"... In PAGE 6: ...igure 7. Potential combinations of human and computer control.............................................. 18 List of Tables Table1 .... ..."

"... In PAGE 6: ... The region of increased uptake corresponding to the tumour has been excluded from the calculations. The results are summarised in Table2 and show that the non-rigid transformation model is better able to correct the motion of the breast than the rigid and affine transformations. 4 Discussion We have developed a fully automated algorithm for the non-rigid registration of 3D breast MRI based on normalised mutual information.... ..."

"... In PAGE 6: ... The region of increased uptake corresponding to the tumour has been excluded from the calculations. The results are summarised in Table2 and show that the non-rigid transformation model is better able to correct the motion of the breast than the rigid and affine transformations. 4 Discussion We have developed a fully automated algorithm for the non-rigid registration of 3D breast MRI based on normalised mutual information.... ..."

"... In PAGE 9: ... To visualize the separability of the pathology groups based on the most discriminatory (according to p value) of the measurements taken, we provide box plots showing the first, second, and third quartiles, the extent of nonout- liers, and the outliers for the mean of distances to cen- troid, the best (according to p value) of the 3D moment features, and the best (according to p value) of the sur- face area and volume measurements in Fig 8. Table5 shows the results of support vector machine classification. At each cell of the table, a value is given that indicates the classification accuracy achieved by the support vector machine in differentiating the pathology groups indicated by the row and column of that cell.... ..."

"... In PAGE 22: ...recision set for T quot;. The results are less sensitive to than in 5.2, due to the total symmetry in of this example. Results for d = 3 and several values are in Table3 , where T is again compared with m TM. The conservative rule for t0 is close to the theoretical time to reach quot;-stationarity here, but the proposed conservative rules for T are again overestimated when the chains mix slower, as shown by Table 4 which gives the chi-square distances.... ..."

"... In PAGE 6: ... If we consider normalized manifestation rate for each category (routing, LUT, and control/clock) which is normalized to the number of sensitive bits as #sensitive bits MTTM , routing bits are the most vulnerable ones. The detailed execution time of the presented SER esti- mation method is reported in Table4 . The total time of this SER estimation method includes a) the time needed to ex- tract the netlist and also to extract SRAM cells information from the XDL file, b) SP computation time, and c) error propagation probability computation time.... In PAGE 6: ... As can be seen in this table, SP com- putation is the most time-consuming part of our estimation method. The accuracy of our SER method versus the random- simulation method has also been reported in Table4 . The accuracy, on average, is about 95%.... ..."

"... In PAGE 23: ...05. Table11 and 12 show the results. Even after 25 million drawings IS and MH with the normal candidate distribution yield completely different results than the other algorithms.... ..."

"... In PAGE 7: ... The results of this categorization appear in Table 1. Table1 divides the processed ESTs into four categories. Many ESTs did not have LC regions within their blastx hits, and are classified as category 1 or 2.... In PAGE 7: ... Masking the LC regions improved the indication of function for ESTs in category 3, whereas masking did not for those in category 4. The results in Table1 show that 21.67% (categories 3 and 4) of the ESTs had LC regions detected by the XNU program1 within the hits found by blastx, and that the additional blastp runs with these LC regions masked out allowed for 17.... In PAGE 7: ... Thus, proper identification of meaningful hits was possible, and the addition of the LCFilter and mblastp modules into our system allowed for more accurate review of blastx similarity results. The ESTs that fell into category 3 in Table1 are divided into A) those in which enough positive hits remained in the blastp results to reveal a possible function, and B) those where no hits remained. We cannot rule out the possibility that the masking of LC regions by XNU caused possible true positive hits to be elimi- nated.... In PAGE 7: ... For this reason, we chose to keep both the blastx results and the blastp results and present them to users, along with infor- mation about whether an LC region occurred in the blastx hits. A statistical analysis of the results in Table1 reveals the confi- dence with which we expect the ratio of ESTs that fell into cate- gory 3 versus those that fell into category 4. Recall that category 3 sequences are cases where XNU and blastp successfully aided our analysis by more accurately determining if an EST could be identified or not.... In PAGE 7: ... On the other hand, XNU and blastp failed to aid our analysis in category 4 sequences. Using the experimental results shown in Table1 , the 375 ESTs that fell into category 3 and 4 can be considered approximately as a binomial random dis- tribution of what we would expect to see in the entire population of Arabidopsis ESTs with LC regions. With category 3 marked as success, the probability of success in this binomial experiment is p =.... ..."

