecutive meetings held in a systems design group over a 2-year period, in total approximately 24 h of video recordings. Subjectivist methods are used for the data analyses. The results suggest that the development of computer-supported collaborative work in health care is situated at three social arenas: the societal arena, the organizational arena and the workplace arena. These are visited by the design group in patterns which correspond to the micro-, meso- and macro-level social structures involved in the design. The study displays that longitudinal analyses of design meeting dialogues provide the opportunity of improving the understanding of external influences on design processes in health care.

A prospective and consecutive study of surgically treated lumbar spinal stenosis. Part I: Clinical features related to radiographic findings.

STUDY DESIGN: A prospective study of consecutive patients undergoing surgery for central lumbar spinal stenosis. OBJECTIVES: To evaluate symptoms and signs in patients undergoing surgery for central spinal stenosis and to correlate the findings to age, preoperative duration of symptoms, and radiographically detected constriction. SUMMARY OF BACKGROUND DATA: The degree of constriction of the spinal canal considered to be symptomatic of lumbar canal stenosis is not clear, nor is the relation between the clinical appearance of the disease and the degree of radiographically verified constriction. MATERIAL AND METHODS: One hundred five consecutive patients scheduled for decompression surgery were included in a prospective study. The day before surgery, all patients were interviewed and examined, using a defined protocol that included data on age, gender, preoperative duration of symptoms, walking ability, and occurrence of pain at rest and at night. Included were data recording straight leg raising test results, reflex disturbance, and extensor hallucis longus muscle weakness. All radiographs were examined by a neuroradiologist. The anteroposterior diameters at each site from L1-L2 to L5-S1 were recorded. For the computer analysis, the site of and width at the narrowest site was registered, as well as the number of sites with an anteroposterior diameter of less than 10 mm. A statistical analysis was performed using chi-square analysis, nonparametric tests, analysis of variance, and logistic regression. RESULTS: Pain at rest and at night was reported by 68 and 60 patients, respectively, and was more common in younger patients (P = 0.065 and 0.015, respectively). A severe reduction of walking ability (< 0.5 km) was reported by 70 patients. The straight leg raising test results were negative in 70 patients, positive > 60 degrees in 16, positive 30-60 degrees in 14, and positive < 30 degrees in 5. Younger patients had a positive straight leg raising result (P = 0.028, analysis of variance) more often. Reflex disturbances correlated to patient age: Older patients had reflex disturbances more often. There was no correlation between preoperative duration and pain or neurologic disturbances: Patients with longer preoperative duration of symptoms did not demonstrate more severe symptoms. There was a total myelographic block of the spinal canal in 13 patients. The mean value of the antero-posterior diameter in the other patients was 6.8 mm (range, 4-11 mm). In patients younger than 70 years L4-L5 was the site for the most pronounced constriction, whereas L3-L4 was the narrowest site in the older patients. Degenerative spondylolisthesis was found in 32 patients, and they had a more pronounced constriction of the spinal canal (5.6 mm compared with 6.7 mm in those without displacement, P = 0.02). There was a (nonsignificant) tendency toward more walking disturbances in patients with a more pronounced constriction of the spine. There was no correlation between reflex disturbances or extensor hallucis longus weakness and radiographically detected constriction. CONCLUSION: Pain was more intense and positive straight leg raising test results were more common in younger patients, whereas reflex disturbances were more common in the elderly. The vertebral site for the lowest anteroposterior value was higher with higher age. Preoperative duration did not affect the severity of symptoms or signs. Patients with more pronounced stenosis tended to have a more severe reduction of walking ability. There was no correlation between symptoms and signs and radiographically detected constriction.

A prospective and consecutive study of surgically treated lumbar spinal stenosis. Part II: Five-year follow-up by an independent observer.

