New imaging anatomy concepts of the conjoined longitudinal muscle
In 1852, XXXX thought that the conjoined longitudinal muscle originated in the longitudinal rectal muscle and strengthened by the levator ani muscle. In 1934, XXXX thought that it was also strengthened by the puborectalis muscle and the deep external sphincter muscle. In 1976, XXXX thought that it consisted of the longitudinal rectal muscle, the vertical portion of the levator ani muscle and the inferior extended portion of the puborectalis muscle [16]. Our MRI images showed that the longitudinal rectal muscle was located in anal canal area, and the vertical portion of the levator ani muscle located in perianal canal. There was a clear anatomical interface between these two structures and there was no evidence for fusion or conjunction [6]. Thus it can be seen that the concept of the conjoined longitudinal muscle is redundant, which has disadvantages to analyze the function of a single muscle.
New imaging anatomy concepts of anatomic layers of anal area
Anal area anatomically comprises five layers, including mucosa, submucosa, the internal anal sphincter, the conjoined longitudinal muscle and the external anal sphincter [15]. In the past two decades, since the 5-layer concept was stiffly followed in the researches on the anal area and pelvic floor, almost all investigators misread the layers of anal area on images. The misreading of the layers of anal area on MRI images is the most regretful thing. The biggest advantage of MRI is the ability of showing the muscles and the lipidic interspaces of the anal area on one image. The black muscles and the white interspaces are references to each other, which is unavailable by any other anatomical tools. Our MRI images showed that the anal area were divisible into 7 layers, including mucosa, submucosa, anal smooth muscle, the inner intersphincteric space, the vertical portion of the levator ani muscle, the outer intersphincteric space and the external anal sphincter [6]. Normally, the anal area can be divisible into 9 layers in MRI. However, the longitudinal rectal muscle is too thin and combined with the internal anal sphincter into one same anal smooth muscle layer, that’s why there are only 7 layers.
In the 39th edition of Gray’s Anatomy, inner-anus MRI images are adopted to show the anatomical layers of anal area. There are only five layers of structures marked on the images, and the external anal sphincter is the only layer marked correctly [13]. Such errors resulted from XXXX primary research, who is the inventor of inner-anus MRI coils. In 1995, she thought that the fifth layer on MRI image also included the anal area by error [17]. These mistakes have been widely adopted by successors[10, 11, 18], but we noticed them[6].
New imaging anatomy concepts of the intersphincteric space
In 1979, XXXX suggested that there were three longitudinal muscles and four longitudinal intersphincteric spaces between the internal and external anal sphincters based on microscopic findings. He then defined four longitudinal channels for perianal abscess and classified anal fistula according to these findings [9]. In 2009, our MRI research showed that there were only two intersphincteric spaces, one is the inner space, which was equal to the deep fascia of anal panel; the other is the outer space, which located between the levator ani muscle and the external anal sphincter [6]. Since there are only two longitudinal channels for perianal abscess, the anatomical basis of the classification of anal fistula has completely changed, which needs to be remodified.
As above mentioned, following are the main new signs and concepts discovered by the imaging experts in recent 20 years: the levator ani muscle descends the anus by isometric contraction and its vertical portion becomes chordae tendineae; the puborectalis muscle actually elevate the anus; the external anal sphincter consists of 3 portions and the inferior orifice of anal canal is blocked by the subcutaneous portion blocks; the extended portion of rectum is considered as anal canal; the conjoined longitudinal muscle is not conjoined; the anal area is divisible into 7 layers on MRI and there are only two intersphincteric spaces. The clinical significance of these new signs is unclear and further analysis by relevant experts is required to improve the development of Chinese colorectal surgery and pelvic floor surgery.

联合纵肌的影像解剖新概念
   1852年XXXX认为联合纵肌（the conjoined longitudinal muscle）来自直肠纵肌，并得到提肛肌的加强；1934年XXXX认为它还得到了耻骨直肠肌和外括约肌深部的加强；1976年XXXX认为它是由直肠纵肌、提肛肌垂直部和耻骨直肠肌下延部共同构成的[16]。我们的MRI图像显示直肠纵肌位于肛管部，提肛肌垂直部则位于肛周部，二者之间有明确的解剖分界面，没有任何融合或联合迹象[6]。由此看来，联合纵肌应当是一个多余的解剖概念，因为这个概念不利于分析单一肌肉的功能。
肛区解剖分层的影像解剖新概念
解剖学将肛区分为5层，包括粘膜、粘膜下、肛门内括约肌、联合纵肌和肛门外括约肌层[15]。在过去的20年间，由于在肛区和盆底研究中，研究者生搬硬套5层观念，几乎所有研究者都误读了肛区影像分层。MRI肛区影像分层的误读最令人遗憾，MRI的最大优点是可以将肛区的肌肉系统和含脂的间隙系统同时显示在一张图像上，肌肉（黑色）和间隙（白色）互为参照物，并且彼此彰显，这是其它解剖工具所不能做到的。我们的MRI图像显示肛区可以分为粘膜、粘膜下、肛门平滑肌、内侧括约肌间间隙、提肛肌垂直部、外侧括约肌间间隙和肛门外括约肌层，共7层[6]。按理说，肛区有可能被MRI显示为9层，但因直肠纵肌极薄，且与肛门内括约肌合并为一个肛门平滑肌影像层，故而表现为7层。
格氏解剖第39版，采用了肛内MRI图片展示肛区解剖分层，图中只标记了5层结构，其中只有肛门外括约肌层被正确标记[13]。此等错误源自肛内MRI线圈的发明人XXXX的最初研究，1995年她错误地认为肛区在MRI图像上也应当表现为5层 [17]。这些错误被后人承袭[10, 11, 18]，并被我们所发现[6]。
括约肌间间隙的影像解剖新概念
1979年XXXX依据显微镜下所见，提出在肛门内外括约肌之间有3个纵行肌和4个纵行的括约肌间间隙（the intersphincteric space），为肛周脓肿划定了4个纵行通道，并据此对肛瘘进行了分类[9]。我们2009年的MRI研究显示括约肌间间隙只有内外2个，内侧间隙相当于肛管的“固有筋膜”，外侧间隙位于提肛肌与肛门外括约肌之间[6]。由于肛周脓肿的纵行通道只有2个，所以肛瘘分类的解剖基础发生了根本性变化，这就要求重新制定肛瘘分类办法。
综上所述，20年来，影像学家发现了以下主要新征象和新概念：提肛肌以等长收缩降肛，其垂直部变成了腱索；耻骨直肠肌是真正的提肛者；肛门外括约肌由3部分构成，皮下部外括约肌封堵肛管下口；直肠延续部被看作肛管；联合纵肌并没有发生联合；肛区在MRI图像上为7层；括约肌间间隙只有2个。这些新征象的临床意义尚不十分清楚，需要相关专家一同分析，以期促进我国的肛肠外科和盆底外科的发展。