"... In PAGE 13: ... The immediate concern in Kentucky is for the established producer using conventional-till methods to produce annual crops on highly erodible land. Assuming a grower is required to use no-tillage in order to remain in conservation com pliance, can he afford to buy the equipment to make the con- version? Normally, the established producer would change to a no-till system only if the additional net cash flow Table3 . Net cash flow advantageof conservationtillagesystems over conventional tillage, established producer, constant yields.... In PAGE 18: ... The lowest yields in the dry years were produced at locations where there was very little or no midsummer rainfall, while the best yields were obtained where heavy summer showers occurred at the right time. Summaries of yield data from all the corn, soybean, and grain sorghum demonstrations are shown in Table3 . In one 1983 demonstration, corn planted in crimson clover yielded 66 bu/acre without extra nitrogen.... In PAGE 19: ... One field, which yielded only 5 bu/acre, received no rain between plant ing and harvesting. On three farms where side-by-side com parisons were possible, two no-till plantings were higher yielding by an average of 8 bu/acre, while no-till and con- Table3 . Summary of no-till demonstrations with corn, grain sorghum, and soybeans.... In PAGE 25: ... average yields were increased by 33 percent and net returns by 46 percent. Excellent yields of high-quality cotton were produced at Halfway in 1987 ( Table3 ). No significant differences in ir rigated cotton yields were found, but net returns were in- creased when compared to conventional tillage cotton with minimum-till continuous cotton, terminated wheat-cotton, forage sorghum-cotton, and wheat-cotton conservation tillage rotations.... In PAGE 26: ...imited tillage - a profitable combination. J. Soil and Water Conserv. 38: 233-236. Table3 . Cotton yield and value, production costs, and relative profitability of irrigated and dryland cropping systems at Halfway, Texas, 1987.... In PAGE 38: ... Fall disk tillage very likely con tributed to the formation of the pan. There may have also been a natural component to formation in that the transition zone between the A and B horizons, which often occurs at Table3 . Effect of on 8-vear mean vield.... In PAGE 41: ... Interseeding soybeans into standing wheat 2 weeks before harvest did not reduce wheat yield. Table3 shows the soil cone index values averaged over the top 15 inches of soil depth for soybean plots one month after planting. Two sets of penetrometer readings were taken for Table 3.... In PAGE 41: ... Table 3 shows the soil cone index values averaged over the top 15 inches of soil depth for soybean plots one month after planting. Two sets of penetrometer readings were taken for Table3 . Cone index values and yield from soybean plots.... In PAGE 41: ... However, there was no statistically significant difference between plots paraplowed only once in fall of 1986 and those which had extra tillage operation with paraplow in June, 1987. Traffic significantly increased soil compaction as shown by penetrometer measurements within the soybean rows ( Table3 ).Figure 4 shows profiles of cone index versus depth for paraplow plots about 8 months after tillage operations.... In PAGE 41: ... This indicates that one tillage operation in the fall, deep enough to remove root inhibiting hardpans, in conjunction with controlled traffic, could eliminate deep tillage of any kind for soybeans. Paraplowing prior to planting soybeans significantly in- creased crop yield ( Table3 ). Statistically, there was no signifi cant difference among chisel plow plots planted with the Clemson interseeder, chisel plots planted with the KMC sub soiler/planter (conventional doublecropping) and paraplow plots with no deep tillage prior to soybean planting.... In PAGE 44: ...796 4.05 113 Table3 . Corn grain yields at SandMountain Substation, 1985.... In PAGE 47: ...10) 5 88 300 472 1.320 Table3 . Corn grain yield in 1987 as affected by tillage and hybrid selection.... In PAGE 47: ... In 1986, when moisture was not limiting because of sup plemental irrigation, grain yield was not affected by tillage or hybrid (average yield, 120 bu/acre). Without supplemen tal irrigation, in 1987, there was