Steering a Tendon-Driven Needle in High-Dose-Rate Prostate Brachytherapy for Patients with Pubic Arch Interference

Abstract

High-dose-rate brachytherapy (HDR BT) is a radiation therapy that places radioactive sources at cancerous tissue using needles. HDR BT offers better dose conformality and sparing of clinical structures, lower operator dependency, and fewer acute irritative symptoms compared to the other form of BT (low-dose-rate (LDR)). However, use of HDR BT is limited for patients with pubic arch interference, where the transperineal path to the prostate is blocked. This study aims to introduce a tendon-driven needle that can bend inside tissue to reach desired positions inside prostate. Initial experiments in a phantom tissue showed the feasibility of the needle to get around the pubic arch for placement at hard-to-reach target positions.

Video Abstract

Download video file ^{(8.9MB, mp4)}

I. Introduction

Prostate adenocarcinoma is the second most common malignancy in American men with about 191,930 new cases in 2020 [1]. Current treatment options include radical prostatectomy, brachytherapy, external beam radiation therapy, hormonal therapy, and cryotherapy. Of these, brachytherapy, an outpatient procedure where local radiation is used to irradiate the cancer inside the prostate, is one of the popular modalities [2], [3].

Patients with early stage prostate cancer are normally selected for low-dose-rate brachytherapy (LDR BT) [4]. In the current state of the art, a ‘stepper’ device allows calibrated linear translation of a transrectal ultrasound (TRUS) through the rectum, while a set of needles is used for percutaneous implantation of radioactive seeds [5]. The physician uses the needle to deposit ~ 80 (ranges from about 50–100) seeds in the prostate. Several methods, such as CT and MRI, are utilized for post-implant dosimetry [6]. Since a high radiation dose with sharp fall-off is delivered to the prostate gland, a precise placement of the seeds is required to ensure an effective treatment [4], [7]. Limited needle actuation near the needle’s entry point into tissue, movement of the target location during needle insertion [8], [9] along with the distribution of 70% of the tumor foci at the peripheral zone of the prostate [10] result in additional difficulties for accurate seed placements and desired dose distribution.

High-dose-rate brachytherapy (HDR BT) involves careful insertion of 18- or 17-gauge needles (catheters) inside the prostate through perineum. This procedure is done manually by a physician under TRUS guidance. Rigid needles and typical templates restrict the needle’s maneuverability, and thereby, puncture of organs at risk such as penile bulb and related vasculature is often unavoidable. After needle placements, a CT scan provides 3D image of the prostate and catheters, as well as the critical organs like the bladder and rectum. The CT images help deliver radiation only to the target areas inside the prostate and not the critical organs. A robotic device, called remote afterloader, is then used to insert a radiation source called Iridium through several channels that are connected to the implanted catheters. The source of radiation is welded on the end of a wire, enabling removal after the procedure. The device is programmed to deliver different radiation doses at different positions in the prostate.

HDR BT eliminates the limitations in seed loss or displacement in LDR BT that can lead to suboptimal dosimetry, including cold areas within the prostate and higher dose than intended to the urethra, rectum, and bladder [11]. Most studies have reported comparable outcomes using either HDR or LDR BT [12]. Benefits of HDR BT include better dose conformality and sparing of clinical structures, lower operator dependency, and fewer acute irritative symptoms. Another study argued that the declining trend in use of BT [13]–[15] is due to lack of radiation oncologists with the skills needed to perform high quality BT implants.

Studies have reported that the existing BT procedures can result in side-effects such as edema in tissue, incontinence, and impotence. The side effects are a result of excessive radiation and needle penetration into sensitive organs such as the urethra, bladder, rectum, penile bulb, cavernous veins, and neuro-vascular bundles. Also, use of HDR BT is limited in patients whose pubic arch obstructs the transperineal path to the prostate, thereby interfering with needle placement [16]. A study showed that the procedure was practical only for 24 out of the 40 patients studied due to pelvic bone arch interference [17]. Substantial pubic arch interference (PAI), which is more likely in patients with a large prostate, makes it difficult to achieve adequate source placement in the anterior and lateral portions of the prostate [18]. Even a narrow pubic arch may prevent proper implantation in a small prostate gland [18]. Known strategies to overcome this problem including oblique catheter insertion and pelvic rotation [19], [20] are not optimal. Other methods suggested insertion of skew-line [21] or oblique needles [20] to pass the pubic arc.

Placement of radioactive seeds or placement of needles at positions close to urethra or rectum in LDR and HDR BT, respectively are also concerning. This could cause unnecessary dose delivery and damage to these healthy organs. Active needle steering can alleviate this concern by precise placement of seeds or needles at desired locations.

