Homogeneity assumptions were generally appropriate under low B0 eld strengths and short read out times. However, demand for faster, higher resolution scans and methods such as functional magnetic resonance imaging (fMRI) require fast methods and higher B0 eld strengths. As fast imaging techniques such as echo-planar imaging (EPI) and spiral scans gain popularity, image artifacts from B0 eld inhomogeneity are visible. These artifacts cause signal loss and result in shifts or blurring in the nal MR images, making qualitative and quantitative analysis difcult. These effects are exacerbated in high B0 elds. Similarly, as MR main elds grow in strength, image artifacts from B1 eld inhomogeneity are visible. At higher eld strengths, the RF wavelength is shortened, and experiences further shortening due to changes in the tissue dielectric constant, resulting in higher inhomogeneity at higher main eld strengths. The nonuniform effect in each voxel gives a possibly different tip angle in each voxel. This gives spatially varying signal and intensity in the image, making both qualitative and quantitative analysis difcult. Therefore, the speed and eld strength requirements of state-of-the-art MR technology further exacerbate the problems of inhomogeneity. Correcting for these artifacts is possible using the appropriate eld map. Given a smooth eld map of B0 inhomogeneity, conjugate phase methods can compensate for phase accrual at each voxel, tailored RF pulses can compensate for signal loss, or iterative reconstruction methods can be used to obtain corrected nal MR images under the condition of an inhomogeneous B0 eld. Similarly, given a map of B1 inhomogeneity, tailored RF pulses, parallel transmit excitation, and dynamic adjustment of RF power can compensate for B1 inhomogeneity. Highly accurate and reliable eld map estimators are required in these intensive imaging environments. Previous estimators have often been based on heuristic algorithms rather than on a statistical estimation theory. These estimators are often limited in scope, dependent on a strict measurement scheme, specic imaging parameters, or ignore complicating physical effects. Additionally, these estimators often satisfy the requirement for smooth eld maps
through low pass ltering and smoothing of calculated eld maps. New statistically based estimators that are based on more comprehensive models are needed. Estimators are needed that incorporate the knowledge that true eld maps are smooth with an understanding of the effect of smoothing on image spatial resolution. This thesis presents three separate estimators that satisfy these desired estimator properties. Chapter II rst presents a short introduction to MRI. Section 2.4 follows with a brief discussion of the effects on eld inhomogeneity - the motivation for new statistical estimators. Chapter III overviews some principles of iterative penalized estimator design, which are used as the solution in this report. Chapter IV tackles the problem of main eld map estimation, considering both current solution and proposing the new solution as well as demonstrating its effectiveness. Chapter V similarly looks at B1 map estimators, considering current solutions and proposing a new iterative estimator and demonstrating its effectiveness. Chapter VI, noting the interdependence of B1 and the longitudinal relaxation time T1 , considers current solutions to T1 mapping and joint B1 , T1 mapping and their limitations and proposes a new joint estimator for T1 and B1 which incorporates slice prole effects and Bloch non-linearity. Finally, Chapter VII concludes, summarizing the proposed solutions in this work and giving future work in the goal of estimating parameter maps in MRI.

no phase unwrapping is required. Our model also takes into account R2 decay, which was
ignored in previous multiple echo techniques.

Multiple Scan Model

We now generalize the conventional model (4.1) to the case of multiple scans, i.e., with more than one echo time difference. The reconstructed images are denoted here by y 0 ,. , y L , where L is the number of echo time differences. Because we are using multiple
echo time differences, R2 decay may no longer be negligible and should be included in our
model. Our model for these images is:
l yj = fj ej l eRj l + l , j
for l = 0,. , L, where l denotes the echo time difference of the lth scan relative to the original scan i.e., (0 = 0), where j denotes the voxel number and where Rj denotes the
value of R2 for the jth voxel. As in most eld map estimation methods, this model assumes
implicitly there is no motion between the scans. As in (4.1), fj denotes the complex transverse magnetization and l denotes the (complex) noise. If we choose the l values j carefully, this data model allows for a scan that is free or largely free of phase wraps but which gives a phase difference lower in SNR, as well as scan(s) with wrapped phase but higher in SNR. Including the scan(s) with a larger echo time difference should help reduce noise in j , whereas the wrap-free scan helps avoid the need for phase unwrapping tools. 4.2.5 Maximum-Likelihood Field Map Estimation
The conventional estimate (4.2) appears to disregard noise effects, so a natural alternative approach is to estimate using a maximum likelihood (ML) method based on a statistical model for the measurements y. In MR, the k-space measurements have zeromean white gaussian complex noise [85], and we furthermore assume here that the additive noise values in y in (4.3) have independent gaussian distributions2 with the same variance
Independence in image space is an approximation. The noise values in k-space data are statistically independent, but reconstruction may produce correlations, especially in scans with non-Cartesian k-space imaging.
2. Under these assumptions, the joint log likelihood for f and given y = (y 0 ,. , y L ) is

m,n scans will also be affected by R2 decay through wj if the data is not scaled to compensate
for this factor. To simplify selecting , we normalize the data by the median of the square root of (4.15) using the approximation (4.9) for wj. Normalizing by this factor allows us to create a standard to FWHM table or graph (e.g., Fig. 4.1). If this normalization were not applied, a similar gure would need to be calculated with each new data set (or at least with each new set of l values) or would need to be chosen empirically. Normalizing based on the analytical result (4.16) enables us to use the same for all scans. We used the inverse 2D DSFT of (4.16) to compute the PSF h[n, m] and tabulate its FWHM as a function of , assuming the previous corrections were made and that the pixel j has dj = 1. Fig. 4.1 shows this FWHM as a function of log2 (), for both p = 1 and p = 2. The FWHM increases monotonically with , as expected, although the knees in the curve are curious. Nevertheless, one can use this graph to select the appropriate given the desired spatial resolution in the estimated eld map. The resulting spatial resolution will be inherently nonuniform, with more smoothing in the regions with low magnitudes and vice versa. One could explore modied regularization methods [35] to make the resolution uniform, but in this application nonuniform resolution seems appropriate since the goals include interpolating across signal voids.
4 2ndorder 1storder 3.2.1.6

FWHM [pixels]

log ()
Figure 4.1: Angularly averaged FWHM of PSF. Shown for eld map estimation as a function of log2 for dj = 1 in (4.16).
Qualitative Example: L = 1
0 Fig. 4.2 shows an example of the data magnitude yj and the usual phase estimate based
on L = 1 (4.2) which is very noisy. This is real data taken from a 3T MR scanner with 1 = 2 ms. The maximum value of |j 1 | is 1.61 radians in nonzero voxels, making the scan free of any phase wraps. Fig. 4.2 also shows the penalized-likelihood estimate based on (4.13) using two different values for and using 150 iterations. Here, we can see the improvement from using a regularized estimator versus the conventional ML estimator. The effect of on the smoothness of the estimate is also seen. The improvement seen is analyzed quantitatively in Section 4.3. Fig. 4.2 also shows the effective FWHM (in pixels) of the regularized algorithm based on (4.16) for both values of. Most of the image has a FWHM corresponding to the chosen based on Fig. 4.1. Areas of low magnitude have a much higher FWHM (such as the sinuses) and areas of high magnitude have the lowest FWHM.