STUDY DESIGN: A prospective and consecutive study of surgical results obtained during serial follow-up investigations in patients who underwent surgery for central lumbar spinal stenosis. OBJECTIVES: To evaluate the result after surgical decompression for lumbar spinal stenosis, at regular intervals after surgery, and to correlate these results with values for preoperative parameters; special interest was focused on the results in relation to the degree of constriction of the spinal canal. SUMMARY OF BACKGROUND DATA: The outcome after surgery for spinal stenosis is debatable; long-term follow-up investigations have indicated deterioration with passing time. Results of studies in nonsurgical patients have demonstrated that the symptoms do not progress with time. Results of a meta-analysis of the literature on surgical results have demonstrated a wide variation of outcomes. MATERIAL AND METHODS: In a prospective study, 105 consecutive patients who underwent surgical decompression (laminectomy with facet-preserving technique, but no fusion) were evaluated at follow-up examinations 4 months and 1, 2, and 5 years after surgery. At the follow-up examinations, the patient's opinion on the surgical result was registered, using a four-grade scale. The occurrence of pain at rest and at night was registered, as well as the patient's walking ability. Statistical analysis was performed, relating the surgical results to patient age, gender, preoperative duration of symptoms and radiographically observed constriction as described in Part I of this study. The radiologist was blinded to patient outcome. Logistic regression analysis was performed. RESULTS: During the follow-up period, 19 patients underwent reoperation, consisting of fusion to treat lumbar pain (n = 4), repeat decompression because of progressive stenosis (n = 13), and repairs in response to surgical complications (n = 2). Follow-up results: The result, related to the recurrence of leg symptoms, deteriorated with passing time. Excellent results were reported by 63% to 67% at 4-month and 2-year follow-ups compared with 52% at the 5-year follow-up. There was a correlation between the constriction of the spinal canal and the outcome at all intervals. Patients with an anteroposterior diameter of 6 mm or less at the narrowest site had significantly better results. The logistic regression analysis demonstrated a significant correlation between a severe reduction of the anteroposterior diameter and excellent results and a tendency toward better results in patients with a shorter preoperative duration of symptoms. Improvement of walking ability was also associated with a pronounced constriction of the spinal canal. CONCLUSION: The results after surgical decompression in patients with central spinal stenosis deteriorated with time. There was a significant correlation between good result and pronounced constriction of the spinal canal. Patients with a preoperative duration of symptoms of less than 4 years and patients with no preoperative back pain tended to have better surgical outcomes. The reoperation rate was 18% within 5 years. When surgery for spinal stenosis is contemplated, these prognostic factors should be taken into consideration: The "ideal patient" has a pronounced constriction of the spinal canal, insignificant lower back pain, no concomitant disease affecting walking ability, and a symptom duration of less than 4 years.

The pragmatics of clinical hypermedia: experiences from 5 years of participatory design in the MEDEA project.

To explore a medical hypermedia design process where requirements engineering and participatory design are used in a naturalistic setting, 5 years of participatory action research (PAR) have been performed in the development of a hypermedia system to be used in primary care practice. In PAR, the participating organizations cooperate with the researchers in deciding and later evaluating organizational actions, and the results are concluded cooperatively. A new type of system emerged from the study, in that the resulting design has its focus on the management of large volumes of hypermedia material, the traceability of authority in the documents, and teamwork support. Regarding the participatory design process, it was found to be essential to deal with social and organizational issues in the design group and its surroundings--and not to cover these over. For instance, an agreement was reached with the managers responsible for health care in the county where the design process took place, whereby the design activity was also given priority at the highest local administrative level. Since health care is a practice organization, there is a potential market for similar systems if only ways of organizing design and marketing the products of participatory design are further developed. Modified versions of structured product specification methods could be a valuable complement. The conclusion of the study is that the development and spread of hypermedia systems in health care may require considerable changes in current design routines and organizations.

Computerized knowledge bases in primary health care: a curse or a blessing for health promotion, prevention and patient quality?.

This paper presents a future scenario analysis of how the introduction of computerized knowledge bases (KBs) can come to affect primary care practice. For the collection and analysis of data, a two-level video method was applied. First, four consultations where a computerized KB was used were video-recorded. A search workshop was then carried out by letting a multi-disciplinary panel comment on the video recordings. The comments were categorized with regard to content and perspective. Analyses of the comments showed a concern for a disregard of patients' health beliefs and for difficulties in portioning out the acquired medical knowledge to the patient during the consultation. Furthermore, the computerized KB was found to easily break the natural flow of the consultation and be perceived as a third party. The conclusion is that the most critical aspects for using computerized KBs in a reformed primary health care concern the integration of the systems into the consultation process. Health promotion, prevention, and patient quality are central here, and the introduction of KB technology must not lead the consultation away from these issues.

Participatory design of computer-supported organizational learning in health care: methods and experiences.

This paper outlines a Computer-Supported Co-operative Work (CSCW) system for primary care and presents from its participatory design process time consumption, costs, and experiences. The system integrates a hypermedia environment, a computerized patient record, and an electronicmessage system. It is developed to coordinate organizational learning in primary care from micro to macro levels by connecting strategic planning to monitoring of patient routines. Summing up design experiences, critical issues for making CSCW systems support cost-effectiveness of health care are discussed.

