Accuracy of a magnetic resonance imaging‐based 3D printed stereotactic brain biopsy device in dogs

Abstract Background Brain biopsy of intracranial lesions is often necessary to determine specific therapy. The cost of the currently used stereotactic rigid frame and optical tracking systems for brain biopsy in dogs is often prohibitive or accuracy is not sufficient for all types of lesion. Objectives To evaluate the application accuracy of an inexpensive magnetic resonance imaging‐based personalized, 3D printed brain biopsy device. Animals Twenty‐two dog heads from cadavers were separated into 2 groups according to body weight (<15 kg, >20 kg). Methods Experimental study. Two target points in each cadaver head were used (target point 1: caudate nucleus, target point 2: piriform lobe). Comparison between groups was performed using the independent Student's t test or the nonparametric Mann‐Whitney U Test. Results The total median target point deviation was 0.83 mm (range 0.09‐2.76 mm). The separate median target point deviations for target points 1 and 2 in all dogs were 0.57 mm (range: 0.09‐1.25 mm) and 0.85 mm (range: 0.14‐2.76 mm), respectively. Conclusion and Clinical Importance This magnetic resonance imaging‐based 3D printed stereotactic brain biopsy device achieved an application accuracy that was better than the accuracy of most brain biopsy systems that are currently used in veterinary medicine. The device can be applied to every size and shape of skull and allows precise positioning of brain biopsy needles in dogs.

imaging can be challenging. [1][2][3][4][5] In a study of brain lesions that were visible on MRI in dogs where the investigators only had to differentiate between gliomas and a presumptive cerebrovascular accident, there was a significant error in identifying the correct disease with 10% to 47% of cerebrovascular accidents are misinterpreted as gliomas. 6 Conversely, up to 12% of gliomas are diagnosed as infarctions. 6 Therefore, clinicians often are expected to make treatment recommendations based on presumptive, potentially erroneous diagnoses.
Histopathological examination of representative brain tissue specimens can increase diagnostic accuracy in those cases. 5,[7][8][9][10] However, the high complexity and vulnerability of the brain requires minimally invasive and highly precise techniques for obtaining brain biopsies. Stereotactic brain biopsy devices are often considered superior to an open surgical approach as they minimize injuries to the brain and surrounding tissues during the biopsy procedure. 11 In the last 2 decades, several CT-and MRI-guided stereotactic brain biopsy devices have been developed for veterinary use. 5,[7][8][9][11][12][13][14][15][16][17][18][19][20][21] The CT-guided modified Pelorus Mark III stereotactic system was 1 of the first biopsy devices in veterinary medicine. The mean needle placement error is determined to be 3.5 mm (SD: 1.6 mm). 12 In another study, a modified Laitinen frame system for CT-guided stereotactic brain biopsy is used. 13 The established needle placement error is 2.9 mm (SD: 1.08 mm). The mean needle placement errors of the Model 1430 MRI KOPF stereotactic system (David Kopf Instruments, Tojunga, CA) are reported with 0.9 mm (SD 0.9 mm) and 1.7 mm (SD 1.6 mm). 15 The first MRI-guided stereotactic neuronavigation system in veterinary use achieves a needle placement error of 1.79 mm (SD: 0.87 mm). 11 Furthermore, a CT-guided frameless stereotactic brain biopsy system shows a mean application accuracy of 2.9 and 4.3 mm. 16 More recently, a targeting error of <3 mm is measured using another stereotactic MRI-guided biopsy device 17 and 1 study determines median needle placement errors of 1.55 mm (range: 1.1-3.4 mm) and 1.5 mm (range: 0.9-2.0 mm) using stereotactic headframes for brain biopsy. 19 All of the aforementioned systems have 1 or more disadvantages: (1) Biopsy accuracy is not sufficient for all lesion types; (2) The system is so expensive that it might be cost prohibitive for routine veterinary use.
The aim of the study was to develop a new MRI-based personalized, 3D printed stereotactic brain biopsy device for dogs that does not require expensive biopsy equipment and that has at least the same accuracy as currently available biopsy systems. This study presents the application accuracy of this new device ( Figure 1) in targeting predefined intracranial points using a biopsy needle. Furthermore, the dog body weight and the depth of the target point were evaluated for any influence on procedural accuracy.

