Search

Improved oncological outcomes after cytoreductive surgery (CRS) with hyperthermic intraperitoneal chemotherapy (HIPEC) in highly selected patients have been well documented. The extensive nature of the procedure adversely affects quality of life (QoL). The aim of this study is to longitudinally evaluate QoL following CRS/HIPEC. This is a retrospective review of a prospectively maintained database of patients with peritoneal malignancies undergoing CRS/HIPEC. Clinicopathological data, oncologic outcomes, and QoL were analyzed preoperatively and postoperatively at 2 weeks, and 1, 3, 6, and 12 months. The Functional Assessment of Cancer Therapy-Colorectal instrument was used to determine changes in QoL after CRS/HIPEC and the impact of early recurrence (12 months) on QoL. Thirty-six patients underwent CRS/HIPEC over 36 months. The median peritoneal cancer index score was 18 and the completeness of cytoreduction-0/1 rate was 97.2 per cent. Postoperative major morbidity was 16.7 per cent with one perioperative death. Disease-free survival was 12.6 months in patients with high-grade tumors versus 31.0 months in those with low-grade tumors (P = 0.03). QoL decreased postoperatively and improved to baseline in six months. Patients with early recurrence had a decrease in global QoL compared with preoperative QoL at 6 (P0.03) and 12 months (P0.05). This correlation was not found in patients who had not recurred. Patients who undergo CRS/HIPEC have a decrease in QoL that plateaus in 3 to 6 months. Early recurrence adversely impacts QoL at 6 and 12 months. This study emphasizes the importance of patient selection for CRS/HIPEC. The expected QoL trajectory in patients at risk for early recurrence must be carefully weighed against the potential oncological benefit of CRS/HIPEC.

Gender alone offers no survival advantage in humans following trauma. However, male gender does predict increased morbidity, specifically ventilator-associated pneumonia (VAP). Previous work has shown that despite lower incidence of VAP, females with VAP have increased mortality. The purposes of this study were to evaluate the impact of VAP and gender on outcome and to determine which characteristics of severe VAP predict mortality in trauma patients.Patients with VAP (≥10 colony-forming units per milliliter in bronchoalveolar lavage) over 8 years were stratified by gender, age, severity of shock, and injury severity. Severe VAP factors were defined as multiple-episode, polymicrobial, multidrug-resistant, nosocomial VAP diagnosed within 7 days of admission (eNVAP), and multiple inadequate empiric antibiotic therapy episodes. Mortality and severe VAP factors were compared using χ analysis. Multivariable logistic regression (MLR) was performed to determine which VAP factors were independent predictors of mortality.A total of 854 patients were identified, 676 men (79%) and 178 women (21%). Despite a higher incidence of VAP among males (3.8% vs. 2.6%, p = 0.001), mortality was higher in females (24% vs. 15%, p = .009). All characteristics of severe VAP were increased in females except multiple episodes (p = 0.15). MLR identified eNVAP as an independent predictor of mortality in females with severe VAP after adjusting for age, Glasgow Coma Scale (GCS) score, Injury Severity Score (ISS), admission base excess, and 24-hour transfusions (odds ratio, 9.97; p = 0.001).That females develop less VAP but experience increased mortality confirms previous studies. Characteristics of severe VAP are increased in females and may contribute to this observed mortality difference. MLR identified eNVAP as an independent predictor of mortality in females with severe VAP following trauma.Epidemiologic/prognostic study, level III.

The 80-hour week was implemented in 2003 to improve outcomes and limit errors. We hypothesize that there has been no change in outcomes postimplementation of the restrictions. Outcomes were queried from the trauma registry from 1997 to 2002 (PRE) and 2004 to 2009 (POST). Primary outcomes were mortality, intensive care unit length of stay (ICU LOS), and length of stay (LOS). Patients were stratified based on demographics, blood pressure, heart rate, and injury severity (Injury Severity Score, Glasgow Coma Score, base deficit). Outcomes were then compared PRE with POST. A total of 41,770 patients were admitted during the study period. The mean age was 38 years with most being male (73%) and blunt mechanism (78%). Although patients admitted in the POST period had a slightly higher blood pressure, they were older and had higher injury severity. ICU LOS, LOS, self-pay, and mortality were higher in the POST period. After adjusted analysis, admission in the POST period was no longer a predictor of mortality (odds ratio, 1.02; confidence interval, 0.92 to 1.14). Whereas patients were more slightly more injured in the POST period, the adjusted analysis shows no difference in mortality and both a longer LOS and ICU LOS. Whether the increase is the result of more severe injury in the POST period or less efficient disposition remains to be elucidated. This study adds to the mounting evidence that the implementation of the limits on work hours does not lead to better outcomes.

Admission red cell distribution width (aRDW) has been shown to predict mortality in trauma patients by an unclear mechanism. It has been speculated that aRDW is a marker of chronic health status, but elevated RDW may also reflect recent hemorrhage. We hypothesized that aRDW is a predictor of major hemorrhage in trauma patients. Shock trauma patients at a Level I trauma center over 6.5 years were evaluated. Patients were stratified by aRDW quintile (Q1: less than 13%, Q2: 13.1 to 13.5%, Q3: 13.6 to 14.0%, Q4: 14.1 to 14.9%, Q5: 15.0% or greater). Massive transfusion (MT) was defined as 10 or more packed red blood cells in the first 24 hours. From multiple logistic regression, odds ratios with 95 per cent confidence intervals (CIs) were determined to evaluate the association between aRDW quintile and MT. Three thousand nine hundred ninety-four met study criteria. Overall MT incidence was 10 per cent and in-hospital mortality was 17 per cent. MT and mortality increased in a stepwise fashion by aRDW quintile ( P < 0.0001). From logistic regression, a threefold increased odds of MT was associated with aRDW Q4 (CI, 1.81 to 4.92), and a 3.5-fold increased odds of MT was associated with aRDW Q5 (CI, 2.70 to 5.83). aRDW independently predicted MT, suggesting that elevated aRDW is an indicator of major hemorrhage in trauma patients. The association between aRDW and mortality in trauma patients may be explained by acute hemorrhage rather than chronic health status.