Hz for the Gaussian ltered estimate, 3.4 Hz for the L = 1 regularized estimate, and 1.9 Hz for the L = = 3 regularized estimate and 1.7 Hz for the L = = 5 regularized estimate. Overall, the ltered conventional estimate performed similar to the PL method with L = 1 over the masked region, but had higher error in the ROI. The PL method with L = 2 showed a decreased error in both the masked region and the ROI. We would expect even higher improvement over any practical Gaussian ltered estimate because a suboptimal would be used. The proposed regularized estimators are more accurate in pixels with low magnitude. Adding additional scans (L > 1) makes the PL estimate even more accurate. Fig. 4.5 shows the improvement (dened as the RMS error for PL estimate with L = 1 divided by the error for PL estimate with L = 2) gained by using an additional set of data for the Gaussian example. For comparison, we also plotted the predicted improvement, given by the square root of the ratio of the expressions (4.19) and (4.20). The experimental gains are actually higher than the improvements anticipated as shown by the dotted lines (the predicted improvement) for some SNR values. Because this is a ratio of RMSEs and the amount of bias can vary between L = 1 and L = 2, the unbiased CRB provides a benchmark of expected ratios rather than an exact upper limit. Also, recall that (4.19) and
(4.20) considered R2 to be a known value when, in fact, R2 is unknown and approximated
through (4.9). The RMSE is low (in voxels with large magnitudes) at high SNRs using either L = 1 or L = 2. At lower SNRs, however, including in voxels with low magnitudes, using L = 2 and higher values of 2 greatly reduces RMS error. We repeated these sim ulations with R2 = 0 (results not shown) and the empirical improvement almost exactly
matched (4.18). Fig. 4.6 shows the improvement gained by using an additional set of data for the brain image. For a low SNR (for example 10 dB), the improvements are close to expected. The brain image has some areas where the magnitude is very low, making estimation using any method quite challenging. In addition, the eld map phase itself is less smooth than in the
RMSE improvement over 2 sets for Gauss data, R* =20 sec1

Space Averaged and |Bias| 3 2.[Hz] 1.0.k1 Improvement in over =1

|bias|

Improvement 2.3.4.5 k5.6.improvement () expected improvement ()
Figure 4.7: Bias and RMSE improvement for Gaussian example. Top: Space-averaged and absolute bias for several 2 values; Bottom: RMSE improvement, empirical and expected, over 2 = 1 for several 2 values.
Table 4.1: Phantom NRMSE for two representative slices Phantom NRMSE (%) for one realization Slice One Slice Three Whole Image Low Magnitude Whole Image Low Magnitude No Field Map 31.1 4.8 20.4 2.9 Conventional 15.0 0.5 6.8 0.6 15.5 0.2 2.2 0.1 Gauss Filter 14.3 0.4 6.1 0.4 15.1 0.2 1.9 0.08 L=1 13.0 0.4 4.0 0.4 15.2 0.2 1.8 0.04 L== 2 13.1 0.4 4.1 0.4 14.8 0.1 1.8 0.03 L== 5 13.5 0.08 4.3 0.2 14.6 0.02 1.8 0.01 L=3 13.5 0.09 4.4 0.1 14.6 0.02 1.8 0.01
Space Averaged and |Bias| 2 1.5 [Hz] |bias|
4 k1 Improvement in over =1
Improvement improvement () expected improvement () 2 2.3.4.5 k5.6.5 7
Figure 4.8: Bias and RMSE improvement for brain example. Top: Space-averaged and absolute bias for several 2 values; Bottom: RMSE improvement, empirical and expected, over 2 = 1 for several 2 values.

|x| (8 shot)

Conventional

Gauss Filter

L=2 =2

L=2 =5

L = 3 Field map

|x| (1 shot)

conventional

Filtered

L = = 2

L = = 5

Figure 4.9: MR phantom data eld map reconstructed using proposed method. First Slice - Top: Reconstructed 8-shot image, Conventional eld map, Gaussian ltered eld map, regularized eld map L=1, regularized eld map L=2; 2 = 2, regularized eld map L== 5, regularized eld map L=3. The eld maps are displayed with a common color scale from -35 Hz to 50 Hz; Bottom: Reconstructed one-shot image with no eld map and with each of the eld maps above. The images are all on the same color scale. These are all from one representative realization. 4.3.2 MR phantom data: Application to Spiral Trajectories

+ time and enable rapid B1 mapping, such as [20]. Some fast methods have been developed
that concurrently estimate or correct the B1 eld, (e.g., [24]) to circumvent the difculty of a quick direct mapping. Some methods have been developed that are T1 oblivious over the relevant range of T1 values (e.g., [39]) to circumvent needing T1 information at all.
+ All current B1 mapping have disadvantages that need to be corrected (e.g., ow artifacts,
off-resonance, suceptibility effects), but most have low noise and low bias [81]. Because the proposed method is built around a very general cost function, it is also applicable to fast methods developed for the DAM. Our proposed method seeks to map both the magnitude and phase of the B1 eld. This method uses a statistical cost function that incorporates noise and slice selection effects ignored by the conventional estimate. Including regularization into our cost function also circumvents the need for later ltering.
Signal model for multiple coils, multiple tip angles/coil combinations
Suppose there are K coils. We take M measurements by transmitting with different coil combinations and receiving from a common coil. (This method could be generalized to use multiple receive coils.) For each measurement, one or more coil(s) are driven simultaneously by the same RF signal b1 (t) with possibly different known amplitude scaling factors mk , where k = 1,. , K denotes the coil number, m = 1,. , M denotes the measurement number, and is a M K array containing the scaling factors mk. For the problem to be tractable, we require that M > K. The complex coil patterns sum together due to linearity to make the total transmitted B1 eld. This general model encompasses the
conventional model (5.1) if we let K = 1, M = 2, and 1 = . 2 We model the resulting M reconstructed images as follows:

yjm = fj F

mk zjk

+ jm ,

for m = 1,. , M and j = 1,. , N , where fj denotes the underlying object transverse magnetization in the jth voxel (multiplied by the sensitivity of the receive coil) and jm
+ denotes zero-mean complex gaussian noise. The B1 map, constrained to be real in the + conventional model, is actually a complex quantity. zjk denotes the unknown complex B1
map that relates RF amplitude to tip angle at the jth voxel for the kth coil. When multiple coils are driven by the same signal b1 (t) (with possibly different amplitudes), then the elds from those coils will superimpose and the complex coil patterns will add by linearity, hence the sum over k in (5.3). If the units of the amplitudes mk are gauss, then the units of zjk will be radians per gauss. More typically, the units of mk are arbitrary, and all that is known is their relative values. In this case zjk will have units such that the product of mk and zjk has units of radians. This should sufce for RF pulse design. The function F in (5.3) replaces the typical sin seen in the double angle formula and inherently incorporates slice selection effects. The function F is explained further in Appendix B. The model (5.3) expands the one used in [41, 42] and includes both slice selection effects and linear transceive coil combinations. Recent B1 mapping methods [10, 90] have introduced linear combinations of transmit coils. These methods have the advantage of using much smaller tip angles while still collecting enough signal to produce accurate results. The proposed method accomodates this matrix transmit technique with a comprehensive

