Benchmarking Human Performance for Visual Search of Aerial Images.

Aerial images are frequently used in geospatial analysis to inform responses to crises and disasters but can pose unique challenges for visual search when they contain low resolution, degraded information about color, and small object sizes. Aerial image analysis is often performed by humans, but machine learning approaches are being developed to complement manual analysis. To date, however, relatively little work has explored how humans perform visual search on these tasks, and understanding this could ultimately help enable human-machine teaming. We designed a set of studies to understand what features of an aerial image make visual search difficult for humans and what strategies humans use when performing these tasks. Across two experiments, we tested human performance on a counting task with a series of aerial images and examined the influence of features such as target size, location, color, clarity, and number of targets on accuracy and search strategies. Both experiments presented trials consisting of an aerial satellite image; participants were asked to find all instances of a search template in the image. Target size was consistently a significant predictor of performance, influencing not only accuracy of selections but the order in which participants selected target instances in the trial. Experiment 2 demonstrated that the clarity of the target instance and the match between the color of the search template and the color of the target instance also predicted accuracy. Furthermore, color also predicted the order of selecting instances in the trial. These experiments establish not only a benchmark of typical human performance on visual search of aerial images but also identify several features that can influence the task difficulty level for humans. These results have implications for understanding human visual search on real-world tasks and when humans may benefit from automated approaches.

Alvimopan Is Associated With a Reduction in Length of Stay and Hospital Costs for Patients Undergoing Radical Cystectomy.

OBJECTIVE: To evaluate the impact of alvimopan in patient undergoing radical cystectomy (RC) for bladder cancer. We hypothesize that alvimopan can decrease cost for RC by reducing length of stay (LOS). METHODS: We identified patients who underwent elective RC for bladder cancer from 2009 to 2015 in the Premier Healthcare Database, a nationwide, all-payer hospital-based database, and compared patients who received and did not receive alvimopan in the perioperative period. Hospitals that had no record of administering alvimopan for patients undergoing RC were excluded. The primary outcomes were LOS and the direct hospital costs. The secondary outcomes were 90-day readmission for ileus and major complications. RESULTS: After applying the inclusion criteria, the study cohort consisted of 1087 patients with 511 patients receiving perioperative alvimopan. Alvimopan was associated with a reduction in hospital costs by -$2709 (95% confidence interval: -$4507 to -$912, P = .003), decreased median LOS (7 vs 8 days, P < .001), and lower likelihood of readmission for ileus (adjusted odds ratio: 0.63, P = .041). While alvimopan use led to higher pharmacy costs, this was outweighed by lower room and board costs due to the reduced LOS. There was no significant difference between 2 groups regarding major complications. These results were robust across multiple adjusted regression models. CONCLUSION: Our data show that alvimopan is associated with a substantial cost-saving in patients undergoing RC, and suggest that routine use of alvimopan may be a potential cost-effective strategy to reduce the overall financial burden of bladder cancer.

An extremely diverse CD4 response to vaccinia virus in humans is revealed by proteome-wide T-cell profiling.

CD4 T cells are required for the maintenance and recall of antiviral CD8 T cells and for antibody responses. Little is known concerning the overall architecture of the CD4 response to complex microbial pathogens. In a whole-proteome approach, 180 predicted open reading frames (ORFs) in the vaccinia virus genome were expressed and tested using responder cells from 20 blood samples from 11 vaccinees. Validation assays established the sensitivity and specificity of the system. Overall, CD4 responses were detected for 122 ORFs (68%). A mean of 39 ORFs were recognized per person (range, 13 to 63). The most frequently recognized ORFS were present in virions, including A3L and A10L (core proteins), WR148 (a fragmented homolog of an orthopoxvirus protein that forms inclusions in cells), H3L (a membrane protein), D13L (a membrane scaffold protein), and L4R (a nucleic acid binding protein). Serum immunoglobulin G profiling by proteome microarray detected responses to 45 (25%) of the ORFs and confirmed recent studies showing a diverse response directed to membrane and nonmembrane antigens. Our results provide the first empirical whole-proteome data set regarding the global CD4 response to full-length proteins in a complex virus and are consistent with the theory that abundant structural proteins are immunodominant.

Dominance and diversity in the primary human CD4 T cell response to replication-competent vaccinia virus.

