===== s012 ===== ===== Negative ROIs ===== | | loc1 | loc2 | loc3 | loc4 | | V1 | loc1_rV1_NEG( ) | loc2_lV1_NEG( ) | loc3_lV1_NEG( ) | loc4_rV1_NEG( ) | | | split up | | | | | V2 | loc1_rV2v_NEG( ) | loc2_lV2v_NEG( ) | loc3_lV2d_NEG( ) | loc4_rV2d_NEG( ) | | V3 | | | loc3_lV3d_NEG( ) | | | V4 | | | | | | V3A | | | loc3_lV3AB_NEG( ) | | | all | loc1_negs(21) | loc2_negs(40) | loc3_negs(54) | loc4_negs(43) | ==== each roi ==== === right side === loc1_rhV4(9): very nice curve. single gamma. loc1_rV1(27): two patches, W shaped. fits don't work. **loc1_rV3v(27): nice tiering, double gamma.** (rV2v: W shaped, lots of negative HDR. if we move inward, finding nice curves, but not obviously tiered.) **loc1_rV2v(18): took inside half of voxels in loc range. no tiering.** loc1_rV2v_NEG(6): negative HDRs. outward from the localizer range. loc1_rLO1(5): ugly. loc1_rV3AB(1): nothing… loc4_rV1(10): WRONG intersects rV2d and rV3d. based on localizer was not clean at all. lots of negative responses. so hand drew a circular patch around one particularly characteristic voxel → plausible HDR responses. not bad. loc4_rV1_NEG(34): “stripes” with the loc1_rV1 roi. negative HDR followed by positive HDR, no tiering. loc4_rV3d_doesntwork(10): overlaps with V1 voxels… not sure, but very good responses in these voxels. drew an roi of nice responses in rV1,V2d,V3d (33) then intersected V1, V2d, V3d. loc4_rV3d_intersect_niceResponses(6): nice enough HDR loc4_rV2d_intersect_niceResponses(11): nice enough HDR loc4_rV1_intersect_niceResponses(5): not so good… loc4_rV1_handDrawn(3): very noisy. fit1 1/2 doesn't work. **loc4_rV2d(3): works! hdrlen17** **loc4_rV3d(3): works! hdrlen17** **loc4_rV1(9): works. nice and round.** **loc1_rV1(6): works. split, but true to localizer. flat, then drops off to inverted gamma...** **loc1_rV3AB(10): ziggidy HDR, but in terms of localizer, correct. maybe intersects with V7, but patch seem probabile based on localizer.** **loc4_rV3AB(6): just outside of V3d, but strong HDR and doesn't look quite like V3d responses.** **loc1_hV4(7): good responses, kind of sharp response, but tiered.** **loc4_hV4(23): very strong, nice, fatter curve.** === left side === loc3_lhV4(14): beautiful HDR!! not tiered. loc2_lhV4(7): beautiful HDR!! 50% is highest... sort of tiered. loc2_lV3AB(3): not very good... lopsided W, no real peak or tiering. loc3_lV3AB(10): good sharp positive response, tiered. loc3_lV3AB_NEG(16): nice sharp W, left dip bigger. loc3_lV2d(11): double gamma, tiered. loc3_lV2d_NEG(24): left dips strongly, clean. loc3_lV3d(8): ok. double gamma. loc3_lV3d_NEG(17): typical. loc3_lV1(5): flat looking HDR response with huge post peak dip, so took the highest 5 voxles (peaks rose to 0.2 rather than flat 0). loc3_lV1_NEG(9):typical. loc3_lV7(0):deleted. loc2_lV3v(14):super nice, tiered, strong. loc2_lV2v(10): hand drawn, but good. a bit of splitting, but tolerable. loc2_lV2v_NEG(12): very negative, but not as overpowering as other regions. loc2_lV1(11): hand drawn, but doesnt intersect with loc3_lv! and round. thresholded. loc2_lV1_NEG(43): huge coverage... very W shaped, pretty prominent initial dip , then a second dip. loc2_lV7(9): really nice HDR, tiered closely. ==== concats ==== [loc1_rV2v loc4_rV2d] = none of the fits work. very W shaped. hdrlen needs to be about 20. [loc1_rV2v loc4_rV2d_intersect_niceResponses] = better than non intersect, but not very clean. [loc1_rV3v loc4_rV3d_intersect_niceRepsonses] = ok, but not very good. **[loc1_rV2v loc4_rV2d]**: works! cueOne cued is lower than uncued, though. hdrlen17, fit1 **[loc1_rV3v loc4_rV3d]**: works! hdrlen17, fit1 [loc1_rV1 loc4_rV1]: doesnt work unless you have loc2/3. the initial drop makes single gamma fit fail. == pedOnly == V1: hdrlen [13-16 xx 13-16 13-14], ampType[1 2 1 1] concat[13-15],fit1. hdrlen15,fit1. skewed left and underestimates, but fails otherwise. V2: hdrlen [15 15 xx 15], amp[1]. hdrlen15,amp1. V3: hdrlen15,amp1 == cuedUncued(7) == V3: {15,1} 20110208 V2: {15,1} 20110208 V1: {15,1} 20110208 V3A: {15,1} 20110209 similar to V123 hV4: {15,1} 20110209 not similar to V123 == cuedTarg(6) == V3: {15,1} 20110208 V2: {15,1} 20110208 V1: {15,1} 20110208 V3A: {15,1} 20110209 hV4: {15,1} 20110209