Background
The gold standard for surgical treatment of diaphyseal lower limb long bone fractures is the use of intramedullary nails [
1]. These nails are locked into place with proximal and distal screws to prevent rotation and shortening of the fractured limb. For proximal interlocking specialized jigs offer easy placement of the screws. However, due to deflection of the nail during insertion into the intramedullary canal no such automated process exists for the distal screws. The common practice for insertion of distal locking screws is a freehand technique under fluoroscopic control. The C-arm is tilted until alignment of interlocking hole axis and image intensifier beam is achieved indicated by a round appearance of the hole projected on the image. The drill is then manually oriented under repeated fluoroscopic control. The process is technically demanding and requires experience of the operator [
2]. Depending on the skill level of the surgeon, procedure time and, even more important, radiation exposure to patient and surgical personnel can increase markedly [
3]. In addition, a high rate of screw misplacements is reported [
4].
A variety of approaches have been suggested to solve the distal locking problem. Recent advancements range from various mechanical aiming devices [
5,
6] continued to the idea of drilling from the inside to the outside (Dgimed Ortho. Inc., Minnetonka, US) over the use of surgical navigation systems [
7‐
9] over laser illumination of the hole from the inside of the nail [
4] to ultrasound [
10] or electromagnetic hole locators [
11,
12]. Solutions appear to be technically irreproducible, are restricted to a specific nail/procedure or face the problem of extensive equipment requirements, which seems unjustified in the face of the comparatively simple problem of inserting a screw into a drill hole. Requirements for additional staff and long training periods with gradual learning curves lead many of these new ideas out of favor [
13].
A new fluoroscopy based technique (guided freehand) was developed for simplified distal interlocking, aiming at reducing radiation exposure and operation time. The technique is generally based on the freehand standard and additionally guides the surgeon by means of visible landmarks projected on the C-arm image. A computer program plans the exact drilling trajectory by 2D-3D conversion of the locking hole projections from a single fluoroscopy shot in an arbitrary orientation and provides mentioned guiding landmarks in real-time on the familiar fluoroscopy screen. Interlocking holes can be drilled by visually aligning the drill to the planned trajectory. No additional tracking or navigation equipment is required.
Object tracking by utilizing X-ray projections of cylindrical holes carries potential for a variety of applications within trauma and orthopedics for positioning implants and instruments, such as screw insertions, guide-wire placements, positioning of plates, nails or prostheses was well as anatomical fracture reduction. The hereby introduced guided freehand distal locking system is considered as pilot application of the overall concept.
The aim of the study was to investigate this technique on human cadaveric specimens in an operating room (OR) environment, with realistic OR settings. The guided freehand procedure was compared with the conventional freehand technique in terms of radiation exposure and operational time.
Discussion
Distal locking of intramedullary nails is a frequently employed, but technically demanding procedure in trauma surgery. In this paper a newly developed technique for implant independent distal interlocking (guided freehand) is introduced. The method was compared to the common freehand technique in an experimental cadaveric setting exemplified on conventional tibia nails. The guided freehand technique has proven to reduce radiation exposure by more than 50% and operation time by 20% when compared to freehand.
When a nail is inserted into a long bone, it is likely to bend according to the curvature of the intramedullary canal [
14]. Exact orientation of the distal interlocking holes is, hence, difficult to predict. Usually surgeons use repeated fluoroscopy to insert the screws in a freehand manner. Radiation is a growing problem amongst orthopedic surgeons, associated with a relative risk for cancer of 5.37 with respect to the general population [
15]. Malignancies of exposed personnel range from cancers of solid organs (i.e. thyroid and pancreas), to skin and hematopoietic cancers [
16]. In female orthopedic surgeons the standardized prevalence ratio for all cancers is 1.9 and 2.88 specifically for breast cancer when compared to the general population [
17]. In the literature radiation expenditure for freehand locking differs widely. E.g. Gugala et al. reported a fluoroscopy time of 36 s for placement of two screws in the tibia [
3], whereas Suhm et al. stated intense use of fluoroscopy during freehand locking of 108 s per screw [
9]. Factors such as experience level of the operator or experimental/clinical setup might contribute to this scattering. In this study all screws were placed by a single 4th year orthopedic resident to test the system on a young experienced surgeon. Recorded average radiation time during conventional freehand locking was 35 s for two locking screws, which is in the lower range of the reported values. In the investigation of Kirousis et al. [
2] a complete tibia nailing procedure required 72 s of radiation and resulted in an effective dose of 0.04 mSv for the operating surgeon and 0.11 mSv for the C-arm technician. In view of the annual dose constraint of 10 mSv (International Commission on Radiological Protection [
18]) efficient and responsible use of radiation is of utmost importance.
Opposed to the mentioned disadvantages of freehand locking such as radiation exposure and handling complexity, marginal requirements for recourses and equipment make the technique practicable all over the globe. Guided freehand does not aim to replace but to enhance the freehand gold standard. Tilting of the C-arm to achieve round hole-projections implies 3D vision of the surgeon and often generates the major portion of time and radiation. Some operators rotate the patient rather than the C-arm, which could end in loss of fracture reduction and anatomical mal-rotation. With guided freehand the C-arm can be maintained at a convenient position. The demanding step of aligning the drill to the image intensifier projection axis is replaced by intuitive matching of targeting structures, easy to follow by the inexperienced surgeon.