Dysautonomia in traumatic brain injury patients may contribute to secondary injury. We hypothesize that propranolol is the best β-blocker (BB) to block the excess catecholamines and improve mortality in this patient population.Patients with traumatic brain injury admitted during a 48-month period who received BB were compared with those who did not after excluding patients who received preinjury BB, deaths within 48 hours, and head Abbreviated Injury Scale (AIS) score of less than 3 or greater than 5. In addition, propranolol was also compared with all other BBs.A total of 1,755 patients with traumatic brain injury were identified during the study period after exclusions. Patients who received BB (427) were older (49 years vs. 40 years; p0.0001), were more severely injured (Injury Severity Score [ISS], 30 vs. 24; p0.001), and had a more severe head injury (head AIS score, 4.2 vs. 4.0; p0.001). By univariate analysis, BB patients had a higher mortality (13% vs. 6%; p0.001); after adjusted analysis, no difference was identified (adjusted odds ratio, 0.850; 95% confidence interval, 0.536-1.348). Seventy-eight patients (18%) received propranolol during the study period. Propranolol patients were younger (30 years vs. 53 years; p0.001) but more severely injured (ISS, 33 vs. 29; p = 0.01; head AIS, 4.5 vs. 4.2; p0.001), with longer stay (44 days vs. 26 days, p0.001). Mortality was less in the propranolol group (3% vs. 15%, p = 0.002). Adjusted analysis confirmed the protective effect of propranolol (adjusted odds ratio, 0.199; 95% confidence interval, 0.043-0.920).Propranolol is the best BB to limit secondary injury and decrease mortality in patients with traumatic brain injury.Therapeutic, study level III.

Operative management at our institution for all colon injuries have followed a defined algorithm (ALG) based on risk factors originally identified for penetrating injuries. The purpose of this study was to evaluate the applicability of the ALG to blunt colon injuries.Patients with blunt colon injuries during 13 years were identified. As per the ALG, nondestructive (ND) injuries are treated with primary repair. Patients with destructive wounds (serosal tear of ≥50% colon circumference, mesenteric devascularization, and perforations) and concomitant risk factors (transfusion of6 U packed red blood cells and/or presence of significant comorbidities) are diverted, while patients with no risk factors undergo resection plus anastomosis (RA). Outcomes included suture line failure (SLF), abscess, and mortality. Stratification analysis was performed to determine additional risk factors in the management of blunt colon injuries.A total 151 patients were identified: 76 with destructive injuries and 75 with ND injuries. Of those with destructive injuries, 44 (59%) underwent RA and 29 (39%) underwent diversion. All ND injuries underwent primary repair. Adherence to the ALG was 95%: three patients with destructive injuries underwent primary repair, and five patients with risk factors underwent RA. There were three SLFs (2%) (one involved deviation from the ALG) and eight abscesses (5%). Colon-related mortality was 2.1%. Stratification analysis based on mesenteric involvement, degree of shock, and need for abbreviated laparotomy failed to identify additional risk factors for SLF following RA for blunt colon injuries.Adherence to an ALG, originally defined for penetrating colon injuries, simplified the management of blunt colon injuries. ND injuries should be primarily repaired. For destructive wounds, management based on a defined ALG achieves an acceptably low morbidity and mortality rate.Prognostic/epidemiologic study, level III; therapeutic study, level IV.

Background Traumatic disruption of the pelvic ring is a significant cause of life-threatening hemorrhage. For those patients who survive the initial injury, these fractures are associated with long periods of immobilization and intense rehabilitation. There is little published information available regarding long-term functional outcomes in these patients. This study evaluated the impact of severe pelvic fractures on those long-term outcomes. Methods All patients with severe pelvic fractures over an 18-year period were identified. Severe pelvic fractures were defined as those with vascular disruption, open-book component with symphysis diastasis, or sacroiliac disruption with vertical shear. Functional outcome was measured using the Boston University Activity Measure for Post-Acute Care to assess mobility (normal, >84) and daily activity (normal, >84). Multiple linear regression analysis was used to identify predictors of functional outcome after severe pelvic fracture. Results From January 1996 to September 2014, 401 patients were identified: 240 (60%) men and 161 (40%) women. Overall mortality was 29%. Of the 285 survivors, follow-up was obtained in 145 patients (51%). Mean follow-up was 8.3 years, with a maximum of 20 years. Mean age and Injury Severity Score were 53 years and 27, respectively. Mean Activity Measure for Post-Acute Care scores for mobility and daily activity were 55 and 63, respectively; both signifying significant impairment when compared with normal. Multiple linear regression analysis using age, traumatic brain injury, transfusions, Injury Severity Score, ventilator days, presence of associated lower-extremity fractures, and time to operative pelvic fixation identified time to pelvic fixation as the only predictor of decreased mobility (β = -0.74, p = 0.04) and activity (β = -0.21, p = 0.03) following severe pelvic fracture. Conclusions Prolonged time to operative pelvic fixation led to worse long-term functional outcomes in patients with severe pelvic ring disruption. Early fixation of the pelvic ring is the only potentially modifiable risk factor for decreased functional outcomes in patients with severe pelvic fractures. Level of evidence Therapeutic study, level IV.

Introduction The purpose of this study was to evaluate outcomes following pancreaticoduodenectomy(PD) for ampullary adenocarcinoma(AAC). Methods We evaluated patients having undergone PD for AAC and the impact of clinical/histopathologic factors and adjuvant therapy(AT) on survival. Results 52 patients underwent potentially curative PD. Perineural and lymphovascular invasion were associated with decreased survival. There was no difference in survival between patients treated with PD vs. PD+AT, however, AT was more often administered to patients with N1 vs. N0 and stage II/III vs. I disease. Among patients receiving AT, we observed a trend towards improved survival when radiation was included. Recurrence occurred in 7/18(39%) stage I patients, only 2(7%) of which received AT. Conclusion AT did not improve survival, however was more commonly administered in advanced disease. Stage I patients had high recurrence rates but rarely received AT. Prospective evaluation of appropriate AT regimens and use in early stage patients should be considered.

Catecholamine surge after traumatic injury may lead to dysautonomia with increased morbidity. Small retrospective studies have shown potential benefit of beta-blockers (BB) in trauma patients with and without traumatic brain injury (TBI). This study evaluates a large multiply injured cohort without TBI that received BB. Patients were identified from the trauma registry from January 1, 2003 to December 31, 2011. Patients who received1 dose of BB were compared to controls. Patients with TBI, length of stay (LOS)2 days, and prehospital BB were excluded. Outcomes were mortality, intensive care unit (ICU) LOS, and LOS. Stepwise multivariable regression was used to identify variables significantly associated with mortality. During the study period, 19,151 eligible patients were admitted. The mean age was 39 years. Most were male (74%) and most sustained blunt mechanism (75%). A total of 1854 (11%) patients received BB. BB patients had longer LOS (16 vs 6 days), ICU LOS (7 vs 1 days), and higher mortality (2.8 vs 0.5%) (all P0.001). Multivariable regression demonstrated no benefit to BB after adjusting for potential confounding characteristics [odds ratio (OR) 0.952; confidence interval (CI) 0.620-1.461]. In conclusion, in this largest study to date, patients receiving BB were older, more severely injured, and had a higher mortality. Unlike TBI patients, multivariable regression showed no benefit from BB in this population.