| MATERIAL AND METHODS
The cadaver preparation, the computer-aided manufacturing of the brain biopsy device, the brain biopsy needle placement, and the determination of the needle placement error were performed as described. 22

| CADAVER PREPARATION
Biopsy needle placement accuracy was tested in 22 canine cadavers.
The cadavers were divided into small breed dogs (group 1) with a body weight of up to 15 kg and large breed dogs with a body weight of more than 20 kg (group 2). All dogs were euthanized for reasons unrelated to the study.
Each cadaver head was prepared as follows: Three specifically designed titanium bone anchors were screwed to predetermined bony points. These suitable points were located at the left and the right zygomatic arches and at the occipital protuberance. In very large dogs, the frontal bone overlying the frontal sinus was used instead of the occipital protuberance, as the latter was too solid for the attachment of the bone anchors without the need for predrilling. These bony points were determined because of the following conditions and tested in a pilot test on a cadaver dog: (1) The bone anchors should be grouped around the brain, so that the markers in the diagnostic imaging are nearly homogenously allocated; (2) The grouping around the brain was needed for a good and stable stand of the biopsy frame after attaching the frame to the bone anchors; and (3) The bony point should be superficially located, covered just by the skin, and they should be easy to palpate. The titanium bone anchors were small selfcutting screws. The self-cutting thread had a diameter of 2 mm and a length of 4 mm. The head of the bone anchors without markers extended 4.5 mm over the bony surface, allowing for skin closure over

| BIOPSY NEEDLE PLACEMENT
The dog head was placed in sternal recumbency, and the individual frame was secured with its 3 legs to the 3 bone anchors using specifically designed screws ( Figure 4A). Subsequently, minimally invasive access to the brain was obtained. A 1-cm incision was made into the skin and the temporal fascia, and the temporalis muscle was bluntly separated until the skull surface was exposed. A round craniotomy opening of 3 mm in diameter was drilled (Electric Pen Drive, DePuy Synthes, West Chester, Pennsylvania) after placing a drill sleeve on the tool guide for the planned trajectory. The dura mater was perforated using a 0.3-mm hypodermic needle. The drill sleeve was replaced by the biopsy needle sleeve, which was placed in the tool guide.
A spacer was secured to the Sedan side-cutting biopsy needle (Sedan side cutting needle, ELEKTA, Stockholm, Sweden; outer diameter of 2.5 mm) in order to restrict the needle advancement to the desired depth ( Figure 4B). The biopsy needle was advanced through the tool guide with the needle sleeve into the brain parenchyma up to the maximum depth allowed by the spacer. A second CT scan of the skull with the biopsy needle in place was performed using the same scan parameters as in the first scan ( Figure 4C). The entire procedure was repeated for the second target point. The range was added in normally distributed data where it appeared to be appropriate. The comparison between the groups was performed using the independent Student's t test or the nonparametric Mann-Whitney U Test. P-values of .05 or less were considered statistically significant.

| RESULTS
Twenty-two canine cadaver heads were sampled. Eleven dogs were classified as small breed dogs with a body weight of less than 15 kg There was no linear correlation between the target point deviation and the free biopsy needle length (R 2 = 0.2067, Figure 6).