Table 5.2: Simulation NRMSE (%) for proposed method M = 5 versus conventional DAM method M = 8 averaged over 20 realizations (truncated sinc pulse with SNR=30dB) |z|(0) |z|(n) z (0) z (n) # of iters LOO M = 5 proposed fj > 0.1 max fj 43.5 19.9 12.2 4.Low Magnitude 49.4 22.8 25.7 4.fj > 0.1 max fj 43.5 13.2 12.2 3.Low Magnitude 49.4 13.5 25.7 4.fj > 0.1 max fj 43.5 9.1 12.2 3.Low Magnitude 49.4 8.0 25.7 3.LOO M = 8 DAM fj > 0.1 max fj 5.8 11.1 Low Magnitude 31.6 45.2 OAAT M = 5 proposed fj > 0.1 max fj 80.5 17.3 35.5 8.Low Magnitude 122.7 24.5 54.8 7.fj > 0.1 max fj 80.5 15.9 35.5 5.Low Magnitude 122.7 24.5 54.8 3.fj > 0.1 max fj 80.5 14.6 35.5 4.Low Magnitude 122.7 22.1 54.8 2.OAAT M = 8 DAM fj > 0.1 max fj 58.1 17.7 Low Magnitude 116.8 59.3 this method (for example, improving the initial estimate based on information about the relative coil patterns as suggested above) may yield even greater improvements in reduced scan regularized estimation. We tested the improvement seen by using the correct slice prole for estimation versus assuming an ideal sinc pulse prole. The results are summarized in Table 5.3. We see that using the correct slice prole gives slightly better error results for OAAT for the truncated sinc pulse (though curiously not for the gaussian pulse). This holds true for pixels with a high signal value as well for signal voids. We would not expect a very large difference for OAAT, because the ip angles are very small and the difference in F at these ip angles is also small. For LOO we see bigger relative differences, especially among the gaussian pulse. This is as we would expect, as the gaussian pulse differs greatly from the ideal sinc pulse at the ip angles seen in LOO. Thus, the improved slice prole is most advantageous at higher ip angles.
Table 5.3: Simulation NRMSE (%) using the correct slice prole for estimation estimation Excitation Assumed pulse pulse Trunc sinc Trunc sinc OAAT M = 8 fj > 0.1 max fj Trunc sinc Ideal sinc OAAT M = 8 fj > 0.1 max fj Trunc sinc Trunc sinc OAAT M = 8 Low Mag Trunc sinc Ideal sinc OAAT M = 8 Low Mag Gaussian Gaussian OAAT M = 8 fj > 0.1 max fj Gaussian Ideal sinc OAAT M = 8 fj > 0.1 max fj Gaussian Gaussian OAAT M = 8 Low Mag Gaussian Ideal sinc OAAT M = 8 Low Mag Trunc sinc Trun sinc LOO M = 8 fj > 0.1 max fj Trunc sinc Ideal sinc LOO M = 8 fj > 0.1 max fj Trunc sinc Trunc sinc LOO M = 8 Low Mag Trunc sinc Ideal sinc LOO M = 8 Low Mag Gaussian Gaussian LOO M = 8 fj > 0.1 max fj Gaussian Ideal sinc LOO M = 8 fj > 0.1 max fj Gaussian Gaussian LOO M = 8 Low Mag Gaussian Ideal sinc LOO M = 8 Low Mag

+ relaxation than on the B1 map.
[124] also uses a parametric method using the the standard SSI (FLASH method) to
+ estimate both T1 and f , but assumes that B1 is known. Like [11], the initial estimate is
found via regression and then non-linear regression is used to nd a more accurate solution under the natural constraints of the T1 and f. The need for at least three or four ip angles, in contrast to the standard two, to obtain accuracy is also discussed. [122] generalizes the model for both the AFI (actual ip angle imaging) and SSI methods to create a new MTM model using multiple repetition times. This new model, which averages multiple measurements, can be solved either analytically or numerically using
+ model tting. The model has the benet of giving accurate maps of both B1 and T1 when
solved analytically, although the accuracy of the T1 maps was not further analyzed because
+ of the focus on B1 maps. The Cramer Rao bound was also used to quickly determine
optimal scan parameters. [137] analyzes the importance of spoiling on measurement effects for both the AFI
+ and the VFA methods for B1 mapping (corrected for T1 ). Because diffusion is critical in
RF-spoiled sequences, the optimal angles and measurement of T1 are strongly dependent on the strength of the spoiling. Spoiling is not considered in this thesis, but obviously is necessary to consider in the future work.
6.3 Limitations of Current Methods and Possible Solutions

+ B1 inhomogeneity

+ As mentioned previously, B1 inhomogeneity is the primary source of error in T1 mea-
surements and must be corrected for when using the SSI method. Some of the basic meth+ ods have been explained in the previous section. Almost all methods blur calculated B1
maps to improve SNR without a solid understanding of the effect (i.e., the FWHM of the
+ blur, the effect on T1 calculation). Additionally, many methods nd a low-resolution B1
map or one based on phantoms. Additionally, all methods estimate only the magnitude of
+ the B1 map and ignore the fact that it is a complex quantity. The ideal method would create + an accurate, pixel-by-pixel, in-vivo complex B1 map with high SNR without indiscrimi-
nant blurring and would not require a separate scan.
Slice prole effects, Bloch equation non-linearity, and ip angle miscalibration
Slice selective RF pulses would ideally have a rectangular shape, exciting only the spins in the desired slice, but this is not achievable in practice. Real slice proles have varying ip angles over the slice. Because the measured signal in MRI integrates over the volume, the varying ip angles due to the slice prole can cause error in the accompanying T1 measurements. In addition, spins do not behave linearly in the presence of an RF eld;