Background Our previous experience with colon injuries suggested that operative decisions based on a defined algorithm improve outcomes. The purpose of this study was to evaluate the validity of this algorithm in the face of an increased incidence of destructive injuries observed in recent years. Study Design Consecutive patients with full-thickness penetrating colon injuries over an 8-year period were evaluated. Per algorithm, patients with nondestructive injuries underwent primary repair. Those with destructive wounds underwent resection plus anastomosis in the absence of comorbidities or large pre- or intraoperative transfusion requirements (more than 6 units packed RBCs); otherwise they were diverted. Outcomes from the current study (CS group) were compared with those from the previous study (PS group). Results There were 252 patients who had full-thickness penetrating colon injuries: 150 (60%) patients had nondestructive colon wounds treated with primary repair and 102 patients (40%) had destructive wounds (CS). Demographics and intraoperative transfusions were similar between CS and PS groups. Of the 102 patients with destructive injuries, 75% underwent resection plus anastomosis and 25% underwent diversion. Despite more destructive injuries managed in the CS group (41% vs 27%), abscess rate (18% vs 27%) and colon-related mortality (1% vs 5%) were lower in the CS. Suture line failure was similar in CS compared with PS (5% vs 7%). Adherence to the algorithm was >90% in the CS (similar to PS). Conclusions Despite an increase in the incidence of destructive colon injuries, our management algorithm remains valid. Destructive injuries associated with pre- or intraoperative transfusion requirements of more than 6 units packed RBCs and/or significant comorbidities are best managed with diversion. By managing the majority of other destructive injuries with resection plus anastomosis, acceptably low morbidity and mortality can be achieved.

Background The optimal management of pancreatic injuries, specifically with respect to defining ductal integrity, remains controversial. Our previous experience suggested that decisions based on probability of ductal injury might improve outcome. Consequently, a management algorithm (ALG) was developed and implemented. The purpose of this study was to evaluate the impact of this ALG on outcomes. Methods Consecutive patients more than 13 years with pancreatic injuries subsequent to the development of the ALG were evaluated. Pancreatic injuries were defined as proximal or distal and ductal injuries classified as definite, high, low, or indeterminate (IND) probability. Pancreas-related morbidity (fistula, abscess, and pseudocyst) and mortality were recorded. Patients managed by the ALG were compared with the previous study (PS). Results In all, 245 patients were identified; 35 died within 12 hours and were excluded. Demographics and severity of shock (24-hour transfusions) were similar between groups. Pancreas-related morbidity for proximal injuries was 13.8% in the ALG group and 13.5% in the PS (p = 0.948). Pancreas-related morbidity was significantly reduced in the ALG group for distal injuries requiring drainage alone (11% vs. 25%, p = 0.05) and for distal injuries requiring resection + drainage (26% vs. 58%, p = 0.003) when compared with the PS. There was no pancreas-related mortality in the ALG group (1.6% in the PS group, p = 0.065). Conclusions Adherence to a defined ALG simplified the management of traumatic pancreatic injuries and contributed to a reduction in both pancreas-related morbidity and mortality. The majority of all proximal pancreatic injuries can be treated with drainage alone. For distal injuries, a clinical decision based on defined parameters and suspicion of ductal injury dictates definitive management.

Placement of a halo vest for cervical spine fractures is presumed to be less morbid than operative fixation. However, restrictions imposed by the halo vest can be detrimental, especially in older patients. The purpose of this study was to evaluate the impact of halo vest placement on outcomes by age in patients with cervical spine fractures without spinal cord injury.All patients with blunt cervical spine fractures managed over an 18-year period were identified. Those with spinal cord injury and severe traumatic brain injury were excluded. Patients were stratified by age, sex, halo vest, injury severity, and severity of shock. Outcomes included intensive care unit length of stay, ventilator days, ventilator-associated pneumonia, functional status, and mortality. Multivariable logistic regression was performed to determine whether halo vest was an independent predictor of mortality in older patients.A total of 3,457 patients were identified: 69% were male, with a mean Injury Severity Score (ISS) and Glasgow Coma Scale (GCS) score of 19 and 13, respectively. Overall mortality was 5.3%. One hundred seventy-nine patients were managed with a halo vest, 133 of those 54 years and older and 46 of those younger than 54 years. Both mortality (13% vs. 0%, p0.001) and intensive care unit length of stay (4 days vs. 2 days, p = 0.02) were significantly increased in older patients despite less severe injury (admission GCS score of 15 vs. 14 and ISS of 14 vs. 17, p = 0.03). Multivariable logistic regression identified halo vest as an independent predictor of mortality after adjusting for injury severity and severity of shock (odds ratio, 2.629; 95% confidence interval, 1.056-6.543) in older patients.The potential risk of operative stabilization must be weighed against that of halo vest placement for older patients with cervical spine fractures following blunt trauma. Patient age should be strongly considered before placement of a halo vest for cervical spine stabilization.Therapeutic study, level IV.

Work-hour restrictions were amended in 2011 to limit interns to 16 continuous duty hours, essentially requiring a night float system of 12-hour shifts. We hypothesize that there has been no improvement in outcomes after implementation of the amended work-hour restrictions. Outcomes from trauma admissions were queried from the trauma registry from 2009 to 2011 (PRE) and 2011 to 2013 (POST). The primary outcome was mortality with secondary outcomes intensive care unit length of stay (LOS)and LOS. Patients were stratified based on age, mechanism, gender, blood pressure, heart rate, and injury severity (Injury Severity Score, Glasgow Coma Scale, Base Deficit). Outcomes were then compared from admissions PRE to POST. A total of 9178 patients were included in the study population. The mean age was 42 with most being male (72%) and blunt mechanism (81%). Patient populations were well matched except patients in the POST period were slightly older (43 vs 42 years; P = 0.01). Intensive care unit LOS and LOS were higher in the POST period. After adjusted analysis, admission in the POST period was not a predictor of mortality (odds ratio 0.857; confidence interval 0.655-1.12). The POST period was an independent predictor for LOS (β = 0.74; P = 0.002). This study adds to the mounting evidence that the implementation of the amended limits on work hours leads to furthermore decreased efficiency of care.

The strong association between penetrating colon injuries and intra-abdominal abscess (IAA) formation is well established and attributed to high colon bacterial counts. Since trauma patients are rarely fasting at injury, stomach and small bowel colony counts are also elevated. We hypothesized that there is a synergistic effect of increased IAA formation with concomitant stomach and/or colon injuries when compared to small bowel injuries alone. Consecutive patients at a level one trauma center with penetrating small bowel (SB), stomach (S), and/or colon (C) injuries from 1996 to 2012 were reviewed. Logistic regression determined associations with IAA, adjusting for age, gender, Injury Severity Score (ISS), admission Glasgow Coma Score, transfusions, and concurrent pancreas or liver injury. A total of 1518 patients (91% male, ISS = 15.9 ± 8.4) were identified: 496 (33%) SB, 231 (15%) S, 288 (19%) C, 40 (3%) S + SB, 69 (5%) S + C, 338 (22%) C + SB, and 56 (4%) S + C + SB. 148 (10%) patients developed IAA: 4 per cent SB, 9 per cent S, 10 per cent C, 5 per cent S + SB, 22 per cent S + C, 13 per cent C + SB, and 25 per cent S + C + SB. Multiple logistic regression demonstrated that ISS, 24 hour blood transfusions, and concomitant pancreatic or liver injuries were associated with IAA. Compared with reference SB, S or S + SB injuries were no more likely to develop IAA. However, S + C, SB + C, and S + C + SB injuries were significantly more likely to have IAA. In conclusion, combined stomach + colon, small bowel + colon, and stomach, colon, + small bowel injuries have a synergistic effect leading to increased IAA formation after penetrating injuries. Heightened clinical suspicion for IAA formation is necessary in these combined hollow viscus injury patients.

