Mosquito species identification using convolutional neural networks with a multitiered ensemble model for novel species detection.

With over 3500 mosquito species described, accurate species identification of the few implicated in disease transmission is critical to mosquito borne disease mitigation. Yet this task is hindered by limited global taxonomic expertise and specimen damage consistent across common capture methods. Convolutional neural networks (CNNs) are promising with limited sets of species, but image database requirements restrict practical implementation. Using an image database of 2696 specimens from 67 mosquito species, we address the practical open-set problem with a detection algorithm for novel species. Closed-set classification of 16 known species achieved 97.04 ± 0.87% accuracy independently, and 89.07 ± 5.58% when cascaded with novelty detection. Closed-set classification of 39 species produces a macro F1-score of 86.07 ± 1.81%. This demonstrates an accurate, scalable, and practical computer vision solution to identify wild-caught mosquitoes for implementation in biosurveillance and targeted vector control programs, without the need for extensive image database development for each new target region.

Development of a low-cost imaging system for remote mosquito surveillance.

Targeted vector control strategies aiming to prevent mosquito borne disease are severely limited by the logistical burden of vector surveillance, the monitoring of an area to understand mosquito species composition, abundance and spatial distribution. We describe development of an imaging system within a mosquito trap to remotely identify caught mosquitoes, including selection of the image resolution requirement, a design to meet that specification, and evaluation of the system. The necessary trap image resolution was determined to be 16 lp/mm, or 31.25um. An optics system meeting these specifications was implemented in a BG-GAT mosquito trap. Its ability to provide images suitable for accurate specimen identification was evaluated by providing entomologists with images of individual specimens, taken either with a microscope or within the trap and asking them to provide a species identification, then comparing these results. No difference in identification accuracy between the microscope and the trap images was found; however, due to limitations of human species classification from a single image, the system is only able to provide accurate genus-level mosquito classification. Further integration of this system with machine learning computer vision algorithms has the potential to provide near-real time mosquito surveillance data at the species level.

Free tissue transfer to the traumatized upper extremity: Risk factors for postoperative complications in 282 cases.

BACKGROUND: Complex traumatic upper extremity injuries frequently possess compromised local vasculature or extensive defects that are not amenable to local flap reconstruction. Free tissue transfer is required to provide adequate soft tissue coverage. The present study aimed to evaluate risk factors that contribute to postoperative complications and flap loss in complex upper extremity reconstruction. METHODS: Retrospective chart review was performed for all patients undergoing free tissue transfer for upper extremity reconstruction from 1976 to 2001. Data collected included patient demographic characteristics, timing of reconstruction, location of injury, fracture characteristics, operative interventions, and postoperative complications. Statistical analysis was performed using χ(2) and Fisher exact tests. RESULTS: In total, 238 patients underwent 285 free tissue transfers and met inclusion criteria, from which 3 were excluded because of inadequate information (n = 282). Extremities were repaired within 24 h (75 cases; 27%), in days 2-7 (32 cases; 12%), or after day 7 (172 cases; 62%). Timing of reconstruction did not significantly affect postoperative outcomes. Proximal location of injury was significantly associated with superficial (relative risk [RR], 6.5; P < .01) and deep infection (RR, 5.3; P < .01), and osteomyelitis (RR, 4.0; P < .01), although not with flap failure (P = .30). Presence of an open fracture was significantly associated with developing superficial (RR, 3.1; P = .01) and deep (RR, 1.9; P < .01) infection, as well as osteomyelitis (RR, 1.6; P < .01). Having a closed fracture did not negatively influence postoperative outcomes. CONCLUSIONS: This study supports the safety of early free tissue transfer for reconstruction of traumatized upper extremities. Injuries proximal to the elbow and open fracture were associated with a significantly higher infection rate. Gustilo grade IIIC fractures, need for interpositional vein grafts, and anastomotic revision at index operation resulted in significantly higher risk of flap loss, whereas the presence of fracture, fracture fixation, and injury location were not predictors of flap failure.

Applying lean principles to continuous renal replacement therapy processes.