+ Another problem with current methods is using the signal equations to actually nd B1
and T1. There are no closed form solutions for either variable using the SSI signal model without making many simplifying assumptions and approximations, e.g., [53]. Many approximations, for example, by linearizing the exponential, are only valid in a small range
and are inefcient because of the low signal SNR in that range. The main techniques use regression, least-squares, curve tting, or error-propagation. The conventional technique involves transforming the data (6.6) and then performing a linear least-squares (LLS) t on the transformed data. Because LLS minimizes residuals between the transformed data and the transformed predicted data based on the estimated variable, the tting no longer matches the cost function and is suboptimal, resulting in a biased estimator with low accuracy. Using weighted least squares with careful choice of the weights transforms the residuals to match those of the original non-transformed cost function [14]. LLS gives an estimate of E1 and conventional methods still must estimate T1 in an additional step. Ideally, the estimation techniques used would properly account for noise in the signal model and therefore, avoid problematic ratios and subtractions. Also, the technique would use the most accurate, unbiased methods available.
6.4 Model Selection: A CRB approach
+ Many1 methods have been developed recently that jointly estimate both B1 and T1
(see Section 6.2.4). Making an informed choice between the wide variety of pulse se+ quences where relaxation effects and B1 inhomogeneity feature prominently remains an + open problem. Analysis of the accuracy and precision possible in B1 and T1 estimates and
the inherent trade offs can aid this selection.
+ In this section, we rst construct a general model for joint B1 , T1 mapping. We then use
the Cram r Rao Bound to analyze the lowest possible variance for unbiased joint estimation e
+ of B1 and T1 using several specic pulse sequences. We investigate the variance of both + estimates over a range of B1 and T1 values. We also use this analysis to help optimize
timing and ip angle parameters for each pulse sequence. This analysis extends the large body of research on optimization of parameters and precision for T1 estimation (e.g., [25,
+ 26, 38]) to include joint B1 and T1 estimation. Joint estimation methods usually require a
This section is partially based on [40].
+ higher resolution of one quantity (B1 or T1 ), often utilizing a map of the second quantity
for greater accuracy in the initial mapping. For example, [116] concentrates on T1 mapping
+ + while utilizing a B1 map, while [11] concentrates on B1 mapping but also estimates a
lower-accuracy T1 map. The trade offs and analysis from this section allows comparison of

+ + (reported in Table 6.3) for the joint B1 , T1 estimation is compared to estimating only B1
using the regularization estimation explained in Chapter V, referring to this estimator as the previous estimate. That method ignores T1 effects, as if TR =. We note that the initial T1 estimate here is the conventional T1 estimate for the SSI method described in Section 6.2.2. First, we compared at a high SNR of 60 dB the OAAT method (shown in Fig. 6.28, Fig. 6.29, Fig. 6.30, and Fig. 6.31) and LOO method (shown in Fig. 6.32. Fig. 6.33, Fig. 6.34, and Fig. 6.35.) We note, in regards to the SNR, some current T1 mapping papers report SNRs ranging from 100 - 200 dB in the brain [14] and start to see signicant bias at about 60 dB [16], though these methods use a much lower TR (T R < 10 ms). We used only 12 measurements because both methods perform well, with the most notable error in the T1 map in OAAT in Fig. 6.30. We still see some small drop-out in the T1 map for LOO Fig. 6.34, though the T1 map is denitely improved. We also compared these methods when used at a lower SNR of 30 dB. Here, the OAAT method struggled with only 12 measurements (gures not shown), so we used 16 measurements. Even at 16 measurements, the noise necessitated using the previous method with a small number of iterations as the initial guess. The nal f (see Fig. 6.39) and T1 (see Fig. 6.38) strongly underestimate the interior of the brain which causes some corruption of
+ the B1 magnitude maps (see Fig. 6.36). Clearly, using LOO improves all estimates, shown
in Fig. 6.40, Fig. 6.41, Fig. 6.42, and Fig. 6.43. There is still some overestimation of T1 along the skull, but overall the estimates perform well at the lower SNR and with only 12 measurements. The LOO method works reasonably well at smaller SNRs (results for 20 dB shown in Table 6.3, gures not shown). Overall, the simulation results shows that the proposed method works well, especially
True |B1| Initial est. |B1| 1.0.5 Initial error Previous error 1 87

0.0.1 0.0.1

Previous est. |B1| 1 1.Final est. |B1| 1 0.5

optimal selection of tip angles and repetition times to minimize scan time while achieving a low NRMSE. We plan to further investigate the spatial resolution, especially for the object, 175
and with multiple coils. The model and estimators in this paper provide smooth, less noisy estimates that incorporate T1 effects and greater repetition times selection that allow for a possibly shorter scan time and concurrent T1 estimation.

CHAPTER VII

Conclusion and Future Work
Due to the high eld strength and temporal requirements in modern MRI, eld maps of the main eld B0 and of the radio frequency eld B1 are required for pulse design and image correction. Many current estimators for these elds are heuristic and not based on a comprehensive statistical model. This thesis proposed three new penalized-likelihood estimators based on statistical models. The eld map estimator uses multiple scans and shows an empirical improvement with an improvement in RMSE over the conventional or penalized-likelihood estimator with only two scans. The B1 estimator uses a model that accounts for a multiple coil design and includes slice-selection effects and allows for any number of arbitrary tip angles, an improvement over the double angle conventional esti+ mator. The estimator additionally estimates both the magnitude and phase. The joint B1 ,
T1 estimator accounts for a multiple coil design and allows for any number of arbitrary tip angles and repetition times while estimating the magnitude and phase for each coil and a T1 map. Simulation and MRI studies show the reduced noise and (for the simulation) reduced RMSE when using each new PL estimator over the conventional estimator. Using PL estimators and a statistical model yields better results than just using conventional estimators. These estimators make smoother, less noisy estimates for B0 and B1 and T1 maps for use in pulse design and image correction. Ultimately, each of these methods is a tool that can only help answer the true question of
mapping: the best use of scan time to create the most accurate map. While the preliminary CRB analyses in Section 4.2.10 and in Section 6.4 help guide the user to fortuitous selections of imaging parameters (the echo time in eld map estimation or the repetition times
+ and tip angles in joint B1 , T1 mapping), neither nds the ideal use of scan time for esti-
mating certain parameters. Indeed, most estimation in the literature balances between short scan time and accuracy in estimates, often using approximations that allow for shortened scan time at the cost of accuracy.

(I.11)

APPENDIX J
+ B1 , T1: Derivatives of F
This section derives the derivatives of F in terms of H and its derivatives, which are tabulated (when using a non-ideal slice prole or incorporating B0 eld inhomogeneity) or explicitly derived (assuming an ideal sinc prole as the experiments in this thesis) as explained in Section 6.5.1. To simplify the derivations, we use an equivalent denition for F. We dene z = a + ib and rewrite (6.26) as follows: F (z, t) = ez H( a2 + b2 , t) = ib a + a2 + b 2 a2 + b 2 H( a2 + b2 , t).
The nal line follows from using Eulers formula for ez and expressing cos and sin using the Pythagorean Theorem. Then, a b HR ( a2 + b2 , t) HI ( a2 + b2 , t) a2 + b 2 a2 + b 2 a b HR ( a2 + b2 , t) + HI ( a2 + b2 , t). FI (a, b, t) = a2 + b 2 a2 + b 2