Background Outcome measures after cytoreductive surgery with hyperthermic intraperitoneal chemotherapy (CRS/HIPEC) for peritoneal carcinomatosis in established centers are well defined. However, results from newly emerging US centers have not been reported. Methods This is a retrospective review of a prospectively maintained database of patients with peritoneal malignancies undergoing CRS/HIPEC. Results Fifty-six patients underwent exploratory laparotomy with 36 receiving CRS/HIPEC over 36 months. The median peritoneal cancer index score was 18, and the cytoreduction 0/1 rate was 92%. Postoperative major morbidity was 16.7% with one perioperative death. The median length of hospital stay and intensive care unit days were 9 and 3 days, respectively. Disease-free survival in high-grade vs low-grade tumors was 12.6 and 31.0 months ( P , .03), respectively. Average direct cost for patients undergoing CRS/HIPEC was $25,917. Conclusions Our emerging center's short-term results are comparable with established programs with a trend toward more selective intraoperative judgment on who undergoes CRS/HIPEC.

Blunt abdominal trauma can result in injury to genitourinary (GU) organs. Children may be more susceptible to some GU injuries due to anatomic differences compared to adults. Mechanism of injury (MOI) has been thought to relate to both the likelihood and severity of GU injury in children, although this has not definitively been proven. Our purpose was to determine if MOI has any correlation to the severity of GU injury in children treated at our institution.We reviewed records of all pediatric blunt trauma patients presenting to our institution from January 2005-December 2010 using the LeBonheur Children's Hospital Trauma Registry. All patients with GU injuries were included in this study. Data collected included demographic information, MOI, type and grade of GU injury, associated injuries, and clinical outcome. Continuous variables were tested with ANOVA and categorical variables were tested with chi-square test.Records of 5151 children with blunt trauma were reviewed; 76 patients were found to have GU organ injury. There were 47 males (61.8%) and 29 females (38.2%). Categories of MOI included motor vehicle accident, sports injury, bicycle accident, all-terrain vehicle accident (ATV), pedestrian struck accident, falls, and animal injury. MOI did not have any statistically significant association with the severity of GU organ injury (p = 0.5159). In addition, there was no association between MOI and either gender or side of injury. There was a statistically significant association between MOI and patient age (p = 0.04); older pediatric patients were more likely to experience GU injury due to sports injury and ATV accidents, where as younger patients were more likely to experience GU injury due to pedestrian struck, bicycle accidents or animal bite.Although specific MOI would seem to relate to presence and severity of injury in children, MOI alone does not correlate with the severity of GU organ injury in our pediatric trauma population. Age of pediatric patients is associated with the type of MOI that results in GU organ injury. The possibility of GU injury should be considered in all symptomatic pediatric patients with clinically significant blunt trauma regardless of the exact MOI.

BACKGROUND Current direction in trauma resuscitation includes emphasis on minimizing crystalloid, along with early transfusion of blood products. Although evidence suggests that higher crystalloid volume during the first 24 hours is associated with negative outcomes, the effect of crystalloid administration during initial resuscitation remains unclear. The purpose of this study was to evaluate the impact of the ratio of crystalloid to packed red blood cells (C/PRBCs) infused during initial emergency department resuscitation on pulmonary morbidity and mortality. METHODS Over 6.5 years at a Level 1 trauma center, prospective data were collected on patients that received more than 1 PRBC unit in the resuscitation room. C/PRBC was defined as the ratio of crystalloid infused in liters to the units of PRBCs transfused in the resuscitation room. Patients were stratified by high ratio (>0.75) versus low ratio (

The 80-hour week was implemented in 2003 to improve outcomes and limit errors. We hypothesize that there has been no change in outcomes postimplementation of the restrictions. Outcomes were queried from the trauma registry from 1997 to 2002 (PRE) and 2004 to 2009 (POST). Primary outcomes were mortality, intensive care unit length of stay (ICU LOS), and length of stay (LOS). Patients were stratified based on demographics, blood pressure, heart rate, and injury severity (Injury Severity Score, Glasgow Coma Score, base deficit). Outcomes were then compared PRE with POST. A total of 41,770 patients were admitted during the study period. The mean age was 38 years with most being male (73%) and blunt mechanism (78%). Although patients admitted in the POST period had a slightly higher blood pressure, they were older and had higher injury severity. ICU LOS, LOS, self-pay, and mortality were higher in the POST period. After adjusted analysis, admission in the POST period was no longer a predictor of mortality (odds ratio, 1.02; confidence interval, 0.92 to 1.14). Whereas patients were more slightly more injured in the POST period, the adjusted analysis shows no difference in mortality and both a longer LOS and ICU LOS. Whether the increase is the result of more severe injury in the POST period or less efficient disposition remains to be elucidated. This study adds to the mounting evidence that the implementation of the limits on work hours does not lead to better outcomes.

Admission red cell distribution width (aRDW) has been shown to predict mortality in trauma patients by an unclear mechanism. It has been speculated that aRDW is a marker of chronic health status, but elevated RDW may also reflect recent hemorrhage. We hypothesized that aRDW is a predictor of major hemorrhage in trauma patients. Shock trauma patients at a Level I trauma center over 6.5 years were evaluated. Patients were stratified by aRDW quintile (Q1: less than 13%, Q2: 13.1 to 13.5%, Q3: 13.6 to 14.0%, Q4: 14.1 to 14.9%, Q5: 15.0% or greater). Massive transfusion (MT) was defined as 10 or more packed red blood cells in the first 24 hours. From multiple logistic regression, odds ratios with 95 per cent confidence intervals (CIs) were determined to evaluate the association between aRDW quintile and MT. Three thousand nine hundred ninety-four met study criteria. Overall MT incidence was 10 per cent and in-hospital mortality was 17 per cent. MT and mortality increased in a stepwise fashion by aRDW quintile (P0.0001). From logistic regression, a threefold increased odds of MT was associated with aRDW Q4 (CI, 1.81 to 4.92), and a 3.5-fold increased odds of MT was associated with aRDW Q5 (CI, 2.70 to 5.83). aRDW independently predicted MT, suggesting that elevated aRDW is an indicator of major hemorrhage in trauma patients. The association between aRDW and mortality in trauma patients may be explained by acute hemorrhage rather than chronic health status.