PURPOSE: The application of lean principles to continuous renal replacement therapy (CRRT) processes in an academic medical center is described. SUMMARY: A manual audit over six consecutive weeks revealed that 133 5-L bags of CRRT solution were discarded after being dispensed from pharmacy but before clinical use. Lean principles were used to examine the workflow for CRRT preparation and develop and implement an intervention. An educational program was developed to encourage and enhance direct communication between nursing and pharmacy about changes in a patient's condition or CRRT order. It was through this education program that the reordering workflow shifted from nurses to pharmacy technicians. The primary outcome was the number of CRRT solution bags delivered in the preintervention and postintervention periods. Nurses and pharmacy technicians were surveyed to determine their satisfaction with the workflow change. After implementation of lean principles, the mean number of CRRT solution bags dispensed per day of CRRT decreased substantially. Respondents' overall satisfaction with the CRRT solution preparation process increased during the postintervention period, and the satisfaction scores for each individual component of the workflow after implementation of lean principles. The decreased solution waste resulted in projected annual cost savings exceeding $70,000 in product alone. CONCLUSION: The use of lean principles to identify medication waste in the CRRT workflow and implementation of an intervention to shift the workload from intensive care unit nurses to pharmacy technicians led to reduced CRRT solution waste, improved efficiency of CRRT workflow, and increased satisfaction among staff.

The cutaneous arteries of the anterior abdominal wall: a three-dimensional study.

BACKGROUND: Abdominal perforator flaps represent a natural progression in the quest to minimize abdominal wall morbidity. Their one disadvantage is the significant rate of vascular complications to which they are subject in some series. The authors examined the vascular anatomy of the abdominal integument, to determine why such complications occur and how they may be prevented. METHODS: In 10 fresh cadavers, major arteries supplying the abdominal wall were injected with a lead-based contrast medium. The abdominal integument of each cadaver was imaged using a 16-slice spiral computed tomography scanner, to produce three-dimensional reconstructions of the arterial anatomy. Reconstructions were observed for orientation, course, and morphology of the major perforators within the abdominal integument. RESULTS: Perforators of the deep inferior epigastric artery (DIEA) varied markedly in their orientation, course, and morphology among specimens. By contrast, perforators of the superior epigastric artery (SEA) were relatively consistent in their morphology and orientation. In eight of 10 specimens, SEA perforators with extensive anatomical "territories" orientated toward the umbilicus were present. These SEA perforators pierced the rectus sheath within 4 cm of the costal margin and were present bilaterally in seven of eight specimens. CONCLUSIONS: The unpredictable orientation and course of DIEA perforators indicate that the blood supply of abdominal perforator flaps, raised without clear knowledge of their unique vascular anatomy, may often be more random than axial. This may account for much of the ischemia-related morbidity observed with DIEA-based perforator flaps. Preservation of SEA perforators adjacent to the costal margin during abdominoplasty will likely improve abdominal wall perfusion and reduce donor-site morbidity.

Three-dimensional CT angiography: a new technique for imaging microvascular anatomy.

To date there has been no satisfactory research method for imaging microvascular anatomy in three dimensions (3D). In this article we present a new technique that allows both qualitative and quantitative examination of the microvasculature in 3D. In 10 fresh cadavers (7 females, 3 males, mean age 68 years), selected arteries supplying the abdominal wall and back were injected with a lead oxide/gelatin contrast mixture. From these regions, 30 specimens were dissected free and imaged with a 16-slice spiral computed tomographic (CT) scanner. Using three-dimensional CT (3D-CT) angiography, reconstructions of the microvasculature of each specimen were produced and examined for their qualitative content. Two calibration tools were constructed to determine (1) the accuracy of linear measurements made with CT software tools, and (2) the smallest caliber blood vessel that is reliably represented on 3D-CT reconstructions. Three-dimensional CT angiography produced versatile, high quality angiograms of the microvasculature. Correlation between measurements made with electronic calipers and CT software tools was very high (Lin's concordance coefficient, 0.99 (95% CI 0.99-0.99)). The finest caliber of vessel reliably represented on the 3D-CT reconstructions was 0.4 mm internal diameter. In summary, 3D-CT angiography is a simple, accurate, and reproducible method that imparts a much improved perception of anatomy when compared with existing research methods. Measurement tools provide accurate quantitative data to aid vessel mapping and preoperative planning. Further work will be needed to explore the full utility of 3D-CT angiography in a clinical setting.