FR (a, b, t) =

Using these equations, we can then nd the derivatives d aa2 FR (a, b, t) = HR (|z| , t) 2 HR (|z| , t) + 2 [10] HR (|z| , t) da |z| |z| |z| ab ab + 2 HI (|z| , t) 2 [10] HI (|z| , t) |z| |z| 1 ab ab d FR (a, b, t) = HI (|z| , t) 2 HR (|z| , t) + 2 [10] HR (|z| , t) db |z| |z| |z| b b + 2 HI (|z| , t) 2 [10] HI (|z| , t) |z| |z| 2 a2 d 1 a FI (a, b, t) = HI (|z| , t) 2 HI (|z| , t) + 2 [10] HI (|z| , t) da |z| |z| |z| ab ab 2 HR (|z| , t) + 2 [10] HR (|z| , t) |z| |z| 1 ab ab d FI (a, b, t) = HR (|z| , t) 2 HI (|z| , t) + 2 [10] HI (|z| , t) db |z| |z| |z| b b 2 HR (|z| , t) + 2 [10] HR (|z| , t) |z| |z| d FR (a, b, t) = cos(z) [01] HR (|z| , t) sin(z) [01] HI (|z| , t) dt d FI (a, b, t) = cos(z) [01] HI (|z| , t) + sin(z) [01] HR (|z| , t) dt = exp(iz) [01] H(|z| , t). = exp(iz) [01] H(|z| , t)

APPENDIX K

+ + B1 , T1: Initial estimate for B1 , T1, and f
+ The algorithm requires a good initial estimate for B1 and T1 to ensure the iterates
descends to a good local minimum. We need an initial estimate for the magnitude and
+ phase of the B1 maps, the T1 map, and the object.
Simple approach - Assume T R = We note that the standard approach for T1 estimation (based on (6.14) assumes that the
+ ip angle (and thus, the B1 map) is known. Then, T1 is estimated from the data using a
transformation of the points and using a least-squares t. Therefore, one approach to joint

longer apply. Ignoring these non-uniformities can cause signicant distortions. Accurate maps of the main and RF transmit coil eld inhomogeneity are required for accurate pulse design and imaging. Standard estimation methods yield noisy maps, particularly in image regions having low spin density, and ignore other important factors, such as slice selection effects in B1 mapping and T2 effects in B0 mapping. This thesis uses more accurate signal models for the MR scans to derive iterative regularized estimators that show improvements over the conventional unregularized methods through Cram er-Rao Bound analysis, simulations, and real MR data. In fast MR imaging with long readout times, eld inhomogeneity causes image distortion and blurring. This thesis rst describes regularized methods for estimation of the off-resonance frequency at each voxel from two or more MR scans having different echo times, using algorithms that decrease monotonically a regularized least-squares cost function.
A second challenge is that RF transmit coils produce non-uniform eld strengths, so an excitation pulse will produce tip angles that vary substantially over the eld of view. This

23.Pixel defect---The LCD panel is a very high technology product, giving you finely detailed pictures. Occasionally, a few non-active pixels may appear on the screen as a fixed point of blue, green or red. Please note that this does not affect the performance of your product.

Preparations

Using the Remote Control
<Use the remote control by pointing it towards the remote sensor window of the set.
Objects between the remote control and sensor window may prevent proper operation. Note: the illustration is for your reference only, the remote sensor may locate differently with different model.
Cautions regarding use of remote control
<Do not expose the remote control to shock. In addition, do not expose the remote
control to liquids, and do not place in an area with high humidity.
<Do not install or place the remote control under direct sunlight. The heat may cause deformation of the unit. <The remote control may not work properly if the remote sensor window of the main unit is under direct sunlight or strong
lighting. In such a case, change the angle of the lighting or LCD TV set, or operate the remote control closer to the remote sensor window.
Batteries for the Remote Control
If the remote control fails to operate the LCD TV functions, replace the batteries in the remote control.
1 Open the battery cover.
Insert two size-AAA batteries.
Replace the cover and slide in reverse until the lock snaps.
<(Slide the cover while pressing
<(Place the batteries with their terminals

down.)

corresponding to the (+) and () indications in the battery compartment.)
Precaution on battery use Improper use of batteries can result in a leakage of chemicals and/or explosion. Be sure to follow the instructions below.
<Place batteries with their terminals corresponding to the (+) and () indications. <Different types of batteries have different characteristics. Do not mix batteries of different types. <Do not mix old and new batteries. Mixing old and new batteries can shorten the life of new batteries and/or cause old

batteries to leak chemicals.
<Remove batteries as soon as they are non-operable. Chemicals that leak from batteries can cause a rash. If chemical leakage
is found, wipe with a cloth.
<The batteries supplied with the product may have a shorter life expectancy due to storage conditions. <If the remote control is not used for an extended period of time, remove the batteries from the remote control.

Preparations (continued)

Power connection
Household power outlet Plug into AC outlet. Connect to the DC input socket of the product. Be sure to insert plug into the end fully and confirm it is reliable

AC cord

AC adapter
1. Connecting the female plug to the AC socket on AC adapter. 2. Connecting the AC adapter to the DC power input of the set. 3. Connecting the male plug to the wall outlet as illustrated. Note:
<Always turn the Power Switch of the LCD TV set to OFF when connecting the AC adapter. <This product should be operated only from the type of power source indicated on the marking label. <Always unplug the AC adapter from the product and power outlet when not using for a long period of time.

Antenna Connection

INSTALL the unit in a room where direct light will not fall upon the screen. Total darkness or a reflection on the picture screen may cause eyestrain. Soft and indirect lighting is recommended for comfortable viewing. Optimum reception of color requires a good signal and will generally mean that an outdoor aerial must be used. The exact type and position of the aerial will depend upon your particular area.
75-ohm coaxial cable (round cable)
To antenna input terminal ( ) 300-ohm twin-lead flat feeder
Note: It is recommended that the 75-ohm coaxial cable be used to eliminate interference and noise which may occur due to radio wave conditions. The aerial cable should not be bundled with the power cord and the like.
Identification of Controls

Main unit (front view)

Control Panel

SOURCE

Menu Select Input Signal
Channel Down/Up Volume Down/Up

Power On/ Standby

Speaker
Remote Sensor Power Indicator
A yellow indicator lights when the power is on and a red indicator lights when in the standby mode.
1. SOURCE To access the SOURCE select menu 2. MENU Press this button to access the MENU main page. 3. CH +/Change the TV channel. In OSD Menu, press these buttons to choose the OSD items. 4. VOL +/Increase or decrease the sound volume level. In OSD Menu, press these buttons to adjust the value or setting of each item 5. POWER Press this button to turn the unit ON from STANDBY mode. Press it again to turn the set back to STANDBY.

