中心視野にテクスチャパタンを呈示する直前に高コントラストのマスク刺激を短時間呈示すると，物理的には何も呈示されていない一様輝度の周辺視野にもテクスチャパタンが広がり，テクスチャが視野全体を覆うように知覚される。筆者の発見したこの「マスク誘導充填（mask induced filling-out）」現象の生起要因について，まずテクスチャの規則性が充填の起こりやすさに及ぼす影響を検討した。実験では色と大きさの一様なドットテクスチャの，ドット位置のxy方向へのランダムなズレ（δ）を従属変数とし（δ=0, 2, 4, 6, 10 pixels），マスクの呈示によってこのテクスチャが周辺視野に主観的に拡散（充填）したかどうかを二肢強制選択法で観測した。この結果テクスチャのランダムなズレが大きくなるほど，充填が起こりにくくなることが示された。さらに，テクスチャの充填が観察されている間に物理的に呈示されているテクスチャパタンを入れ替える実験も行った。この結果，中心視野の物理的テクスチャの入れ替えと同時に，周辺視野に充填されている主観的テクスチャも入れ替わることが示された。
一般に充填現象は，物理的な刺激呈示の惹起する網膜〜LGN〜V1のレティノトピックな神経表象が拡散し，物理的には刺激の無い領域にまで伝搬することで生じると説明される（e.g. Komatsu, 2006, Nat Neurosci）。しかし上記の結果は中心視野のテクスチャに対する神経表象が複製され周辺視野に広がっていくと考える従来型の仮説では説明しがたい。例えば単に表象を複製するだけならテクスチャの規則性は充填の起こりやすさに影響しないはずであるし，また伝搬にかかる時間を考えると，中心視野のテクスチャを入れ替えた後しばらくは入れ替え前のテクスチャが視野周辺に充填して見え続けるはずである。本研究の結果は，従来の仮定よりも高次の視覚情報処理が充填現象の生起に関わっていることを示唆する。
When a high-contrast mask stimulus is presented for a short period just before a textural pattern in a central visual field, the textural pattern spreads to the peripheral visual field, where no texture has been physically presented, and the texture is perceived to cover the entire field of view. Regarding the origin factor of this "mask induced filling-out (MIF)" phenomenon discovered by the author, the effect of the regularity of the texture on the ease of filling was examined.
In the experiment, the dependent variable was the random deviation of the dot elements consisting of a uniform dot texture in an x-y direction (δ = 0, 2, 4, 6, 10 pixels), and whether this texture subjectively diffused to the peripheral visual field after presentation of the mask was tested by the two alternative forced-choice task. The results showed that the larger the random displacement of the texture, the less is the MIF. In addition, another experiment was conducted wherein the physically presented textural pattern was replaced with another pattern while texture filling was observed. The results showed that when the physical texture of the central visual field was replaced, the subjective texture filling the peripheral visual field was also immediately replaced.
In general, it has been explained that the filling phenomenon is caused by the diffusion of a retinotopic neural representation of the retina to the lateral geniculate nucleus to V1 which causes the presentation of the physical stimulus and the physical propagation of the stimulus to an unstimulated area (eg Komatsu, 2006, Nat Neurosci). However, the above results cannot be explained with the conventional hypotheses that the neural representation of the texture in the central visual field is replicated and spreads to the peripheral visual field. For example, texture regularity should not affect the likelihood of the occurrence of the MIF phenomenon if the representation is simply duplicated, and considering the time it takes to propagate, the texture before the replacement is for a while after replacing the texture of the central field of view The textural pattern should fill the field of view and considering the time required for propagation, the texture previously presented in the central field should remain in the peripheral visual field for a while after the central texture is replaced. The results of this study suggest that higher-order visual information processing than conventional assumed is involved in the occurrence of the filling phenomenon.