We describe the implementation of a new laboratory-based interdisciplinary undergraduate course on nonlinear dynamical systems. Geometrical methods and data visualization techniques are especially emphasized. A novel feature of the course is a required laboratory where the students analyze the behavior of a number of dynamical systems. Most of the laboratory experiments can be economically implemented using equipment available in many introductory physics microcomputer-based laboratories.

Background For more than a decade, operative decisions (resection plus anastomosis vs diversion) for colon injuries, at our institution, have followed a defined management algorithm based on established risk factors (pre- or intraoperative transfusion requirements of more than 6 units packed RBCs and/or presence of significant comorbid diseases). However, this management algorithm was originally developed for patients managed with a single laparotomy. The purpose of this study was to evaluate the applicability of this algorithm to destructive colon injuries after abbreviated laparotomy (AL) and to determine whether additional risk factors should be considered. Study Design Consecutive patients over a 17-year period with colon injuries after AL were identified. Nondestructive injuries were managed with primary repair. Destructive wounds were resected at the initial laparotomy followed by either a staged diversion (SD) or a delayed anastomosis (DA) at the subsequent exploration. Outcomes were evaluated to identify additional risk factors in the setting of AL. Results We identified 149 patients: 33 (22%) patients underwent primary repair at initial exploration, 42 (28%) underwent DA, and 72 (49%) had SD. Two (1%) patients died before re-exploration. Of those undergoing DA, 23 (55%) patients were managed according to the algorithm and 19 (45%) were not. Adherence to the algorithm resulted in lower rates of suture line failure (4% vs 32%, p=0.03) and colon-related morbidity (22% vs 58%, p=0.03) for patients undergoing DA. No additional specific risk factors for suture line failure after DA were identified. Conclusions Adherence to an established algorithm, originally defined for destructive colon injuries after single laparotomy, is likewise efficacious for the management of these injuries in the setting of AL.

BACKGROUND: The National Institute of Medicine’s report Hospital-Based Emergency Care: At the Breaking Point highlighted the critical issue of emergency department overcrowding. At our institution, patients with anterior abdominal stab wounds (AASW) have been managed with a protocol that uses diagnostic laparoscopy (DL) after positive result on local wound exploration. Patients with negative DL result are eligible for discharge directly from the recovery room. The purpose of this study was to evaluate the use of DL for AASWs in light of the recommendations that suggested serial abdominal examination (SAE) is preferred to determine the need for laparotomy. METHODS: Patients admitted to a Level 1 trauma center from January 2010 through August 2012 with AASWs were included (contemporary period to Western Trauma Association study). Information regarding baseline characteristics, diagnostic workup, injury management, and outcomes were retrospectively reviewed and compared with the SAE AASW algorithm. RESULTS: A total of 158 patients with AASWs were evaluated using our institutional algorithm. Thirty-eight patients (24%) went directly to the operating room for peritonitis, shock, or evisceration; 120 underwent local wound exploration; 99 had positive result (82%). Twenty-eight patients had immediate laparotomy owing to worsening clinical examination findings. Seventy had DL, and 19 of these patients were discharged home from the recovery room, with a mean length of stay of 6.4 hours. When comparing patients managed using the DL algorithm to those managed using the SAE-based algorithm, the nontherapeutic laparotomy rate was lower, although not statistically significant. However, the DL algorithm produced a significantly higher percentage of patients discharged directly home following local wound exploration. CONCLUSION: With some trauma centers suffering from emergency department overcrowding and constrained resources, DL may offer an alternative to SAE to efficiently use available resources. Both SAE and DL are safe and offer similar therapeutic laparotomy rates. The method used to evaluate patients after AASW should be tailored to institutional needs and resources. (J Trauma Acute Care Surg. 2013;75: 1013Y1018. Copyright * 2013 by Lippincott Williams & Wilkins) LEVEL OF EVIDENCE: Therapeutic/care management, level IV.

Ventilator-associated pneumonia (VAP) accounts for almost 90% of infections in mechanically ventilated patients and more than one-quarter of all patients requiring intubation, with associated mortality rates as high as 70%. The rise in methicillin resistance within the community has led to a national increase in methicillin-resistant Staphylococcus aureus (MRSA) rates in early VAP and associated healthcare expenditure.Trauma patients identified via an institutional VAP database were stratified by gender, age, severity of shock (24-h transfusions), and severity of injury. The primary outcome measure was evaluation of the incidence and trend of early MRSA VAP over a 6-y period. Secondary outcomes examined the adequacy of our current empiric antibiotic regimen as it pertained to outcome variables, including mortality.A total of 997 episodes of VAP were identified in 727 patients. Linear regression showed that the incidence of early Staphylococcus aureus (SA) VAP was stable over the 6-y period (slope=-0.911; p=0.490). Over the same 6 y, however, the percentage of MRSA in early SA VAP (slope=3.95; p=0.0154) and the incidence of early MRSA VAP increased. No difference in mortality was detected between early methicillin-susceptible SA and early MRSA VAP. After adjustment for age, ISS, and 24-h transfusion requirements, early MRSA was not an independent predictor of mortality (odds ratio [OR], 0.815; p=0.59).Although the incidence of early SA VAP with methicillin resistance increased significantly within the first 7 d of admission, this study showed no difference in mortality and resource utilization between early VAP from MRSA and other causative organisms, despite lack of empiric MRSA coverage.

BACKGROUND An unresolved concern regarding resuscitation in the setting of massive hemorrhage is potential lung injury from the transfusion of relatively more plasma-rich components. However, the association between plasma-to-packed red blood cell (PRBC) ratio and subsequent pulmonary dysfunction remains unclear. The purpose of this study was to evaluate the impact of plasma/PRBC on PaO2-to-FIO2 (P/F) ratio in the setting of massive transfusion (MT). METHODS During a 5.5-year period, prospective data were collected on trauma patients who underwent MT, defined as 10 or more units of PRBC transfusion by completion of hemorrhage control. Deaths within 48 hours of arrival were excluded. Acute lung injury (ALI) and adult respiratory distress syndrome (ARDS) were defined as P/F ratio of less than 300 and less than 200 at 48 hours, respectively. Stepwise multiple regression analysis was performed to determine variables significantly associated with P/F ratio. RESULTS A total of 199 patients met inclusion criteria; 159 (80%) developed ALI, and 105 (53%) developed ARDS. ALI and ARDS were both associated with subsequent mortality: ARDS at 24% versus no ARDS at 10% (p < 0.05) and ALI at 21% versus no ALI at 2.5% (p < 0.05). Paradoxically, patients with P/F ratio of 300 or greater were found to have received more plasma (5.6 U vs. 4.3 U, p < 0.05) and higher plasma-to-PRBC ratio (1:2 vs. 1:3, p < 0.05) at completion of hemorrhage control. Stepwise multiple regression analysis, however, identified age (p < 0.001) and chest Abbreviated Injury Scale (AIS) score (p = 0.04), but not plasma/PRBC (p = 0.10), to be independent determinants of P/F ratio at 48 hours. CONCLUSION In this cohort of MT patients who survived beyond the first 48 hours, pulmonary dysfunction developed in the majority and was associated with a 10-fold higher risk of subsequent death. However, plasma-to-RBC ratio achieved during hemorrhage control had neither a positive nor a negative impact on subsequent P/F ratio. In fact, only unalterable patient factors including age and severity of thoracic injury were associated with subsequent P/F ratio. LEVEL OF EVIDENCE Prognostic study, level III.