Inside multi-disciplinary design in medical informatics: experiences from the use of an argumentative design method.

This paper reports on a qualitative study using an argumentation-based design method (Argumentative Design) in the development of clinical software systems. The method, which requires visualization of the underlying design goals, the specific needs-for-change, and the probable consequences of the alternative design measures, caused previously implicit argument structures to be exposed and discussed. This uncovering of hidden agendas also revealed previously implicit coalitions and organizational influences on the design process. Implications for software development practices in medical informatics are discussed.

Distribution of adipose tissue and muscle mass in alcoholic men.

An elevated waist to hip ratio (WHR) has been found to be a predictor for several prevalent diseases. To examine the potential role of alcohol in the elevation of WHR, established alcoholic men without severe liver damage who were in adequate nutritional condition were compared with organized teetotalers matched for age, height, and body weight; the groups had similar total body fat content and lean body mass. Computed tomographic (CT) measurements at thigh and trunk levels showed a significant increase in the visceral adipose tissue (AT) areas and a slight decrease of muscle areas in the gluteal and femoral regions of the alcoholics. The alcoholic men had 48% of their AT areas of trunk scans localized retroperitoneally and intraperitoneally compared with 38% for the teetotalers (P < .01). The difference seemed to be more marked for retroperitoneal than for intraperitoneal AT (97 v 60 cm2, P < .01). The elevated visceral AT areas seemed to be independent of smoking. It was concluded that the increased WHR of alcoholics may include not only changes in AT, but also in muscle tissue distribution.

Developing a clinical hypermedia corpus: experiences from the use of a practice-centered method.

This paper outlines a practice-centered method for creation of a hypermedia corpus. It also describes experiences with creating such a corpus of information to support interprofessional work at a Primary Healthcare Center. From these experiences, a number of basic issues regarding information systems development within medical informatics will be discussed.

Alcohol consumption and synthesis of ethyl esters of fatty acids in adipose tissue.

Ethyl esters of fatty acids (EEFA) have been found to be formed during ethanol metabolism. Human adipose tissue contains high concentrations of free fatty acids, the substrate for EEFA synthesis, and might therefore be a tissue with great potential for EEFA formation. In order to explore their potential usefulness as markers of alcohol abuse, the EEFA concentration and the activity of EEFA-synthesizing enzyme were therefore determined in adipose tissue from men belonging to the following categories: teetotalers, social drinkers, alcoholics under treatment, or established alcoholics found to have died as a result of alcohol intoxication. In order to estimate the half-life of EEFA and the synthase activity induction, the alcoholics were examined after different time periods of abstinence from alcohol. Comparisons were also made with several established markers of alcohol abuse. EEFA were not found in teetotalers, and were found in low concentrations in some of the social drinkers. EEFA were found in several alcoholics, and the forensic cases had high concentrations. EEFA-synthesizing enzyme activity was found in all subjects, increasing from teetotalers to social drinkers, and being 2-fold higher in alcoholics and 5-fold higher in dead alcoholics. The induction of the enzyme after abstinence appeared to have a half-life of the order of several weeks. Correlations were found between EEFA synthase activity and previously established markers of alcohol abuse known to remain for a long time period after abstinence, such as mean erythrocyte corpuscular volume. This preliminary study suggests the possibility that EEFA synthase induction in adipose tissue might have a longer half-life than previously used markers of alcohol abuse. It is therefore suggested that the induction of EEFA synthase might be a potentially useful new marker for alcohol abuse because of its apparent proportionality to alcohol intake over a prolonged time period, its presumed specificity, and long-term elevation after alcohol abstinence. This potential marker should be analysed further.

Fat distribution and steroid hormones in women with alcohol abuse.

Anthropometric, hormonal and liver function parameters were examined in 18 premenopausal women with a history of early alcohol abuse, and compared with the data for randomly selected controls of the same age. The alcoholic women showed slightly elevated levels of transaminases, but no clinical or laboratory signs of advanced liver damage. These women were characterized by an increased waist-to-hip ratio, due to enlarged waist circumference. Several endocrine abnormalities were found, including irregular or absent menses as well as low oestrogen, progesterone and delta-4-androstendione levels. The concentration of free testosterone was high and that of sex-hormone-binding globulin was low. These data suggest abdominal distribution of body fat, as well as hyperandrogenicity in alcoholic, premenopausal women. It is postulated that the endocrine abnormalities might be responsible for the abdominal fat distribution.