This work introduces a tendon-driven needle that can bend inside the tissue around the pubic arch to reach the target positions inside the prostate that are difficult or impossible to reach. Contribution of this work relies on the needle and feasibility studies that are specifically designed and shown for HDR BT.

II. Background and Related Work

Passive and active needles [22]–[25] have been proposed in the past decade. Bevel-tip [26], pre-curved [27], kinked needles [28], and concentric pre-curved tubes [29] are among the most effective proposed needle steering methods. Some research groups (Okamura [30], Webster [31], Desai [32]–[34], Esashi [35], and Armand [36]) have used cut-out pattern (notches) of different forms on a superelastic nitinol tube to achieve a higher flexibility. Cable-driven needle bending has also been studied before [31], [33], [37] using a feedback sensor to locate the needle tip to form a closed-loop control system. The steerable needle introduced in this work provides reliable bending in tissue to reach desired positions in prostate gland in LDR and HDR BT. The tendon-driven needle, its flexible section, bending capabilities, and interior tube are specifically designed for application in HDR BT.

III. Robot Design and Modeling

A. Needle Bending Requirements

This section describes a scenario where the patient’s pubic arch interferes with needle insertions in HDR BT. While Pubic Arch Interference (PAI) is more common for patients with a large prostate gland (volume larger than 50 cm³), it has also been observed in patients with a narrow pubic arch and a relatively smaller prostate [38]. PAI has been reported in 10% of patients with early stages of prostate cancer, to whom HDR BT is not a treatment option [38]. Bending the needles makes it possible to reach the target areas that are not reachable by straight needles and to avoid puncturing urethra during needle placements. To estimate the amount of needle bending required to reach the target positions, we consider a scenario with 10mm of interference between the pubic arch and the prostate gland. Figure 1 shows the coronal view of the workspace in BT with needles inserted from the perineum into the prostate. In HDR BT, around 15 needles are placed inside the prostate to cover the prostate volume. However, as shown in the figure, some target areas are not reachable with straight needles. The dimensions of the prostate (44mm in width, 31mm in height, and 38mm in length) were chosen based on average prostate size for male patients between the ages of 40 and 50 [39]. It was estimated that about 28 degree of angular deflection is sufficient for the needle to reach the target (marked in the figure at a depth of 140mm).

Figure 1.

Coronal view of the workspace for prostate BT. Needles are inserted from the perineum into the prostate.

B. Needle’s Flexible Section Design and Manufacturing

Figure 2a shows the flexible section of a tendon-driven needle designed to curve inside the patient’s body to reach the hidden targets in the prostate gland from the perineum avoiding obstacles like the pelvic bone. Figure 2b shows the dimensions of each notch that needs to be carved on the needle tube for improved flexibility. The needle was made from a nitinol tube (Johnson Matthey, London, UK) with an outer and inner diameter of 1.80 and 1.55mm, respectively, and tube thickness of 0.125mm. A series of seven small notches were made on the needle tube to create a compliant section for additional flexibility. The notches were made in the lab using normal machining tools such as a Dremel, and ultra-thin cut-off discs with a thickness of 0.3mm and 0.127mm (Gesswein & Co., Inc., Bridgeport, Connecticut). A custom-made fixture (shown in Figure 2c and 2d) with a micrometer screw gauge was used to hold the needle tube for machining (carving the notches) and move the needle tube precisely to set desired spacing between the notches. The properties of the nitinol tube are affected by the temperature. Therefore, the cutting process was performed slowly to allow an effective heat transfer and to avoid high temperatures. The dimensions of the seven notches are listed in Table 1. Previous studies have shown [32], [40] that other manufacturing methods such as electrical discharge machining (EDM) or femtosecond laser can produce round and smooth corners to alleviate stress concentrations. In this work, however, we relied on traditional machining process. The average cut width (t), depth (d), and distance between the notches (d_n) were 0.14±0.05, 1.57±0.03, and 0.36±0.06mm, respectively. These values were selected to result in the required needle bending to avoid PAI (as described later in Section III C). A round hole was made on each slit with an average diameter of 0.59±0.03mm. The overall length of flexible section was 5.86mm. Two small holes (diameter of about 0.25mm) were made close to the distal end of the needle for a tendon (a nitinol wire of 0.13mm diameter) to get attached to. The tendon was looped in and out of the holes to secure its place. Tension of the tendon assures bending in the direction of the notches.

Figure 2.

(a) Design of the flexible section of the needle, (b) notch parameters, (c) manufacturing fixture, (d) ultra-thin cut-off disc, (e)heat shrink cover, and (f) deflected shape of the needle.