The phenomenon of stochastic resonance is predicted and experimentally verified using an amorphous silicon/ferroelectric liquid crystal optically addressed spatial light modulator. We exploit the bistability of aligned ferroelectric liquid crystal for use as the nonlinear element with the noise induced switching intensity set by an optical beam. Readout of the bistable system is also optical. Our measurements show the characteristic improvement in the output signal-to-noise ratio with increasing input noise intensity and the results agree qualitatively with both computer simulations and the predictions of a standard theory of stochastic resonance.

Background To date, work-hour restrictions have not been imposed on attending surgeons in the United States. The purpose of this study was to investigate the impact of working an overnight trauma shift on outcomes of general surgery operations performed the next day by the post-call attending physician. Study Design Consecutive patients over a 3.5-year period undergoing elective general surgical procedures were reviewed. Procedures were limited to hernia repairs (inguinal and ventral), cholecystectomies, and intestinal operations. Any operations that were performed the day after the attending surgeon had taken an overnight trauma shift were considered post-call (PC) cases; all other cases were considered nonpost-call (NP). Outcomes from the PC operations were compared with those from the NP operations. Results There were 869 patients identified; 132 operations were performed PC and 737 were NP. The majority of operations included hernia repairs (46%), followed by cholecystectomies (35%), and intestinal procedures (19%). Overall, the PC operations did not differ from the NP operations with respect to complication rate (13.7% vs 13.5%, p = 0.93) or readmission within 30 days (5% vs 6%, p = 0.84). Additionally, multivariable logistic regression failed to identify an association between PC operations and the development of adverse outcomes. Follow-up was obtained for an average of 3 months. Conclusions Performance of general surgery operations the day after an overnight in-hospital trauma shift did not affect complication rates or readmission rates. At this time, there is no compelling evidence to mandate work-hour restrictions for attending general surgeons.

Propofol infusion syndrome (PIS) is defined by arrhythmia, rhabdomyolysis, lactic acidosis, and unrecognized leads to death. We sought to determine the incidence of PIS in trauma patients and evaluate the efficacy of a prospective screening protocol in this patient population. Materials and methods In Phase I of the before-and-after study (1st January, 2005–31st December, 2005), trauma patients who received propofol were evaluated. Records were reviewed for demographics, injury severity, propofol time, dose, and rates, laboratory values, and adverse events. Patients were identified with PIS based on two of the following criteria: (1) cardiac arrhythmia/collapse, (2) metabolic acidosis, (3) rhabdomyolysis, and (4) acute kidney injury. Phase II (1st January, 2006–31st December, 2011) consisted of a prospective screening protocol (elevated lactate or creatine phosphokinase (CPK)) to identify patients at risk for PIS. Results 207 patients were identified in Phase I. 6 (2.9%) developed PIS with a 50% mortality. No differences were seen in age, gender, or mechanism. PIS patients were more injured (median ISS 44 vs 26, p = 0.04; median head AIS 5 vs 4, p = 0.003) and received more propofol (median 50,350 vs 9770 mg, p = 0.001) with longer infusion times (413 vs 65 h, p = 0.001). Sodium, creatinine, and CPK levels were higher in those that developed PIS (160 vs 145 mmol/L, p = 0.001; 4.3 vs 1.1 mg/dL, p = 0.005; 59,871 vs 520 U/L; p = 0.002). Pre-screening PIS incidence was 2.9% (6/207), but after screening (January 2006) the incidence dropped to 0.19% (2/1038, p Conclusions PIS is a morbid and lethal entity associated with sedation of critically injured patients. A simple screening procedure utilizing serum CPK (

BACKGROUND In clinical research, massive transfusion (MT) is commonly defined as transfusion of 10 or more red blood cell (RBC) units within 24 hours. However, the clinical relevance of this definition remains poorly understood. In this study, we evaluated whether patients who reach the MT threshold during hemorrhage control differ clinically from those who reach it after hemorrhage control (i.e., after intensive care unit [ICU] arrival) but before 24 hours. METHODS Prospective data were collected on all Level I trauma resuscitations within 5.5 years. Patients transfused 10 or more RBCs in the first 24 hours of hospitalization were identified and stratified according to when the MT threshold was achieved: before ICU arrival (Pre-ICU) versus after ICU arrival but before 24 hours of hospitalization (Post-ICU). Clinical characteristics between groups were compared. RESULTS Three hundred five patients received 10 or more units before ICU arrival, and 46 reached the MT threshold after ICU arrival but before 24 hours. Both groups were clinically similar with respect to age, sex, and Injury Severity Score, but the Post-ICU group had a larger proportion of blunt injuries (71 vs. 53%, p < 0.05), lower mean admission lactate (5.9 vs. 8.1 mmol/L, p < 0.05), and higher systolic blood pressure (112 vs. 96 mm Hg, p < 0.05) compared with the Pre-ICU group. Twenty-four-hour mortality was significantly lower in the Post-ICU group compared with the Pre-ICU group (9 vs. 33%, p < 0.05). In-hospital mortality was not significantly different between groups (33 vs. 46%, p = 0.11). CONCLUSION Patients reaching the MT threshold after ICU arrival comprise a relatively small proportion of those that would be included by the traditional MT definition. However, they have a significantly decreased mortality risk at 24 hours and the potential to dilute the study cohort. For research purposes, restricting the MT definition to 10 or more RBCs during hemorrhage control may result in study cohorts with relatively more uniform mortality risks. LEVEL OF EVIDENCE Prognostic study, level II.