Note: SOURCE, MENU, CH+/-, VOL+/- and POWER on the main unit have the same functions as the corresponding buttons on the remote control. This operation manual provides a description based on operating functions with the remote control.
Identification of Controls (continued)

Main unit (rear view)

DC 12V

PC - AUDIO S-VIDEO

L AUDIO R AV2 IN

L AUDIO R AV OUT

AV1 (SCART)
1. DC POWER input Connect to the DC output of the Power Adapter. 2. VGA input /Audio in Connect to the VGA/audio output 3.5mm jacks on your PC. 3. S-Video input Receive a S-Video signal from external source such as VCR or DVD player. 4. AV 2 inputs (Video, Audio L, R) Receive video/audio signals from external sources such as VCR or DVD player. 5. Headphone 3.5mm jack 6. AV outputs (Video, Audio L, R) Connect to the VCR input jacks to record programs. 7. 21-pin Euro-SCART (RGB) interface/AV1input and Scart output 8. Antenna input

Remote Control

1. POWER Turn the unit on or standby 2. CH+/-,VOL+/CH+/- ---Use to switch channels; To display previous/next page (Teletext) VOL+/----Use to adjust volume; In MENU operation, use CH+/- to select item and VOL+/- to adjust selected item 3. MENU To access the MENU main page 4. STEREO To select stereo, mono or bilingual (select Left, Right or Stereo in AV or D-Sub mode) Red button (teletext) 5. SLEEP To access sleep time setting menu bar Green button (teletext) 6. Teletext To enter/exit teletext mode 7. P.P To select preset picture mode 8. 0~9 digit buttons, -/-- button 0~9: direct channel select; -/-- button: one digit/two digit/three digit channel number selector 9. SOURCE To access source select menu 10. MUTE Sound mute 11. To confirm your operation or setting or to exit menu To display channel status or program table 12. To increase brightness Blue button (teletext) 13. To decrease brightness Yellow button (teletext) 14. RETURN To quickly jump between current channel and last selected channel. 15. S.P To select preset sound mode
Flip the cover, open in the direction of the arrow. 7

16. SCALER MODE To select screen aspect ratio 17. HEADPHONE To load preset sound settings for headphone 18. REVEAL To display hidden information such as solutions to riddles and puzzles. (Teletext) 19. REAL CLOCK To access real time quickly without entering teletext 20. INDEX To display index page (Teletext) 21. MIX To superimpose text on a TV picture (Teletext) 22. POS To quickly select OSD position 23. AUTO To adjust picture automatically 24. TIMER To access timer menu bar 25. SIZE To expand top half or bottom half page of current screen (Teletext) 26. HOLD To hold the current teletext page temporarily (Teletext) 27. To increase contrast (Teletext) 28. To decrease contrast (Teletext)

Connections

Cautions before connecting
Carefully check the terminals for position and type before making any connections. The illustration of the external equipment may be different depending on your model. Loose connectors can result in image or color problems. Make sure that all connectors are securely inserted into their terminals. Refer to the operating manual of the external device as well. When connecting an external device, turn the power off to avoid any issues.
Connect 21-pin EURO-SCART (AV1 IN) Interface on the rear of the Unit

Decoder

21-pin Euro-SCART connector
21-pin Euro-SCART (RGB) 1. Audio right output 2. Audio right input 3. Audio left output 4. Common earth for audio 5. Earth for blue 6. Audio left input 7. BLUE input 8. AV control 9. Earth for green 10. Not used 11. GREEN input 12. Not used 13. Earth for red 14. Not used 15. RED input 16. RGB control 17. Earth for video 18. Earth for RGB control 19. Video output 20. Video input (PAL/SECAM/NTSC) 21. Plug shield

How to connect:

Connect the 21-pin Euro-SCART between the unit and decoder.
TV AV1 AV2 SVIDEO D-Sub Video1

To watch decoder

1. Turn on your LCD TV , press Source button on the remote control. 2. Press CH+/- to select AV1 and press VOL+ to confirm or wait for about 5 seconds. 3. Turn on the decoder. 4. Set the Scart Output item (in System menu) to TV and select a proper channel through the Scart Output Channel item. For detailed operation, please refer to [Setting Scart Output] in page19.

Connections (continued)

Connect Audio/Video inputs (AV2 IN) Interface on the rear of the Unit

S-Video cable

Video Audio cable cable
Yellow (VIDEO) White (AUDIO L/MONO) Red (AUDIO R )
Examples of external devices that can be connected

DVD player

Video Camera

Home video game system

The unit provides Audio/Video inputs for you to connect external devices such as VCR, Video Camera, DVD Player or Home video game system.
Connect the Audio/Video cables between the Audio (L/R)/Video jacks on the unit and external devices.
TV AV1 AV2 SVIDEO D-Sub Video2
To view signal from AV2 inputs
1. Turn on your LCD TV , press Source button on the remote control. 2. Press CH+/- to select AV2 or SVIDEO and press VOL+ to confirm or wait for about 5 seconds. 3. Turn on the external device and play it.
Note: For better video, you can use the S-video terminal if your source supports it. If you are going to use S-video terminal, please select SVIDEO instead of AV2 in SOURCE menu.
The Video input terminal on the AV2 IN and the S-Video input terminal share the same Audio input terminals.
Connect a PC Interface on the rear of the Unit

VGA cable

Audio cable
Connect a VGA cable between the VGA jack on the PC and the VGA input jack on the unit. Connect an Audio cable between the AUDIO output on the PC and AUDIO input 3.5mm jack on the unit.

To Watch the PC screen

1. Turn on your LCD TV , press Source button on the remote control. 2. Press CH+/- to select D-Sub (VGA IN). 3. Press VOL+ to confirm or wait for about 5 seconds. 4. Turn on your PC and check for PC system requirements. 5. Adjust the PC screen.
TV AV1 AV2 SVIDEO D-Sub D-Sub
Connect a VCR for recording Interface on the rear of the Unit

Video cable

VCR for recording

ANT OUT

AV OUT
VIDEO L AUDIO R S - VIDEO

ANT IN

VIDEO L AUDIO R

Rear of the VCR

Connect the Audio/Video cables between the Audio (L/R)/Video jacks on the unit and VCR.
- or Connect the 21-pin Euro-Scart between the Scart Interface on the unit and VCR.
To record program (using AV terminals)
1. Turn on your LCD TV, select a program you wish to record. 2. Turn on your VCR, insert a videotape for recording. 3. Press the Record button to begin recording.
To record program (using Scart interface)

Press MENU to display the menu main page. Press VOL+/- repeatedly to display the Information menu page.
Video Type: PAL/50Hz Mode: 3
Press the button on the remote control, the unit will display current status information such as channel number, channel name (if available), etc.

Memorizing the Channels

Your LCD TV can memorize and store all of the available channels. After the available channels are memorized, use CH+/to scan through the available channels.
Storing Channels in Memory Automatically
Press MENU to display the menu main page. Press VOL+/- repeatedly to display Search menu page.
Channel Number Add/Erase Search Control Hand Search Auto Search Fine Tune Edit Channel Name ADD French

1 ERASE Other

Press CH+/- repeatedly to select Search Control item. Press VOL+/- to select French or Other. Please select French in France, or select Other in other country. The unit searches and memories the SECAM-L/L programs if you select French. Press CH+/- repeatedly to select Auto Search item.

VHFL Searching

Press VOL+/- to store channels in memory automatically. The unit will begin memorizing all of the available channels. Note: <Before Auto Search, we recommend you to set a proper Sound system according to your local area. (Please refer to [Setting Color and Sound System] in this user manual) <The Search Control item may not available depending on your model. If the Search Control item is found unavailable, the steps 3 & 4 can be skipped. <The process of Auto Search will be stopped if you press the MENU button. <Channel Labeling: After the Auto Search is finished, the names of those programs with teletext will be added into the program table automatically.