a tillage x hybrid interaction ( Table3 ) .Sunbelt 1827 demonstrated the largest yield poten tial of the three hybrids.... In PAGE 50: ...05). Table3 . Influenceof tillage and irrigation on grain yield of corn and soybean in 1987.... In PAGE 50: ... Rainfall in tervals that favor a yield response to tillage are greater for soils with high water holding capacity than for soils with a low water holding capacity. Irrigation improved corn yields with both conventional and conservation tillage systems in 1987 ( Table3 ).The yield in- crease was 79 percent for conventional tillage and 63 per- cent for conservation tillage.... In PAGE 59: ....V. (%) 34 (136) 8 (78) 15 ( 48) GS = Grain sorghum; SB = Soybeans; JG = Johnsongrass. Table3 . Annual yield for soybeans grown in a rotational system schemes of greater than one year for this study, there appears .... In PAGE 63: ...tillage systems; in 1985 all paraplow treatments had significantly lower plant densities than both Ro-till 11 to 12-inchdepth treatments ( Table3 ). The Paraplow 4 to 6-inch depth had the lowest plant density and was significantly lower than both Ro-till ll to 12 ,and 14 to 15-inch depths fertilizer surface-incorporated; in 1987, however, the Table 2.... In PAGE 64: ... The Ro-till 7 to 8-inch depth, injected fertilizer treatment had the lowest plant density and was significantly lower than both Ro-till 11 to 12 and 14 to 15-inch depths with injected fer tilizer treatments, no-till, and Paraplow 4t to 6 and 6 to 8-inch depth treatments. Soybean yields were higher in 1985 than both 1986 and Table3 . Effect of reduced tillage systems and fertilizer placement on soybean plant population, height at maturity, and yield on a Providence silt loam soil in 1985-87 at the Pontotoc Ridge-Flatwoods Experiment Station.... In PAGE 69: ....................... Conventional 2,090 1,700 Strip tillage 2,370 1,850 2,550 No tillage 2,400 2,200 Table3 . Soybean cyst nematode counts found in 1985 through 1987 at Crossville, AL, with different tillage and crop rotation svstems.... In PAGE 69: ... This rate of yield loss with time could be due to the SCN populations building up very rapidly in the early stages of the tests under the conventional tillage, with these yields dropping during the second and third years. The number of SCN counts60 days after planting with con ventional tillage declined with time ( Table3 ). However, yields of Essex soybeans continue to be lowest with conventional tillage, indicating that other factors may be reducing the SCN populations at a lower levels during the early years of the tests.... In PAGE 71: ... Soil Chemical Analysis As expected, application of lime after wheat harvest in 1984 increased soil pH and Ca levels of samples obtained in March 1986. Tillage influenced the distribution of P, pH, and Ca with depth in the soil ( Table3 ). In the disked and subsoil plus disk treatments, lime increased soil pH from 5.... In PAGE 71: ... ns = not significantly different. Table3 . Soil pH, Ca, and P values from soil tests conducted during March of 1986 as influenced by tillage every spring and a lime application made during May 1984, Baton Rouge, LA.... In PAGE 78: ... Fixed costs for Paraplow and Ro-till implements were about the same as the chisel + disk system. Direct Costs Direct costs ( Table3 ) ranged from $97 to $140/acre and were higher for both Pontotoc and Raymond than Verona. The higher costs at both Pontotoc and Raymond were associated with the postemergence herbicide system for weed Table 1.... In PAGE 78: ...49 24.49 *Depth of fertilizer placement and tillage Table3 . Estimated 1987 direct costs for reduced tillage systems on three soil resource areas.... In PAGE 85: ...4.50/hour. Interest rate on operating capital was computed at 10 percent annual percentage rate Results and Discussion Wheat The first year, 1984, only the 24-inch wheel track spacing was used. The 24-inch wheel tracks with the relay planted treatments produced less yield than the 28- and 32-inch in 1986, but not in 1985 ( Table3 ). The tractor we used had 20-inch wide tires and the 24-inch wide skip for the tractor wheels allowed only 4 inches for wheel track error in the field planting operation.... In PAGE 86: ...Table3 . Effect of relay planting and tractor wheel track width Table4.... ..."