Smith & Nephew, Inc. and John O. Hay-hurst, M.D. (together “S & N”) appeal the *1306summary judgment of the United States District Court for the District of Oregon, holding that Ethicon, Inc. did not infringe, induce infringement, or contribute to infringement of United States Patent No. 5,601,557 (“the '557 patent”) entitled “Anchoring and Manipulating Tissue.” 1 We vacate the summary judgment and remand for further proceedings.
BACKGROUND
Bones are ’ composed of two kinds of bony tissue — a relatively thin, harder outer shell called “cortical bone” and a softer inner bone material called “cancellous bone.” The '557 patent is directed to a method of affixing a suture within a bone by inserting a suture anchor having resilient legs into a hole drilled into the bone. After the anchor is pushed into place the resilient legs automatically expand into the cancellous bone, anchoring the suture. The specification states that “Whenever tension is applied to the suture, the ends of the legs dig into the bone and resist removal of the anchor member from the hole.” The invention is illustrated in the '557 patent as follows:
[[Image here]]
The '557 specification explains:
Once expelled from the needle into the hole, the resilience of the anchor member urges the outer edges of the legs to bear upon the bone within the hole. With the outer edges 87 of the legs bearing upon the bone, any tension applied to the suture 82 causes the sharp edges 87 to dig into the bone to secure the anchor member within the hole. The barbs 88 also dig into the bone to supplement the anchoring effect of the legs 86.
Preferably, the anchor member is sized so that when it is positioned within the hole, the outer edges of the legs are beneath a relatively dense bone layer that is located at the surface of the bone, and is known as the cortical layer. As a result, tension in the suture (in conjunction with the intrinsic resilient force of the anchor member that forces the leg edges apart) tends to lodge the edges of the anchor member legs beneath the cortical layer, rendering the anchor member substantially irremovable from the hole.
'557 Patent, col. 9, lines 42-60. Claim 1, the broadest claim, follows:
1. A method of anchoring in bone a member and attached suture, comprising the steps of:
forming a hole in the bone;
attaching a suture to a member;
lodging the member within the hole by pressing the member with attached suture into the hole; and
attaching tissue to the suture so that the tissue is secured against the bone.
Surgeons are instructed to tug on the suture after the anchor is pressed into the bone, to assure that it is securely seated. *1307After the anchor is secure, the suture is used to re-attach damaged tissue, such as a ligament or tendon, so that it is correctly placed during healing.
The Ethicon anchor bearing a suture, pictured below, is also pressed through a needle into a hole drilled in the bone:
[[Image here]]
Ethicon states that instead of the resilient spreading of the S & N anchor legs after the anchor enters the bone, the Ethicon anchor has two or four legs or “arcs” that are spread into the bone by manipulation by the surgeon, who “sets” the legs by pulling backward on the suture, causing the legs to spread and to prevent further movement. S & N disputes this mechanism, stating that the Ethicon legs are resilient and recover their shape when expelled from the needle.
Ethicon moved for summary judgment of noninfringement on the ground that the setting action assertedly required for its anchor, whereby the surgeon pulls on the suture to spread the arcs of the anchor and drive them into the bone, avoids infringement of correctly construed claims. Ethicon argued that the clause “lodging the member within the hole by pressing the member with attached suture into the hole” limits the '557 claims to a method wherein the anchor is permanently and fixedly embedded in the bone upon being pressed into the hole, without any further manipulation or movement. Ethicon stated that its anchor moves “significantly” on the surgeon’s pull, and thus is not “lodged” without this manipulation. Ethicon states that the movement of its anchor averages 1.1 to 1.8 millimeters. The amount of movement of the Ethicon anchor was disputed, S & N stating that the Ethicon tests used degraded pig bone and involved the application of excessive force.
The district court referred the claim construction issues to a magistrate judge, who construed the “lodging the member” term as follows:
This phrase requires that the member, once pressed into the hole, may not be removed.* However, this phrase does not require that the member be immovable when pressed into the hole, nor does it preclude movement or manipulation after lodging. Therefore, once the member is pressed into the hole, some manipulation may occur, but manipulation beyond pressing must not be necessary in order to secure the member in the hole. The further movement that is permitted, but not required, may or may not result in the device being even more securely lodged in the bone. For instance, tension on the suture may, in some configurations result in further deformation of the portion of the device in the cancellous layer of bone resulting in an even more secure placement of the device. The key is, however, that this may make the device even harder to remove from the bone, it is not required to keep it within the bone.