110MHz

Storing Channels in Memory Manually
Press MENU to display the menu main page. Press VOL+/- repeatedly to display Search Press CH+/- to select Channel Number item. Press VOL+/- repeatedly to select a channel number you want to store. Press CH+/- repeatedly to select Hand Search item. Press VOL+/- to start searching. When pressing VOL-, the unit searches towards lower frequency; When pressing VOL+, the unit searches towards higher frequency. When a program is located, the searching stops and the program is stored in the specified channel number. Repeat the above steps if you want to store another program in other channel number. menu page.

9 ERASE Other

VHFH SearchEnd

224MHz

Memorizing the Channels (continued)

Manual Fine Tuning

Channel Number Add/Erase Search Control Hand Search Auto Search ADD French
Press CH+/- repeatedly to select Fine tune item. Press VOL+/- to fine tune till the best possible picture and sound are obtained.
Fine Tune Edit Channel Name

VHFH FineTune

Adding and Erasing Channels
Use number buttons to directly select a channel that will be added or erased. Press MENU to display the menu main page. Press VOL+/- repeatedly to display Search menu page.
Press CH+/- repeatedly to select Add/Erase item. Press VOL+/- to select ADD or ERASE. When ERASE is selected, the selected channel will be erased. When ADD is selected, the selected channel will be added. The erased channels can not be selected by using the CH+/buttons unless you use number buttons to input channel number directly.

Editing Channel Name

Use number buttons to directly select a channel you want to edit its name. Press MENU to display the menu main page. Press VOL+/- repeatedly to display Search menu page.
Press CH+/- repeatedly to select Edit Channel Name item. Press VOL+/- to access channel name editing page. Use CH+/- to select character. Use VOL+/- to move cursor right or left. Use OK to confirm and return.
Set Channel Name Press <OK> to Confirm Changes
Note: This LCD TV enables you to edit names for up to 112 channels.

Basic Operations

Changing Channels
Using the Channel Buttons (CH+ or CH- ) Press the CH+ or CH- to change channels. When you press the CH+ or CH-, the unit changes channels in sequence. You will see all the channels that the unit has memorized. You will not see channels that were erased or unavailable. Direct Accessing Channels Press the number buttons to go directly to a channel. To select a one-digit channel: Input the channel using the 0-9 number button directly. To select a two-digit channel: Press -/-- to display "--", then input the channel number. To select a three-digit channel: Press -/-- to display "---", then input the channel number. Note: Be sure to enter the channel within 3 seconds. When you use the number buttons, You may directly select channels that were erased. Using the Return button
Press this button to switch between the current channel and the previous channel. Using the Program Table This LCD TV enables you to edit names for up to 112 channels and those channels with name can be listed by program table. You can quickly select channel by using program table.

Press the

button twice to display the program table.
AAA__ 1 _____5 CCTV_9 _____13 _____17
BBB__ 2 _____6 _____10 _____14 _____18
ABC__ 3 _____7 _____11 _____15 _____19
_____4 _____8 _____12 _____16 _____20

Bright

WIDE Mode Settings
Press MENU to display the menu main page. Press VOL+/- repeatedly to display System menu page.
Blue Screen Scaler Mode Video Standard Audio Standard Sleep Time Scart Output
Off Fill All Auto B/G 0 minutes Monitor
Press CH+/- repeatedly to select Scaler Mode item. Press VOL+/- to select Fill All or Wide.
Fill All The Fill All mode stretches the input vertically and horizontally to fill the display. Wide If the signal is 16:9, this mode acts as if it is in Normal mode. If the signal is 4:3, the top and bottom sides of the picture are compressed with a black stripe on top and bottom of the screen. Note: You can quickly select the Scaler mode by using the button.

Setting System

Setting Blue Screen Background
Press MENU to display the menu main page. Press VOL+/- to display System menu page.
Blue Screen Scaler Mode Video Standard Audio Standard Sleep Time Scart Ouput
Press CH+/- repeatedly to select Blue Screen item. Press VOL+/- to select On or Off.
When there is no input signal, a blue background appears if you set this item to On. To cancel this function, please set this item to Off. If there is no input signal and the Blue Screen is turned on, the unit goes to Standby in 15 minutes automatically.
Setting Color and Sound System
Press CH+/- repeatedly to select Video Standard item. Press VOL+/- to select a proper color system. Normally, set the Video Standard to Auto. Press CH+/- repeatedly to select Audio Standard item. Press VOL+/- to select a proper sound system.

Setting Scart Output

The unit allows you to output TV or AV signal through the Scart.
With the unit working in AV mode, press MENU to display the menu main page. Press VOL+/- to display System menu page.
Press CH+/- repeatedly to select Scart Ouput item. Press VOL+/- to select TV or Monitor. When you set this item to Monitor, the Scart outputs current screen picture. When you set this item to TV, you select channel through setting the Scart Output Channel item, then the Scart outputs TV program of the specified channel.
Blue Screen Scaler Mode Video Standard Audio Standard Sleep Time Scart Output Scart Output Channel
Off Fill All Auto B/G 0 minutes TV 9
Note: <The Scart Output Channel item is available only when the unit works in AV1 mode and the Scart Ouput item is set to TV. <The Scart Output item is unavailable in TV and VGA mode.

Setting Sound

Adjusting Sound Settings
Press MENU to display the menu main page. Press VOL+/- repeatedly to display Audio menu page.
Bass Treble Balance Stereo Mono On On Off Off VBE Spatial
Press CH+/- to select item you are going to change. Press VOL+/- to adjust the value of the item.
Adjustment item Bass Treble Balance VBE Spatial

Choice/Value range

---Off Off On On
Description Press Adjusting Bass sound effect. Adjusting Treble sound effect. Adjusting sound Balance

Right (

To turn on/off Virtual Bass Enhancement To turn on/off space surround effect ( * The choices differ depending on whether or not a NICAM or IGR signal is received.) For detail, please refer to next two pages.

Stereo

NICAM Mono Mono NICAM Stereo Mono NICAM A NICAM B Mono Mono FM Stereo FM A FM B
VBE: turn on this function if the signal has a poor bass effect and you hope the unit enhance the bass effect. Spatial : turn on this function when the unit is located in a small room and you hope the unit produce a space

surround effect.

