By learning a multilevel statistical model of the surgical process, the proposed methods use both the knowledge of the temporal process of the surgery and the knowledge of the relationships between the surgical steps and phases. The system provides the most probable sequences of surgical steps and phases, given present and past information only, by assigning a label for steps and phases to each image of the video.
Building an annotated medical dataset is a challenging task, first because of the sensitivity of the data.
Our system needs to deal with this limitation of data, especially to train a model of the surgical process. Our system also needs to deal with the specifity of the medical data recorded from video monitored surgery. In the case of the cataract surgery, the system has to deal with the motion of the eye during the surgery, and some zoom or level variations.
The remainder of this paper is organized as follows. Related work is pre- sented in 2. The proposed methods are described in 3.
In 4, we discuss our experimental setup and results. Finally, 5 presents some concluding remarks. Every surgical process analysis method describes the surgery at a given level of abstraction. A high level of description, into surgical phases [11] or tasks [19], provides a global description of the surgery with a simple sequencing.
Indeed, surgeries generally have the same phase sequencing [15, 10]. Automatic recognition at a low level of description, on the other hand, is a challenging task because of the large number of possible temporal sequences, sometimes marginally represented in the dataset.
So, a low level of description, into surgical gestures [29], activities [10] or steps, allows a more precise analysis of the surgery, but implies a more complex surgical process. Some authors propose to work with several levels of description.
For in- stance, Padoy et al. This action detection supports the recog- nition of high-level phases. Forestier et al. Those two methods allow an on-line recognition of high- level phases, but low level information presence of surgical tools or activities is manually provided by surgeons.
Lalys et al. In terms of methodology, several methods reuse data recorded during a video-monitored surgery for the automated analysis of surgical processes [10, 29]. In particular, some of these methods rely on Content-based video retrieval CBVR , whose goal is to find similar videos or sub-videos inside a dataset [5, 18, 20, 13]. But those methods do not model the temporal sequencing of the surgical process. Different kinds of models were used to model this process, like Dynamic Time Warping DTW averaging [15, 10], which builds an av- erage surgery.
But this method does not allow on-line computations because it requires the entire video to be known past, present, but also future infor- mation.
CRFs seems to provide better results than the HMMs in the context of automatic surgical video analysis [19, 25]. A majority of methods for the analysis of video data are developed without the real-time constraint: they are applied to automatic documentation and report generation [24, 10], fast search of similar cases in a database [1] or educative video construction [2].
A few of them allow on-line analysis [15, 7], but at a high level of description into surgical phases and do not allow accurate analysis of the surgery. The method presented in this paper extends a previous solution from our group, presented at a conference [4]. That system performs an on-line anal- ysis of a cataract surgery video at two different levels of description. It uses high-level phase recognition to help low-level step recognition, but it also uses information from step recognition to refine the recognition of phases.
Those models aim to model the tem- poral process at each level of description as well as the relationships between steps and phases. The ad- vantage of HHMM is that is can jointly model the temporal process at multiple description levels, using a simple relationship model between levels. We then evaluate two novel models which handle separately the relation relationships between steps and phases and the temporal process.
These models can take as inputs different kinds of observations, computed throughout the surgery. These observations are presented in the following section. Next, the models are presented. Each sub-sequence contains the same number of frames, as presented in Fig. For each sub-sequence, observations are computed. In this paper, observations derive either from the presence of surgical tools or from the analysis of motion in videos through the content-based video retrieval CBVR paradigm.
As motion extraction can be disturbed by scale variation or motion of the eye, we also evaluate the system after a spatial normalization of images from the video stream.
In a scenario where the system is used jointly with a surgery simulator, we can easily assume that this information will be provided by the simulator. In a scenario where the system is used during a real surgery, this information may also be obtained using barcodes or RFID chips [21, 28].
Another solution consists in automatically tracking surgical tools in the surgical scene by computer vision methods. In this paper, this information is manually provided by surgeons, which allows us to validate the models. Disruptive motion could appear, induced by camera or eye motion dur- ing cataract surgeries for exemple. The influence of a spatial normalization of images that composed the video on the performance of the system is evalu- ated. This normalization step aims to refine feature extraction based on pupil center and scale tracking [3].
Then, the zoom level is estimated from the illumination pattern reflected on the cornea. Knowing this information for each frame of the video, we can pre- processe them to balance eye motion, zoom or level variation before the feature extraction step. First, eye motions are balanced by registering all frames on the same iris center. A simple coordinate system change is ap- plied, which places the iris center at the image center. This should allow eliminate motion induced by eye or camera motion and make tool motion more relevant.
Then, all frames are scaled on a same scale level to balance zoom or level variations. After this last preprocessing step, all irises should have the same radius. Finally, a circular mask centered on the iris center is applied to select a region of interest, because all relevant actions should appear in a region closed to the iris location. The OpenCV 2 library1 is used to select strong corners and the optical flow between each pair of consecutive frames is computed at each strong corner by the Lucas-Kanade iterative method [14].
Finally, the motion contained in the sub-sequence as a whole is char- acterized by one 8-bin amplitude histogram, two 8-bin amplitude-weighted spatial histograms one for the x-coordinates and one for the y-coordinates and one 8-bin amplitude-weighted directional histogram.
STIP points are first detected locally within each sub-sequence. During a learning step, those local feature vectors are used to build a dic- tionary of visual words. Cataract Surgery Advancements. Cataract Surgery. View Live Surgery Videos. Upper and Lower Eyelid Surgery Blepharoplasty. Watch a Lower Eyelid Surgery. Muellerectomy Eyelid Tightening Surgery.
Levator Repair Surgery. Watch an Endoscopic Brow Lift Surgery. Direct Brow Lift Surgery. Watch an Ectropion Surgery. Watch an Entropion Surgery. Upper Eyelid Surgery. In the News. Laser Cataract Surgery on News 3. We Made the News! Satisfied Patients. Cataract Surgery Testimonial. Frances after Cataract Surgery.
Georgia after Cataract Surgery. John after Cataract Surgery. Bunny after Eyelid Surgery. David after Laser Eye Surgery. Normand after Laser Eye Surgery. Also, you may need cataract surgery if you start to see starbursts around lights at night or hazy images during the day. Once the decision is made, cataract surgery typically is done in a clinic and does not require an overnight hospital stay.
In fact, an uncomplicated cataract procedure usually takes less than 10 minutes. Prior to surgery, your eye will be dilated with eye drops. Next, your surgeon will numb your eye with drops or an injection of anesthesia. A special surgical tool then will be used to make at least one small incision into the eye.
A fluid substance called viscoelastic is injected into the eye to help stabilize the interior and maintain eye pressure. A hollow tip then is inserted into the eye in a common procedure known as phacoemulsification.
No, each video within the collection is its own file. What is the length of each video? The average length of each video is minutes. This collection includes 13 videos in English and 13 in Spanish. Still have questions? Normal business hours of the Academy's Service Center are a.
Pacific Time, Monday through Friday. Advanced Search. Product Number V. Be the first to review this product. Please log in to receive member pricing. Add to Cart. Print this page. View Samples.
0コメント