| DISCUSSION
The MRI-based personalized, 3D printed stereotactic brain biopsy device achieved excellent precision in targeting predefined intracranial points in the caudate nucleus and the piriform lobe. The overall median target point deviation was 0.83 mm (range: 0.09-2.76 mm).
The results are better than most of the previous studies of other stereotactic brain biopsy devices that are currently used in veterinary medicine, which have mean needle placement errors ranging from 0.9 mm to 4.3 mm 11-13,15-17 and median needle placement errors of 1.5 mm and 1.55 mm. 19 The error of system tested here is similar to the accuracies of stereotactic devices for brain surgery in human medicine. [23][24][25][26] The high application accuracy of the device presented here was achieved by rigid fixation of the biopsy device to the skull using bone anchors. Therefore, it does not allow any movement between the device and skull that can potentially occur when using a bite plate 11,[13][14][15][16][17]19,21 or a face mask. 18 In accordance with the results, the MRI-based personalized, 3D printed stereotactic brain biopsy frame can be used to perform brain biopsies in dogs of varying weight and skull conformation. The smallest dog in this study was a Chihuahua with a dome-shaped skull and a body weight of 1.8 kg. The largest dog was a dolichocephalic mixed breed dog with a weight of 38 kg. In contrast, some stereotactic brain biopsy devices have limitations with regard to the size of skulls they could be applied to, 15 or the devices only were tested in a homogenous cohort of canine skulls. 11,13,14,16,20 There are difficulties in applying other systems to small and dome-shaped skulls. 12 A bite plate system had problems in brachycephalic dogs caused by the specific round shape of their skulls and the malformations of the maxilla. 11 A 3D printed patient-specific facemask with an attached biopsy port can be used for brain biopsy, 18 but might be problematic in brachycephalic dog breeds because of the use of the bridge of the nose and the nasal planum for the facemask. In the study presented here, 36% of dogs were brachycephalic breeds, such as French Bulldog, Pug, English Bulldog, or Shih Tzu breeds. In addition, some very small, brachycephalic dogs with dome-shaped heads, such as Chihuahua or Yorkshire Terrier breeds, were included. It was easier to place the bone anchors and gain minimally invasive access to the brain in these dogs than in large normocephalic dogs as the thick masseter muscles in large breed dogs complicated the approach to the skull. Figure 6 illustrates the relationship between free biopsy needle length (distance between the lower end of the biopsy port and the biopsy needle tip) and needle placement error. A previous study shows a negative correlation between body weight and needle placement error, 12 but there is no significant relationship between the needle placement error and the target depth in another study. 11 Two other separate studies interestingly conclude that more superficial F I G U R E 5 Box and whisker plots displaying the target point deviation of the 43 predetermined target points in mm using the MRIbased personalized, 3D printed stereotactic brain biopsy device. The results are displayed for both target points separately (target point 1: caudate nucleus; target point 2: piriform lobe; whiskers represent minimum and maximum values) F I G U R E 6 Comparison between the target point deviation and the free brain biopsy needle length for all 43 target points. Black line-best fit curve, broken lines-95% confidence interval lesions are biopsied less accurately than deeper lesions. 15,16 In the study presented here, there was no correlation between the free biopsy needle length and the target point deviation.
Our device can be used on the basis of MRI imaging, which is considered to be the gold standard for diagnosing intracranial lesions.
Titanium bone anchors and markers filled with diluted gadolinium, which are visible in T1-and T2-weighted MR images, facilitate MRI compatibility. Nevertheless, the markers are also visible on CT images.
Therefore, the device could be used for CT-based stereotactic brain biopsy or for intraoperative or postoperative CT examinations.
There are several advantages and a few disadvantages of the device presented here. It can be fixed to every skull size and shape as previously mentioned and the device can be rigidly fixed to the skull just using 3 screws without the need for fixing the skull to the table.
This surgical setting allows more flexibility for the surgeon performing the procedure. Furthermore, the biopsy procedure is easy to perform, and it requires only 2 people. The surgeon does not need intensive training, like that needed for other biopsy systems, before using this stereotactic device. 12,16,20,27 However, the construction of the biopsy frame itself based on the MR images should be performed by a specifically trained person.
Additionally, the system allows the sampling of multiple brain specimens. The frame can either be constructed with multiple biopsy ports with different trajectories or, alternatively, a spacer secured to the biopsy needle can be used to vary the biopsy depth, allowing tissue sampling from different areas of the lesion along 1 trajectory.
One disadvantage of the system presented here is the separation of imaging and the biopsy procedure itself. Because the construction, 3D printing, and steam sterilization of the frame require up to 3 days, there are 2 procedures under general anesthesia that are necessary.
Furthermore, the dog must have the bone anchors placed and covered by skin closed with 1 skin suture while the biopsy device is being constructed. Those bone anchors must stay in place until the biopsy has been performed. Therefore, there is a potential risk that the position of bone anchors may change or that they may loosen completely due to patient manipulation. This risk can be reduced by the application of a neck collar, but it cannot be eliminated completely. However, the bone anchors are screwed to the skull, and they protrude only 4.5 mm over the surface of the bone. Therefore, the risk of loosening should be low. Another limitation in use of the MRI-based personalized, 3D printed stereotactic brain biopsy device is, that the trajectories that can be used for brain biopsy are fixed by the design. Therefore, the surgeon does not have the option to choose alternative trajectories if he/she encounters procedural technical problems or if the results of the smear preparation of the biopsy specimen indicate insufficient sampling. So the surgeon has to end the biopsy procedure with an alternative brain biopsy method (eg, image-guided free-handed brain biopsy).
Other potential sources of errors could be the accuracy of the 3D printer during the print, mistakes of the surgeon while attaching the markers or the frame to the bone anchors or while adjusting the needle depths, as well as mistakes made by the engineer while constructing the 3D biopsy frame on the PC.
There are some limitations of the study. The brain biopsy device is MRI-based, but the target point deviation was determined with the help of CT examinations. Therefore, image fusion of both modalities was necessary to perform the study. Consequently, the determined needle placement error is the summation of the procedural error of the device itself and the error of the CT and MRI image fusion process. One might speculate that the study design might have artificially reduced the determined needle placement error and therefore improved the accuracy of the device. However, it is much more likely that the combination of 2 steps, each having intrinsic error, would have increased the total error of the procedure. Therefore, the needle placement error of the device presented here might even be lower than what was measured.
Another limitation is the fact that the accuracy was tested for 2 localizations only.
In conclusion, the MRI-based personalized, 3D printed stereotactic brain biopsy frame is a relatively inexpensive, highly precise, and economical way of sampling brain tissue in dogs of all sizes and with different head shapes.

ACKNOWLEDGMENTS
We acknowledge support from the German Research Foundation

CONFLICT OF INTEREST DECLARATION
Authors declare no conflict of interest.

OFF-LABEL ANTIMICROBIAL DECLARATION
Authors declare no off-label use of antimicrobials.

INSTITUTIONAL ANIMAL CARE AND USE COMMITTEE (IACUC) OR OTHER APPROVAL DECLARATION
Authors declare no IACUC or other approval was needed.

HUMAN ETHICS APPROVAL DECLARATION
Authors declare human ethics approval was not needed for this study.