Using the Preset Sound Mode
There are three preset sound modes (Movie, Music and News) and one userset sound mode (User). You can quickly select sound mode by pressing the S.P button. Each preset mode has its own Audio settings (Treble and Bass). Movie: Select for a movie program picture. Music: Select for a music program. News: Select for a speech or conversation program. Adjusted settings are stored in User mode.
Loading Preset Sound Settings for Headphone
In normal cases, the adjustments of sound settings is for the speakers only. If you want to adjust sound settings for the headphone independently, press the Headphone button, a icon appears on the screen and the unit enables you to adjust sound settings for the headphone. The adjusted sound settings will be saved as headphone preset. Next time when you listen with headphone, press this button to quickly load preset sound settings for headphone.
Note: <When you do not listen with headphone, please be sure to press the Headphone buttonto quit headphone mode, the icon should disappear. <When you connect a headphone to the headphone jack to enjoy the sound through headphone, the sound of speaker will be muted.
Setting Sound (continued)
NICAM Broadcast Selection
This enables the selection of the reception mode when receiving a NICAM signal. To let you enjoy NICAM broadcasting, this LCD TV set receives NICAM system stereo, bilingual, and monaural broadcasts. Stereo When the LCD TV set is receiving a stereo broadcast. Each time you press STEREO , the mode changes between NICAM Stereo and Mono. When Mono is selected, the TV sound is output monaurally.

NICAM Stereo

Bilingual When the LCD TV set is receiving a bilingual programme. Each time you press STEREO , the mode changes as follows:

NICAM A

NICAM B
Monaural When the LCD TV set is receiving a NICAM monaural broadcast. Each time you press STEREO , the mode changes between NICAM Mono and Mono.

NICAM Mono

IGR (German stereo system) Broadcasts
This enables the selection of the reception mode when receiving an IGR signal. To let you enjoy IGR broadcasting, this LCD TV set receives IGR system stereo and bilingual broadcasts. Stereo When the LCD TV set is receiving a stereo broadcast. Each time you press STEREO , the mode changes between FM Stereo and Mono. When Mono is selected, the TV sound is output monaurally.

FM Stereo

Bilingual When the LCD TV set is receiving a bilingual programme. Each time you press STEREO and FM B. , the mode changes between FM A
Note: <The selection of Stereo Mode can be performed by directly pressing the STEREO
<With working in AV (SVIDEO) or VGA mode, the selection of
button or using the Audio
Stereo Mode is between Left, Right or Stereo.

Adjustment in VGA Mode

With working in VGA mode, this unit allows you to perform several adjustments.
Press MENU to display the Picture

menu main page.

brightness contrast color temp
If the Picture menu does not display, press VOL+/- repeatedly to display Picture menu page. Press CH+/- to select the item you wish to change. Press VOL+/- to adjust the value of the item. Press MENU to exit.
Adjustment item brightness contrast Color temp customer temp setting Press VOLLess bright Decrease contrast Value range 5000K, 7300K, 9300K, User
5000K 7300K 9300K user customer temp setting
Press VOL+ More bright Increase contrast
Press VOL+ or VOL- to access the submenu to adjust RGB individually.
Note: You can quickly adjust the brightness by using the

button.

Customizing the Color Temp
You need to set the Color Temp item to User if you want customize color temp by yourself.
Press MENU to display the Picture menu main page. If the Picture menu does not display, press VOL+/- repeatedly to display Picture menu page. Press CH+/- repeatedly to select the customer temp setting item. Press VOL+/- to access the submenu. Press CH+/- to select the User red, User green or User blue item. Press VOL+/- to adjust item.

Press the button to expand the top half of the display. Press it again to expand the bottom half of the display. Press it once more to resume normal size. HOLD Press the button to stop updating Teletext pages automatically. Press it again to release hold mode. INDEX Press the button to display the index page. MIX Press the button to superimpose the teletext on a TV picture. Press it again to cancel. CONTRAST Press the buttons to adjust Teletext contrast.

Troubleshooting

Before calling for repair service, check the following items for possible remedies to an encountered symptom. Symptoms
Ghost or double images No power

Check item

This may be caused by obstruction to the antenna due to high rise buildings or hills. Using a highly directional antenna may improve the picture. Check that the AC power cord is plugged into the mains socket. Unplug the power cord, wait for 60 seconds. Then re-insert plug into the mains socket and turn on the unit again.

No picture

Check antenna connections at the rear of the unit to see if it is properly connected to the unit. ! Possible broadcast station trouble. Try another channel. ! Adjust the contrast and brightness settings. ! Select a correct input. ! Is a non-compatible signal being input?
Good picture but no sound
Increase the VOLUME. Check that the unit is not muted.
Good sound but poor color Poor picture
Adjust the contrast, color and brightness settings. Sometimes, poor picture quality occurs when having activated an S-VHS camera or camcorder connected and having connected another peripheral at the same time. In this case, switch off one of the other peripherals. ! Check whether the room is too bright

Horizontal dotted line

This may be caused by electrical interference (e.g. hairdryer, nearby neon lights, etc.) ! Turn off the equipment.
Television not responding to remote control
Check whether the batteries are working. Replace if necessary. Clean the remote control sensor lens on the unit. ! Do not use the remote control under strong or fluorescent lighting. ! The batteries should be inserted with polarity (+, -) aligned.

! ! ! ! !

Snowy picture and noise No stable or not synchronized VGA picture No output from one of the speakers
Check the antenna connection. Check if you have selected the correct VGA mode in your PC. Adjust Balance in the Audio menu.

Note: If your problem is not solved, restart your TV by turning it off and then on again once. Never attempt to repair a defective TV yourself.

Specifications

Type Display Size diagonal Display Feature LCD panel Resolution Pixel Pitch Maximum Colors Brightness Contrast Viewing Angle Response Time TV System Channel Coverage System PAL-D/K PAL-B/G PAL-I SECAM-B/G SECAM-D/K SECAM-L SECAM-L Teletext Speaker AC Adapter Input AC Adapter Output Power Consumption Unit Weight (kg) Unit Dimensions (WxHxD) (mm) Terminals AV1 IN AV 2 IN S-VIDEO VGA/AUDIO IN AV OUT HEADPHONE Accessories 21-pin EURO-SCART VIDEO, AUDIO S-VIDEO D-Sub 15-pin/3.5mm Stereo mini phone Jack VIDEO, AUDIO 3.5mm Stereo mini phone Jack 1 User Manual, 1 AC adapter, 1 AC Power Cord 1 Remote Control, 2 AAA Batteries 1 Audio/Video Cable, Supported Display Format 640 x 480 @ 60Hz 720 x 400 @ 70Hz 800 x 600 @ 60Hz 800 x 600 @ 75Hz 640 x 480 @ 75Hz VHF 1-12 2-12 2-12 2-12 1-12 2-4 5-12 21-69 X~Z C57, S1~S41 TOP/FLOF TELETEXT (10 pages) 8W/6 ohm x 2 AC230~240V, 50Hz DC12V 55 W 9.x 444 x 192 UHF 13-57 21-69 21-69 21-69 21-69 X~Z+2, S1~S41 CATV Z1~Z37 X~Z+2, S1~S41 TFT-LCD 800 x600 0.51mm x 0.51mm 16,700,nit 500:1 120o/160o <16 ms PAL/SECAM BG, DK, PAL-I, SECAM-L, SECAM-L, LCD TV 20

Dimensional Drawings

637 71

380 444

178 192
Design and specifications are subject to change without notice.
PRINTED IN RECYCLED PAPER

604-L20H3S9-00