Most studies examining suture line failure after penetrating colon injuries have focused on right- versus left-sided injuries. In our institution, operative decisions (resection plus anastomosis vs. diversion) are based on a defined management algorithm regardless of injury location. The purpose of this study was to evaluate the effect of injury location on outcomes after penetrating colon injuries.Consecutive patients with full thickness penetrating colon injuries for 13 years were stratified by age, injury location and mechanism, and severity of shock. According to the algorithm, patients with nondestructive injuries underwent primary repair. Destructive wounds underwent resection plus anastomosis in the absence of comorbidities or large preoperative or intraoperative transfusion requirements (6 U of packed red blood cells); otherwise, they were diverted. Injury location was defined as ascending, transverse, descending (including splenic flexure), and sigmoid. Multivariable logistic regression was performed to determine whether injury location was an independent predictor of either morbidity or mortality.Four hundred sixty-nine patients were identified: 314 (67%) underwent primary repair and 155 (33%) underwent resection. Most injuries involved the transverse colon (39%), followed by the ascending colon (26%), the descending colon (21%), and the sigmoid colon (14%). Overall, there were 13 suture line failures (3%) and 72 abscesses (15%). Most suture line failures involved injuries to the descending colon (p = 0.06), whereas most abscesses followed injuries to the ascending colon (p = 0.37). Multivariable logistic regression failed to identify injury location as an independent predictor of either morbidity or mortality after adjusting for 24-hour transfusions, base excess, shock index, injury mechanism, and operative management.Injury location did not affect morbidity or mortality after penetrating colon injuries. Nondestructive injuries should be primarily repaired. For destructive injuries, operative decisions based on a defined algorithm rather than injury location achieves an acceptably low morbidity and mortality rate and simplifies management.Prognostic study, level III.

BACKGROUND Among patients subjected to massive transfusion (MT), some will require considerably more blood than others, depending on the rate and quantity of hemorrhage. In analyses concerning plasma to red blood cell (RBC) ratios and platelet to RBC ratios, this has yet to be examined. We sought to evaluate the effect of the number of RBC units transfused on both plasma:RBC and platelet:RBC and their association with mortality in MT patients. METHODS Prospective data were collected on trauma patients taken directly to surgery from the resuscitation room who received ≥ 10 RBC units by completion of operation. MT protocol was in place for all patients. To account for survival bias, intra-operative deaths were excluded. Patients were stratified by plasma:RBC and platelet:RBC (HIGH > 0.5, MID 0.33-0.5, LOW < 0.33). Crude and adjusted risk ratios (RRs) for hospital mortality were determined, using the HIGH ratio as the reference group. RESULTS One hundred thirty-five patients met inclusion criteria. There were no significant differences with respect to demographics, injury characteristics, or shock severity. However, the mean number of intra-operative RBC units transfused was significantly different between plasma:RBC groups (HIGH: 16.2, MID: 19.7, LOW: 25.1; p < 0.001). The crude risk for mortality was significantly higher for the LOW group relative to the HIGH group (RR 1.99, 95% confidence interval [CI] 1.02-3.89). However, after adjustment for the number of RBCs transfused, the risk was not significantly different (RR 1.54, 95% CI 0.75-3.15). The adjusted mortality risk for the LOW versus HIGH platelet:RBC groups was also not statistically different (RR 1.92, 95% CI 0.99-3.71). CONCLUSIONS Among patients subjected to MT, those who receive relatively higher quantities of RBCs are both more likely to receive a lower plasma:RBC and are more likely to die. Any analysis concerning transfusion ratios should take the potential confounding of this heterogeneity among MT patients into account. LEVEL OF EVIDENCE Prognostic study, level III.

The application of a bistable, binary, amorphous silicon ferroelectric liquid crystal optically addressed spatial light modulator to speckle photography is described. The system presented has a higher space-bandwidth product than a television system and can operate at greater than 500 Hz frame rates. The output is also amenable to further optical processing. The spatial light modulator is characterised for sensitivity, contrast ratio and resolution and speckle correlation fringes from a moving object are successfully obtained. Measurement of displacements in the range 20 μm to 70 μm with an rms error of 3 μm is demonstrated and the suitability of these devices to practical optical metrology is discussed.

A set of methods has been developed to study the adhesion between four Lactobacillus reuteri strains and the milk fat globule membrane (MFGM) components in dairy products. By combining sucrose density gradient (SDG) centrifugation and bacterial DNA quantification it was found which strains of L. reuteri were more strongly associated with the dairy products, and the results were corroborated by direct binding rate and force measurements made with optical tweezers. It was determined that strong binding was associated with hydrophobicity of the bacteria and that this hydrophobicity is correlated with the presence of LiCl-extractable protein on the surface of the bacteria. Confocal laser scanning microscopy (CLSM) allowed for the visualization of interactions between bacteria and MFGM. This study demonstrates that these methods can be used in combination to characterize, both qualitatively and quantitatively, the adhesion of lactic acid bacteria strains in dairy products.

Background Although damage control strategies and the open abdomen have improved survival, they present their own unique set of challenges in caring for the multiply injured trauma patient. We previously reported the technique of staged abdominal wall closure for the management of the open abdomen. The purpose of this study was to evaluate the efficacy of various techniques of abdominal wall reconstruction (final stage of management) on long-term outcomes after planned ventral hernia, and to better define risk factors for recurrence. Study Design Patients undergoing abdominal wall reconstruction over a 15-year period were identified and stratified by gender, age, severity of shock, injury severity, and method of repair: secondary fascial closure ± prosthetic, standard components separation (SCS) ± prosthetic and modified components separation (MCS) ± prosthetic. Long-term outcomes (recurrence) were determined using hospital records, telephone interview, and physical examination. Multivariable logistic regression analysis was performed to determine independent predictors of recurrence. Results One hundred fifty-two patients were identified. Fourteen (9%) patients underwent secondary fascial closure ± prosthetic, 47 (31%) underwent SCS ± prosthetic, and 91 (60%) underwent MCS ± prosthetic. Long-term follow-up (up to 14.6 years, mean 5.3 years) was obtained in 114 (75%) patients. Sixteen patients (14%) had a recurrence. Prosthetic use increased recurrence 4-fold. There were 2 known recurrences (5%) in patients with MCS without prosthetic. Logistic regression identified both female gender and body mass index as independent predictors of recurrence. Conclusions The MCS technique is the procedure of choice for repair of giant abdominal wall defects. This approach can avoid the need for prosthetics. In fact, MCS without prosthetic resulted in an acceptably low hernia recurrence rate (5%).

We present the design, fabrication, and testing of a novel liquid-crystal-on-silicon optically addressed spatial light modulator for use as a weight matrix in an ART-1 optoelectronic neural processor. Each pixel in the 50 x 83 element array occupies 75 microm x 75 microm and consists of a photodetector, a threshold circuit, a 1-bit (flip-flop) memory element, and a liquid-crystal modulating mirror. The array is designed to switch all the pixels initially to the ON state. Subsequently each pixel is independently switched to the OFF state if a superthreshold amount of light falls upon the pixel's photodetector. The device has a contrast ratio of 20:1, a switch-on time (10-90% rise time) of 500 micros, and a switch-off time of ~500 micros (depending on the externally set threshold). Measured device uniformities and interpixel coupling are also described.

We report using a fast-shuttering CCD camera to determine the transverse stiffness of an optical tweezers trap. By utilizing the relation between the particle position variance and the trap stiffness we are able to circumvent the sampling requirements of the position power spectrum approach. We find that for trap stiffness of order 10 -5 N/m exposures shorter than 0.001 sec seem to be required in order to give valid measurements. For longer exposures (~0.01 sec) the trap stiffness can be overestimated by a factor of three or more.