A variety of options for improved distal interlocking has been proposed in the past decades. None has yet found its way into accepted clinical practice. With fluoroscopy based navigation for distal interlocking an impressive reduction in radiation could be reported (2 s radiation per screw). However, the authors state that fluoroscopy based surgical navigation markedly increases the need for resources [
19]. Up to an additional 40 min were required prior to skin incision and after skin closure as set up and take down time for the navigation system. Moreover, an especially trained technician was needed [
20]. Ultrasound based techniques using differences in resonance between nail and bone [
10] lack accuracy. Electromagnetic solutions offer acceptable accuracy without direct involvement of radiation. However, radiographic imaging is still required to confirm proper screw position at the end. Electromagnetic tracking implies the insertion of a pre-calibrated single-use probe into the cannulation of the nail with a cable connection to the outside [
12,
21]. Besides additional bridging between sterile and non-sterile fields, the blocked cannulation restricts the surgeon to a specific sequence of screw insertion progressing from distal to proximal, which is contradictory to clinical practice. Moreover, all magnetic metals such as stainless steel implants, screwdrivers or drill-bits cannot be used. It is a point of discussion whether it is worth interrupting the clinical workflow and setting up of an additional support system with considerable restrictions for the sole task of distal interlocking.
The hereby proposed concept is consequently geared to existing surgical workflows. Besides a simple targeting jig and software, no additional equipment is required. A conventional C-arm, as available in the bulk of operation theaters worldwide, is essential for nailing. It appears, therefore, reasonable to further involve the C-arm for distal interlocking when it is already employed for nail insertion, fracture reduction and implant positioning beforehand. The authors believe that the potential of conventional radiographic devices is only barely taped since the intended use is confined to plain visualization. Elevating the role of the conventional C-arm from pure imaging could significantly contribute to an improved surgical outcome. Hidden information within 2D projections could be more efficiently extracted by innovative algorithms to aid the surgeon in the daily routine. The idea of reconstructing 3D information from 2D projections is certainly not a novelty. Cone beam algorithms, for example, build the basis for 3D reconstruction by computed tomography [
22]. Some work has already been done on utilizing the image intensifier projections of distal locking holes to identify their orientation [
23,
24]. However, solutions appear clinically infeasible. At least two projections are required from different angles, computational time is long and no simple procedure has been suggested to finally position a tool for drilling.
The hereby proposed method assists in planning (identification of the hole orientation) and navigation of the drill by means of a 2D C-arm. The 3D transformation problem is solved from a single projection of two interlocking holes. No tilting of the C-arm is required to obtain a second view angle. Computational time for two holes is kept below 1 s and robustness is increased by extracting only significant landmarks from the hole contour rather than processing the entire projection shape. The technique is independent from a specific implant type or brand and could be universally applied to tibia, femur, humerus or to other relevant regions.
However, some issues need to be addressed in the future. As the algorithm processes two interlocking holes simultaneously, recalculation due to movement of the patient might be difficult after the first screw is already inserted. Currently, the prototype system simulates the second hole on basis of the previous calculation. Furthermore, patient movements relative to the C-arm can have adverse effects on the procedure. With the next X-ray the algorithm identifies and corrects these motion artifacts by real-time monitoring of shifts between images and performs automated recalculations. Nonetheless, the nature of the method remains static. Events occurring between distinct snapshots cannot be detected. The advantages of a freehand procedure are essentially to be seen in maximized freedom and usability for all kinds of applications independent from specific devices and implant families. However, from a handling perspective a freehand procedure remains demanding. In the present case, the skill level of the operator determines the extent of radiation needed for iterative control. For example a handheld drill-bit happens to slip on the cortex requiring restart of targeting. Slippage and soft tissues wrapping around the drill-bit could be improved by the proposed targeting jig. Still, the screw is freely inserted into the drill-hole. Particularly in reduced bone quality, mal-placement of the screw might occur even if the drill channel was correct.
Abstracting the underlying idea as a future outlook, an intramedullary nail with interlocking holes can be regarded as a radio-opaque object with two cylindrical holes, which can be spatially tracked from 2D projections. Holes could therefore be regarded as tracking markers. Even though, all kinds of geometries are thinkable to serve as markers, cylindrical holes appear favorable because they are easy to produce (or already existent) and their projection is explicit to track. It is therefore credible to establish a support system for surgical routine interventions, e.g. for controlling the entire nailing procedure from implant positioning to distal interlocking.
Competing interests
The authors are not compensated and there are no other institutional subsidies, corporate affiliations, or funding sources supporting this work unless clearly documented and disclosed.
Authors' contributions
MW developed the concept, programmed the software algorithms and drafted the manuscript. JS performed the surgeries and drafted the manuscript. LF planned and supervised the experiments, evaluated the data and helped in writing the manuscript. BD designed and produced the prototype targeting jig. ML contributed to study planning and supported the paper draft. GR supported study planning and the concept development process and revised the manuscript. All authors read and approved the final manuscript.