Upon Ethicon’s motion for clarification, the magistrate judge added the following footnote to the first sentence, at the asterisk:
* Therefore, “lodged” means that the device is ready to secure tissue to the bone.
*1308The magistrate judge provided no explanation of the footnote, and the parties debate its meaning. Ethicon states that the footnote means that all movement of the anchor is excluded after it is pressed into the hole, since an anchor that moves is not “ready to secure tissue to the bone,” and that the claim is limited accordingly. S & N states that this interpretation contradicts the several statements of the magistrate judge that there is no requirement that the anchor be immovable immediately after it is inserted, stressing the magistrate judge’s explicit statement that the claim “does not preclude movement or manipulation after lodging.”
The district court accepted the magistrate’s claim construction, but also accepted Ethicon’s view that the “lodging the member by pressing” step of claim 1 requires an anchor that need not be pulled by the surgeon before the suture is attached to tissue. The court ruled that if pulling to set the anchor is required, there can not be infringement. The district court concluded that “in the absence of some evidence that the step of applying pressure to the suture anchor is, in fact, a needless step, there is no evidence from which a reasonable juror could conclude that surgeons using the accused suture anchors directly infringe the '577 patent.”
The district court concluded that in the Ethicon method the suture is not ready to secure tissue when the anchor is pressed into the hole, but only after a subsequent pulling/seating step. On this claim construction, the district court granted summary judgment of non-infringement, both literal and under the doctrine of equivalents.
DISCUSSION
I
Claim construction is reviewed as a matter of law, including de novo determination of any fact-based issues underlying the legal conclusion of claim scope, without deference to the rulings of the district court. Cybor Corp. v. FAS Technologies, Inc., 138 F.3d 1448, 1456, 46 USPQ2d 1169, 1174 (Fed.Cir.1998) (en banc). The question of infringement, which was decided on Ethicon’s motion for summary judgment, is reviewed for correctness on the record before the district court, drawing all reasonable inferences in favor of the non-movant S & N.
S & N states that the district court’s claim construction is incorrect, in that it departed from the magistrate judge’s correct construction that the claim “does not preclude movement or manipulation after lodging.” S & N points out that the magistrate judge rejected Ethicon’s requested construction that no further movement or manipulation of the anchor is permitted by the claim. S & N also states that it is irrelevant whether further movement or manipulation of the anchor occurs within the bone, for that limitation is not included in the claim. Thus S & N argues that the district court erred in construing the claim as barring any movement or manipulation of the anchor by the surgeon after the anchor has been inserted in the bone.
We conclude that the magistrate judge correctly ruled that the term in claim 1 of “lodging the member within the hole by pressing the member with attached suture into the hole” does not bar the surgeon’s tug and any ensuing small movement of the anchor after insertion. The claim construction was correct that “lodging” means that the anchor can not be withdrawn after it is pressed into the hole, and that “further movement ... is permitted, but not required.” This construction does not preclude the action whereby the surgeon tugs *1309on the suture after insertion. Both sides agreed that any prudent surgeon would assure that an anchor is seated within the bone structure. In the S & N method the pull on the anchor tests whether it is lodged, because the legs automatically spread into the cancellous layer of bone; the pull confirms that the legs have spread. There was evidence before the district court that both sides issue instructions that surgeons should perform such a pulling step.
The Ethicon anchor, according to Ethi-con, is not lodged until after the surgeon pulls on the suture, Ethicon stating that it is the ensuing movement through the bone that spreads the legs, and that without this manipulation the Ethicon anchor is not “lodg[ed] within the hole” as set forth in claim 1. S & N disputes this view of the Ethicon method, and also argues that even if a pulling step were necessary to lodge the Ethicon anchor, the claims are infringed because all of the claim steps are performed, whether or not this additional pulling step is also performed.
Ethicon also argues that S & N is estopped from a claim construction covering a method that requires or employs manipulation or movement of the anchor after it is pressed into the hole, pointing to arguments made during prosecution of the '557 patent to overcome a reference to Freedland. Ethicon is partly correct, for estoppel arises, but not to the extent proposed by Ethicon. The Freedland reference shows a fastener inserted into a hole drilled into a broken bone at the fracture; the fastener serves not as a suture anchor, but as a device for holding the fracture together during healing. The Freedland fastener has arms that project into the cancellous bone tissue in the area of the fracture, but the arms of the fastener must be extended by manipulation. In the Freedland method the surgeon uses forceps to pull on a cord that extends through a tube, opening the Freedland fastener like an umbrella, thereby pinning itself inside the hole in the bone and stabilizing the bone against movement while the fracture heals.