Table 1.

Dimensions of the seven notches carved on the needle tube. Units are in mm.

Notch	1	2	3	4	5	6	7
Cut Width (t)	0.12	0.15	0.10	0.09	0.13	0.12	0.25
Round Hole Diameter (dh)	0.61	0.61	0.61	0.60	0.60	0.55	0.56
Cut Depth (d)	1.60	1.59	1.58	1.53	1.52	1.56	1.60
Distance between the Notches (dn)	N.A.	0.29	0.45	0.40	0.30	0.39	0.34

The flexible section of the needle was covered with a biocompatible ultrathin (wall thickness of 76μm) medical-grade heat shrink tubing made from Polyolefin (Cobalt Polymers, Cloverdale, CA) to prevent rupture of the tissue at the sharp edges of the notches. The heat shrink cover also provides insulation to avoid tissue from entering the needle tube. Figures 2e and 2f show the needle in straight and bent configurations, respectively. The heat shrink cover slightly increased the outer diameter of the needle to 1.876mm and the stiffness of the flexible section. The flexible joint of the needle was bent about four degrees when covered with the heat shrink tubing. To test the insulation, the needle tube was filled with water with both ends of the tube closed. The cover was observed for 30 minutes, and no leakage was reported. The unsharpened cylindrical needle tip was blocked with epoxy to avoid tissue penetration into the needle tube.

To make sure that the needle can host a source of radiation in HDR BT, or to pass the radioactive seeds (0.8mm diameter and 4.5mm long) in LDR BT, a flexible guidewire of 0.65mm diameter was passed inside the needle tube, while the needle was bent at its maximum bending angle. The study showed that this guidewire can easily pass inside the needle tube when bent.

C. Modeling and Bending Calculations

Equations 1 to 4 explain methods to estimate the bending angle for the needle’s flexible section. For the notched needle tube presented in Section III A, the neutral bending axis can be calculated using the following equation [41]:

$$\overline{y}=\frac{4\left(r_o^3{\mathit{\text{sin}}}^3\left({\varphi }_o\right)-r_i^3{\mathit{\text{sin}}}^3\left({\varphi }_i\right)\right)}{3\left(r_o^2\left(2{\varphi }_o-\mathit{\text{sin}}2{\varphi }_o\right)-r_i^2\left(2{\varphi }_i-\mathit{\text{sin}}2{\varphi }_i\right)\right)}$$ (1)

where $\overline{y}$ is the distance moved from the center of the tube, r_o and r_i are the outer and inner radii of the tube, respectively, ϕ_o and ϕ_i are found by the following equations:

$${\varphi }_o=\mathit{\text{arccos}}\left(\frac{d-r_o}{r_o}\right)$$ (2)

$${\varphi }_i=\mathit{\text{arccos}}\left(\frac{d-r_o}{r_i}\right)$$ (3)

where d is the notch depth. The bending angle for each notch can be estimated as:

$${\theta }_i=\frac{h}{r_o+\overline{y}}$$ (4)

where h is the height of the neutral axis for each, which in this work is equal to the height of the cut width (t). For the notch dimensions shown in Figure 2b, an angular deflection of 4.58 degrees is estimated on each notch. Assuming that the total deflection is distributed equally along all the notches, the bending angle of the whole flexible section can be found by multiplying the bending angle of each notch by the number of notches (seven notches for our needle). Thereby, an overall bending angle of is about 32 degrees for the notched needle introduced in this work.

D. Tendon Actuation and Control System

The tendon tension (F) that is required to realize a specific bending on the flexible section of the needle should be carefully estimated to design a functional actuation and control system. The following equation was used to calculate the amount of torque (T) required to pull the tendon and consequently bend the needle, similar to lowering down a load on a lead screw:

$$T=\frac{F{d}_m}{2}\left(\frac{l+\pi f{d}_m\mathit{\text{sec}}\alpha }{\pi d_m-fl\mathit{\text{sec}}\alpha }\right)$$ (5)

where d_m, l, f, and α are the mean diameter of a single thread, tendon displacement, coefficient of friction, and thread angle, respectively. We estimated that a maximum of 22N of force (tendon tension) must be sufficient to bend the notched needle (presented in Section III A) inside the tissue. With this amount of tension, an 8mm diameter lead screw, friction coefficient of 0.25 for threaded pairs of a steel and dry screw, and thread angle of zero for square threads, the maximum torque was estimated as 0.41N.m. Accordingly, a DC motor, gear, and a lead screw combination was selected and purchased from Maxon Group, Sachseln, Switzerland.