Vaccination with replication-competent vaccinia protects against heterologous orthopoxvirus challenge. CD4 T cells have essential roles helping functionally important Ab and CD8 antiviral responses, and contribute to the durability of vaccinia-specific memory. Little is known about the specificity, diversity, or dominance hierarchy of orthopoxvirus-specific CD4 T cell responses. We interrogated vaccinia-reactive CD4 in vitro T cell lines with vaccinia protein fragments expressed from an unbiased genomic library, and also with a panel of membrane proteins. CD4 T cells from three primary vaccinees reacted with 44 separate antigenic regions in 35 vaccinia proteins, recognizing 8 to 20 proteins per person. The integrated responses to the Ags that we defined accounted for 49 to 81% of the CD4 reactivity to whole vaccinia Ag. Individual dominant Ags drove up to 30% of the total response. The gene F11L-encoded protein was immunodominant in two of three subjects and is fragmented in a replication-incompetent vaccine candidate. The presence of protein in virions was strongly associated with CD4 antigenicity. These findings are consistent with models in which exogenous Ag drives CD4 immunodominance, and provides tools to investigate the relationship between Ab and CD4 T cell specificity for complex pathogens.

Innate immune responses to herpes simplex virus type 2 influence skin homing molecule expression by memory CD4+ lymphocytes.

Herpes simplex virus (HSV) infections of humans are characterized by intermittent, lytic replication in epithelia. Circulating HSV-specific CD4 T cells express lower levels of preformed cutaneous lymphocyte-associated antigen (CLA), a skin-homing receptor, than do circulating HSV-specific CD8 T cells but, paradoxically, move into infected skin earlier than CD8 cells. Memory CD4 T cells develop strong and selective expression of CLA and E-selectin ligand while responding to HSV antigen in vitro. We now show that interleukin-12, type I interferon, and transforming growth factor beta are each involved in CLA expression by memory HSV type 2 (HSV-2)-specific CD4 T cells in peripheral blood mononuclear cells (PBMC). A reduction of the number of monocytes and dendritic cells from PBMC reduces CLA expression by HSV-2-responsive CD4 lymphoblasts, while their reintroduction restores this phenotype, identifying these cells as possible sources of CLA-promoting cytokines. Plasmacytoid dendritic cells are particularly potent inducers of CLA on HSV-reactive CD4 T cells. These observations are consistent with cooperation between innate and acquired immunity to promote a pattern of homing receptor expression that is physiologically appropriate for trafficking to infected tissues.

Diversity in the acute CD8 T cell response to vaccinia virus in humans.

Orthopoxviruses have complex proteomes. Infection provokes a brisk CD8 response, which is required in some systems for recovery from primary infection. Little is known concerning the Ags and epitopes recognized by CD8 T cells. We examined the fine specificity of cloned and bulk human vaccinia-specific CD8 CTL by expressing polypeptide fragments from a library of vaccinia genomic DNA. This epitope discovery method emphasizes virus-specific biological activity, as the responder cells are all reactive with whole vaccinia virus. Sixteen novel epitopes, restricted by several HLA A and B alleles, were defined to the nomamer peptide level in diverse vaccinia open reading frames. An additional seven epitope were mapped to short regions of vaccinia proteins. Targets of the CD8 response included proteins assigned to structural, enzymatic, transcription factor, and immune evasion functions, and included members of all viral kinetic classes. Most epitopes were conserved in other orthopoxviruses. Responses to at least 18 epitopes were detected within a single blood sample, revealing a surprising degree of diversity. These epitopes will be useful in natural history studies of CD8 responses to vaccinia, a nonpersisting virus with long-term memory, and in the design and evaluation of attenuated and replication-incompetent vaccinia strains being tested for variola and monkeypox prevention and for the delivery of heterologous Ags.

Expression of cutaneous lymphocyte-associated antigen and E-selectin ligand by circulating human memory CD4+ T lymphocytes specific for herpes simplex virus type 2.

Virus-specific memory T lymphocytes traffic to sites of viral infection. Herpes simplex virus (HSV) type 2-specific CD4(+) and CD8(+) T lymphocytes differ with regard to their homing kinetics to infected tissues. We studied the expression of cutaneous lymphocyte-associated antigen (CLA) and E-selectin ligand (ESL) by HSV-2-specific CD4(+) T lymphocytes. Virus-reactive T lymphocytes were identified ex vivo by CD154 or interferon-gamma up-regulation. We detected selective expression of CLA by HSV-2-reactive CD4(+) T lymphocytes, but at levels lower than those we previously observed for CD8(+) T lymphocytes. Short-term HSV-2-reactive CD4(+) lines generated from peripheral-blood mononuclear cells preferentially express CLA, compared with cytomegalovirus- or influenza-specific cells. CLA is expressed by HSV-2-reactive cells that are initially CLA negative before restimulation. Short-term culture-expanded HSV-2-specific CD4(+) T lymphocytes also selectively express ESL. These findings have implications for the optimization of vaccines for HSV and other cutaneous pathogens.