The hypothesis that people classified as Type A and Type B have different personality profiles based on five major personality factors (Neuroticism, Extraversion, Openness, Agreeableness, and Conscientiousness) was tested using the Student Jenkins Activity Survey and the Revised NEO Personality Inventory. Results based on discriminant function analysis of data from 243 psychology undergraduates (105 males and 138 females) strongly supported the hypothesis indicating that Type A and Type B groups have significantly different Revised NEO Personality Inventory profiles and that the standardized discriminant function coefficients were large for Agreeableness and Conscientiousness and moderately large for Extraversion.

In this paper we consider the formation of morphological templates using adaptive resonance theory. We examine the role of object variability and noise on the clustering of different sized objects as a function of the vigilance parameter. We demonstrate that the fuzzy adaptive resonance theory is robust in the presence of noise but that for poor choice of vigilance there is a proliferation of prototypical categories. We apply the technique to detection of abnormal cells in pap smears.

An optoelectronic system has been designed to pre-screen pap-smear slides and detect the suspicious cells using the hit/miss transform. Computer simulation of the algorithm tested on 184 pap-smear images detected 95% of the suspicious region as suspect while tagging just 5% of the normal regions as suspect. An optoelectronic implementation of the hit/miss transform using a 4f Vander-Lugt correlator architecture is proposed and demonstrated with experimental results.

Since it was first adduced as a possible explanation for the periodic occurrence of the earths ice ages [1] [2], the phenomenon of stochastic resonance has been extensively investigated, both theoretically and experimentally. An intuitive understanding of the phenomenon may be developed by considering figure 1(a) which shows a particle in a double potential well. Under a large periodic perturbation of the wells (1(b)) the particle can escape from well to well, and if we regard the particle position as the observable or signal then the power spectrum of this signal will exhibit a large peak at the perturbation frequency. Under large random forcing the particle also makes the transition between the two wells but this time in a random fashion and the output spectrum is typically Lorentzian. In the case where there is a small periodic perturbation the particle is unable to escape from the right hand well where it resides initially (fig 1(c)). However, if random noise is now added the particle can make the transition with a greater probability from right to left when the right well is higher and vice versa when it is in the left well. As may be imagined, if there is insufficient noise then no transitions are made while if the noise is too large then the random switching dominates. It is this notion of an optimum amount of noise coupled with a small periodic component in a nonlinear system that forms a definition of stochastic resonance.

Pap-smears are slides of cellular material used to screen for cervical cancer. Currently pap-smears are examined manually, a repetitive and tedious task which leads to about 18 % false negative rate (% of abnormal slides going undetected) [1]. Automated pap-smear examination is desirable as both a quality control mechanism for detecting abnormal slides missed by human inspection and as a primary diagnostic cytology screen. Automated screening is challenging since it is a typical example of the “needle-in-a-haystack” problem, where the features of interest are hidden in a vast search area. In a pap-smear one in 10,000 cells screened may be abnormal. Detecting this cell requires high computation power and throughput. Each slide is 2.5 cm x 5.0 cm on a side. Features of interest (the cell nuclei) are on average 10 microns in diameter and sampling at 0.8 μm/pixel, (equivalent to screening the slide with a 20x objective) therefore requires at least 37,000 images of 256x256 pixels to be processed for each slide screened. In this paper we present an optoelectronic implementation of the morphological hit-or-miss transform which can scan each pap-smear slide and detect the regions of interest (ROI) in that slide in under five minutes.

In this paper we describe our recent work developing automated methods for generation of kernels or structuring elements for use in the hit-or-miss transform. We show how a neural network algorithm (Fuzzy Adaptive Resonance Theory) generates hit and miss structuring elements that can be used with a fuzzy morphology to detect a class of objects and we illustrate with computer simulations.

Automation of the Pap-smear cervical screening method is highly desirable as it relieves tedium forthe human operators, reduces cost and should increase accuracy and provide repeatability. We presenthere th design for a high-throughput optoelectronic system which forms the first stage of a two stagesystem to automate pap-smear screening. We use a mathematical morphological technique called thehit-or-miss (HOM) transform to identify the suspicious areas on a pap-smear slide. This algorithm isimplemented using a VanderLugt architecture and a time-sequential ANDing smart pixel array. 1 Introduction Cervical cancer is a major cause of malignant death in women in the "industrial world" and the most common cause of malignant death in India and South America. As many as 7000 women die every yearin the U.S from cervical cancer. However, the disease is largely preventable by regular screening. HencePapanicolaou smears (pap-smears) are increasingly used to screen women from the age of 16 to 60 for thiscancer. A pap-smear slide is prepared by obtaining a cellular sample from the patients cervix which issmeared on to a glass slide, then fixed with an alcohol based fixative and stained using the Papanicolaoustain. This slide can now be screened for cervical cancer by examining the squamous cells on the slide.Figure 1 shows a typical 480x512 image taken from a pap-smear slide at a resolution of 0.7 im/samp1inginterval.Automating the pap-smear screening is desirable to the health care industry as it relieves humansfrom the tedious job of microscope based examinations, reduces the cost of a pap-smear examination,should increase its accuracy and make the screening process repeatable. However, automated screeningis challenging due to the large computational resources required and the low tolerance for error. A slidecover glass has dimensions of 2.5x5.0 cm and the features of interest (the cell nuclei) are on average 10 jtmin diameter. This implies processing a large area at fine resolution which translates to high space(time)-bandwidth product. For example, using a sampling resolution of 0.7 tm/sample requires a minimum of39,000 images to be processed for each slide screened. Also, among the few pap-smears slides which dodisplay abnormal cells, the actual number of abnormal cells may be as few as 10 of the 100,000 or so cellspresent on the slide. This calls for precision in pre-processing and classification of each sub-area. Focalplane processing (FPP) has the potential to offer high throughput as required in this application, and suchFPP can be implemented using optoelectronic devices. In this paper we will develop an optoelectronicallyimplementable algorithm to identify suspicious regions (also referred to as regions of interest (ROT)) on apap-smear slide and discuss its implementation.

An optoelectronic morphological processor for detecting regions of interest (abnormal cells) on a cervical smear slide using the hit/miss transform is presented. Computer simulation of the algorithm tested on 184 Papsmear images provided 95% detection and 5% false alarm. An optoelectronic implementation of the hit /miss transform is presented, along with preliminary experimental results.

The design and test results of an optoelectronic smart detector array for use in a fingerprint ridge extraction system are presented. The system comprises a microlens array for taking multiple Fourier transforms and an array of detectors with 8 or 16 elements. Each detector is composed of a set of wedge photodetectors with associated winner-take-all circuitry and encoders for reporting the maximally stimulated wedge element. The size of one pixel is 440 X 440 μm2. The rises and fall times of the detector were 80 and 750 μs, respectively. By using the system, the extraction of fingerprint ridge directions has been successfully demonstrated.