The combination consists of a 0.5W Maxon DC motor RE 8 Ø8 mm, Precious Metal Brushes) with an 8mm diameter lead screw drive (GP 8 S Ø8mm, Metric spindle, M3 × 0.5), and an encoder (MR, Type S, 100CPT). The Maxon motor was connected to a computer using a EPOS4 digital positioning controller and programmed using Maxon Group’s software, EPOS Studio. The software communicates with Maxon Group’s EPOS4 positioning controller to give the user control on the position, velocity, and acceleration profiles. In order to run under the safe conditions of the motor system at high incremental position value (600000inc or 3mm), an rpm of 10000 was chosen with a 500rpm/s acceleration and deceleration rate to form a trapezoidal profile. This allows the system to quickly reach the target position without overshooting.

Two holders were designed, and 3D printed to hold the motor and the lead screw (shown in Figure 3a). Holder 1, housing a nut on the lead screw, was mounted on a 50mm long linear rail guide (MGN9H) to translate rotational movement of the motor shaft to linear movement. Holder 2 was stationary, keeping the position of the motor. The holders are designed to ensure that all the movement occurs on the same plane. The free end of the tendon was fixed on Holder 1. The tendon is threaded through the midpoint of the holder to ensure colinear movement.

Figure 3.

Actuation and control system to pull the tendon and bend the notched needle: (a) assembled on a linear stage for needle axial insertion, and (b) Maxon motor, gear box, and lead screw.

The Maxon motor, gear box, and lead screw (shown in Figure 3b) were mounted on a linear stage for needle insertion tests in a phantom tissue. The Maxon motor pulls the tendon axially to bend the needle. The motor is fixed to a linear motion guide rail with the lead screw nut being held in place on a movable platform along the rail. The needle is considered at the maximum bending position when the cable tendon is pulled 3mm in the axial direction due to the notches in the needle touching each other.

IV. Results

A. Needle Deflection in Air

Figure 4 shows the angular bending of the needle realized by the tendon displacement in air. The error bars were calculated to show the higher and lower values in five trials. A maximum bending angle of 30.92 degrees was demonstrated, which is comparable with our bending calculations of 32 degrees estimated in Section III B. The agreement was reasonable considering our methods for manufacturing the needle.

Figure 4.

Angular bending of the tendon-driven needle in air.

B. Needle Bending in Phantom Tissue to Avoid PAI

Figure 5a shows the deflection of the needle when inserted to a depth of 140mm into a phantom tissue. The phantom material was a polyvinylchloride (PVC) gel (M-F Manufacturing Co., Ft. Worth, TX), with 3:1 ratio of plastic (PVC suspension) to softener that gives an indentation elastic modulus of about 25.6±0.6kPa [42] that is comparable to the modulus of normal prostate tissue (22.74 kPa).

Figure 5.

(a) Tendon-driven needle inserted to a depth of 140mm into a phantom tissue, (b) - (d) lateral, vertical, and 3D position of the needle tip scanned by an ultrasound device, respectively.

To avoid the obstacle in the scenario described in Section III A, the needle was first inserted to a depth of 120mm and then the tendon was pulled at a rate of at 0.1mm/s while being inserted to a depth of 140mm to reach the target. The insertion and actuation rate were realized by motorized stages and the values were set manually in an open-loop manner. Upon completion of the needle insertion task, the needle was scanned by an ultrasound device (Chison, Eco 5) similar to [43]. The lateral and vertical position of the needle tip are shown in Figure 5b and 5c, respectively. The 3D position of the needle tip is plotted in Figure 5d. The out-of-place displacement of the needle was due to its slightly curved shape of the needle prior to actuation.

To show the needle’s capability to bend inside the phantom tissue, needle bending was measured at different tendon displacement ranging from 0.5 to 3.0mm. The angular bending of the needle inside the phantom tissue is shown in Figure 6. Smaller needle bending was observed inside the phantom tissue compared to the needle bending in air (Figure 4), which was expected due to the tissue resistance.

Figure 6.

Angular bending of the tendon-driven needle in phantom tissue.

V. Discussion

This work introduced a tendon-driven notched needle that can be used for placement of needles inside the prostate gland for patients with pubic arch interference. The needle was designed and manufactured (with traditional machining tools in the lab) to curve inside the patient at a maximum bending angle of 30 degrees to pass a pubic arch with 10mm of overlap with the prostate gland. The needle can also reach to the out-of-plane target positions via axial rotation at its end. Steering the needle inside the tissue requires careful consideration of tissue and target movement with a proper and synchronized control of needle insertion and bending. Acquiring target positions is left for future studies. Future work will also focus on needle arrangement planning to ensure that curved needles can meet the dose requirements in the HDR BT.