During prosecution of the '557 patent the inventor, Dr. Hayhurst, argued that a difference between his invention and that of Freedland is that the '557 anchor is securely embedded simply upon pressing it into the hole where it resiles into the can-cellous bone, and that no further manipulation of the anchor is necessary. At his deposition Dr. Hayhurst repeated this description of how his invention works. Ethicon states that the inventor thus disclaimed coverage of its anchor that requires manipulation to move it into place and open its legs. S & N responds that the Ethicon anchor does not require such manipulation, and that the surgeon’s “pull” on both the '557 anchor and the Ethicon anchor after they are pushed into the hole is not a Freedland-type manipulation to open a closed device in a fracture, but a routine cautionary safeguard to assure that the anchor is in place in the bone. S & N states that the '557 inventor did not disclaim his own procedure which included this safeguard of pulling on the anchor, and that Dr. Hayhurst merely distinguished Freedland’s elaborate manipulation that opens an umbrella-like fastener inside a fractured bone.
We agree with the magistrate judge that the '557 claims, correctly construed, neither exclude nor require a surgeon’s tug to assure that the anchor is set within the cancellous bone, and do not preclude a small movement as a result of that tug. A claim interpretation that *1310would exclude the reasonable practice of the method taught in the patent “is rarely the correct interpretation; such an interpretation requires highly persuasive evi-dentiary support.” Modine Mfg. Co. v. United States Int'l Trade Comm’n, 75 F.3d 1545, 1550, 37 USPQ2d 1609, 1612 (Fed.Cir.1996).
Ethicon is incorrect in its assertion that since a pulling step (sometimes called a “tensioning” step) is “necessary” to set the Ethicon anchor, the step of pulling on the anchor must be included in the '557 claims in order for the claims to be infringed. Our colleague in dissent has adopted Ethieon’s argument, and would hold that since the step of tugging on the anchor is not stated in the '557 claims, the claims can not be infringed. However, claims are infringed when all of the steps thereof are performed, unless the invention itself is the elimination of additional steps. The '557 patent, its specification, prosecution history, and testimony in the summary judgment proceeding, make clear that the instruction to surgeons to tug on the anchor before using it is a matter of prudent surgical practice, not a limitation of claim 1.
Both parties point to claim 6 of the '557 patent, which includes the limitation that the anchor is not “manipulated”:
6. A method of anchoring in bone a member and attached suture, comprising the steps of:
providing a deformable member having a width dimension “D”;
attaching a suture to the member;
forming a hole in a bone in a manner such that the hole has a diameter that is not greater than the width dimension “D”; and inserting the member into the hole with the member oriented such that the member lodges within the hole in the absence of any manipulation of the member other than inserting the member into the hole.
Ethicon states that claim 6 shows that S & N’s invention excludes manipulation of any sort after insertion of the anchor, while S & N states that this claim describes a preferred embodiment, in accordance with the doctrine of claim differentiation. This doctrine reflects the presumption that separate claims are of different scope. See Irving Kayton, 1 Patent Practice (6th ed.) 3.1, 3.3 (1995):
[Pjatent practitioners typically draft a series of claims approximating a spectrum of patent protection.... The first way in which a claim may be made narrower is by adding a limitation to it in the form of an additional element.
We discern no basis for reading claim 6 as countermanding the claim construction of the magistrate judge that claim 1 “[does not] preclude movement or manipulation after lodging.”
We confirm the district court’s claim construction, with the modification or clarification that claim 1 neither excludes nor requires the step of pulling on the suture after it is inserted. However, we also confirm the claim construction that the '557 claims do not cover a device whereby the suture must necessarily be manipulated and moved through the bone in order to open the anchor and spread the legs, for S & N is estopped by the prosecution history from covering a device that must be manipulated and opened in order to fix it in the bone. However, a device that resiles after insertion into the bone may be manipulated provided that “manipulation beyond pressing [is not] necessary in order to secure the member in the hole.”
*1311II
The signal that additional steps may be performed in carrying out a claimed method is the word “comprising.” See Vivid Technologies, Inc. v. American Science & Engineering, Inc., 200 F.3d 795, 811, 53 USPQ2d 1289, 1301 (Fed.Cir.1999) (the signal “comprising” is “generally understood to signify that the claims do not exclude the presence in the accused apparatus or method of factors in addition to those explicitly recited”); Moleculon Research Corp. v. CBS, Inc., 793 F.2d 1261, 1271, 229 USPQ 805, 812 (Fed.Cir.1986) (stating “the general proposition that an accused method does not avoid literally infringing a method claim having the transitional phrase ‘which comprises’ (or ‘comprising’) simply because it employs additional steps”). Our esteemed colleague in dissent criticizes use of the signal “comprising,” calling it a “weasel word” to somehow avoid precision in claiming. This signal nonetheless appears in the vast majority of patent claims, for it implements the principle that claims are intended to provide a concise statement of the claimed invention as distinguished from what has gone before; a claim is not a handbook for practice of the invention. Dr. Hayhurst’s testimony and instructions that any prudent surgeon would test the anchor to assure that it is lodged beneath the cortical layer was not included in any claim. Nor was it required to be claimed. It is neither a “shortcoming” nor a “weaseling” to use “comprising” to recognize that inventions may be practiced with steps in addition to those listed in the claims.
A claim is not defective when it states fewer than all of the steps that may be performed in practice of an invention. See, e.g., Moleculon Research Corp., 793 F.2d at 1271, 229 USPQ at 812 (method as practiced can include steps in addition to those stated in the claim). Infringement arises when all of the steps of a claimed method are performed, whether or not the infringer also performs additional steps. See, e.g., Vivid Technologies, 200 F.3d at 811, 53 USPQ2d at 1301 (inclusion of steps in addition to those stated in the claim does not avoid infringement); Stiftung v. Renishaw PLC, 945 F.2d 1173, 1178, 20 USPQ2d 1094, 1098 (Fed.Cir.1991) (claim using “comprising” reads on devices which add additional elements). This court did not hold in Maxwell v. J. Baker, Inc., 86 F.3d 1098, 39 USPQ2d 1001 (Fed.Cir.1996) that unless all disclosed procedures are included in the claim, the patentee has “dedicated to the public” not only the unclaimed procedures but the entire claimed process. Our colleague in dissent, finding the disputed fact that Ethicon’s method “indisputably requires the ‘tensioning1 step,” proposes that Maxwell requires that by not including the tensioning step in the claims, S & N dedicated to the public any procedure in which that step is included. However, Maxwell does not treat such a situation. The issue in Maxwell was whether equivalent subject matter that was disclosed in the specification but not claimed can be reached under the doctrine of equivalents.2 There was no issue in Maxwell of literal infringement when every step of the claimed method is in fact practiced by the accused infringer. In Maxwell the accused equivalent subject matter was disclosed in the patent but was *1312plainly outside the scope of the claims. Such a situation is not here presented.
Ill
The '557 invention is directed to an anchor with resilient legs that open and re-sile after the anchor is pushed into the cancellous bone, lodging the anchor against the cortical layer. Although Ethi-con states that the legs of its anchor do not work in this way, there was evidence before the district court supporting S & N’s statement that the Ethicon legs “are made of a highly resilient alloy called ‘Nitinol,’ which will spring back to its relaxed position automatically after being compressed,” and that “No action of the surgeon is required to open the arcs.” Brief at 32-33.
S & N points out that Ethicon’s own product literature and patent do not require a “tensioning” step to move the anchor after insertion, instead stating that the anchor has no tendency to move when inserted. Indeed, the district court agreed with S & N, and ruled: “In sum, it is undisputed that the ares of the accused suture anchors substantially return to their- relaxed state upon being pressed into cancellous bone and can not be removed.” We can not reconcile this ruling with the district court’s apparent conclusion that the surgeon’s pull on the Ethicon anchor is necessary to spread the legs and lodge the anchor. Ethicon’s position that its anchor is not “lodged” until after it has been moved and manipulated within the bone was disputed, and was material to the summary judgment. On the record before the district court this disputed material fact could not be resolved adversely to S & N.
In addition to the issues of literal infringement, S & N argues that even on the district court’s unmodified claim construction infringement would he under the doctrine of equivalents. The questions of infringement are factual issues, and require determination in view of our modified claim construction.
The summary judgment is vacated, and the case is remanded for further proceedings in accordance with the claim construction as modified herein.
VACATED AND REMANDED
. Smith & Nephew, Inc. v. Ethicon, Inc., No. 98-76-MA, 54 U.S.P.Q.2d 1888, 1999 WL 1938857(D.Ore. Dec. 17, 1999).
. This question is being considered by this court en banc, in Johnson & Johnston Associates Inc. v. R.E. Service Co. See 238 F.3d 1347 (Fed.Cir.2001) (Order directing supplemental briefing on questions of "Whether and under what circumstances a patentee can rely upon the doctrine of equivalents with respect to unclaimed subject matter disclosed in the specification.'')