FIELD OF THE INVENTION

This invention relates to hydrocephalus shunt monitoring.

BACKGROUND OF THE INVENTION

Hydrocephalus is caused by the abnormal buildup of cerebrospinal fluid in the ventricles of the brain. Currently, there is no cure for hydrocephalus. The most common way to manage hydrocephalus is to insert a shunt, a flexible tube positioned to drain excess fluid from the ventricles to another part of the body, such as the abdomen.

More than 125,000 shunts are implanted each year in the United States at a cost of $2 billion. Although shunt placement is the most common procedure performed by neurosurgeons, shunts remain among the most failure-prone life-sustaining medical devices implanted in modern medical practice. In pediatric patients, shunts have a 50% failure rate within the first two years of placement and will fail often throughout a patient’s life due to blockage. Nearly half of the cost associated with shunt implantation is associated with shunt revisions.

Most patients with shunt failure experience non-specific symptoms such as irritability, headache, nausea, vomiting and lethargy, but the consequence without timely treatment is fatal. This problem is more prominent for pediatric patients who have more difficulty communicating their symptoms. To detect a shunt failure, a patient must get a series of X-ray scans of the head, chest, and abdomen, while the size of the ventricles is assessed by Computed Tomography (CT) of the head or Magnetic Resonance Imaging (MRI). This requires a visit to the emergency room or medical facility with the capability to manage shunts. Additionally, the need to confirm shunt failure by x-rays or CT scans increases the long-term effects of ionizing radiation. Excessive exposure to ionizing radiation is of greater concern particularly for pediatric patients. Though MRI does not expose patients to radiation, machines are often only available at hospitals.

To reduce hospital visits and the exposure to radiation in hydrocephalus patients, an at-home, nonradioactive solution to detect shunt failures is preferred. This is especially relevant for pediatric patients, who, besides susceptibility to radiation, are less capable of communicating their symptoms. Several systems based on the thermal flow meter are available for this purpose.

The thermal flow meter measures the flow rate in a duct by heating or cooling the flow with a temperature actuator, and monitoring the consequent spatial and temporal temperature profile to infer the flow rate within the duct. Among the available thermal flow meters for hydrocephalus shunt, one system cools down a large area on the neck with an ice pack and measures the temperature drop downstream. However, the system is limited to exclusive use in clinics due to the cooling mechanism and the bulky temperature sensing system, and yields an inadequate accuracy. Another system is also placed on the neck, but uses a miniature temperature actuator to generate a constant heating power above the skin surface, and detects the temperature changes upstream and downstream of the heater. Despite the reduced form factor and improved thermal efficiency, the amplitude of the detected temperature change is attenuated by heat transfer through skin between the device and the shunt. Such small temperature signals compromise the reliability of measurements due to the variation of ambient temperature. Although higher power may increase the signal amplitude, it would also increase the temperature of the body tissue, causing discomfort and potentially thermal damage to the patient.

Accordingly, there is a need in the art to identify new technology to at least address some of the current issues/problems. SUMMARY OF THE INVENTION

Embodiments of the invention can be summarized as follows. A method of measuring the flow rate of cerebrospinal fluid in a hydrocephalus shunt. The method includes heating or cooling the cerebrospinal fluid in the hydrocephalus shunt with one or more thermal elements, such as Joule heaters or thermoelectric coolers. The spatial or temporal temperature profile induced by the thermal elements differs with different flow rates of the cerebrospinal fluid in the hydrocephalus shunt, including the flow rate of zero as in the case of blockage. One or more sensors are used to measure the temperature-related changes induced by the thermal elements. The measurement results are used to infer the flow rate of cerebrospinal fluid in the hydrocephalus shunt. The heating or cooling powers of the thermal elements are modulated by a temporal pattern of a sinusoidal frequency, a combination of multiple sinusoidal frequencies, or any waveform designed to separate the effects related to the flow rate of cerebrospinal fluid from other interferences unrelated to the flow rate of cerebrospinal fluid. The frequencies of the thermal element modulation are in a range of ImHz to lHz.

The average power of all thermal elements combined over the duration of a measurement of the flow rate is no less than 1 mW and no higher than the level that causes thermal damage with the embodiment. In another embodiment, the average power of all thermal elements combined over the duration of a measurement of the flow rate is no less than 0.5 mW.

A hydrocephalus shunt monitoring apparatus having one or more thermal elements to alter the temperature of the cerebrospinal fluid flow in a region, and one or more sensors deployed to measure the temperature-related changes induced by the thermal elements. The thermal elements or the sensors are placed in one or more implantable enclosures to closely contact the hydrocephalus shunt tube. The smallest distance between the apparatus and the shunt tube is no more than 2 mm, and the distance between at least one of the thermal elements and one of the sensors is no more than 5 mm. The implanted part of the apparatus is powered by an external part either through wires or wirelessly. The measurement results of the cerebrospinal fluid flow rate are transmitted to the external part for display or analysis.

It is noted that the invention can be embodied as an implantable thermal flow meter, but also as a method of sinusoidal heating or cooling, and a system with an implantable device, which could be independent of each other. It is further noted that the sinusoidal temperature modulation is a method applicable to not only implantable devices, but also other thermal flow meter for hydrocephalus shunt monitoring, including skin-mount (epidermal) devices.

The present invention provides a method of measuring a flow rate of a cerebrospinal fluid in a tube of a hydrocephalus shunt. The method distinguishes applying a temperature to the cerebrospinal fluid in the tube of the hydrocephalus shunt. The power output of the applied temperature is modulated by a sinusoidal pattern having a central frequency ranging from ImHz to lHz. In one example, the modulation lasts for at least one sinusoidal period. The temperature can be applied by one or more thermal elements. At least one of the thermal elements is placed no more than 2 mm from the hydrocephalus shunt. The temperature application could involve heating of the cerebrospinal fluid or cooling of the cerebrospinal fluid. The temperature-related changes of the cerebrospinal fluid in the hydrocephalus shunt induced by the temperature application are then measured by one or more sensors. At least one of the sensors is placed no more than 2 mm from the hydrocephalus shunt. From these measured temperature changes, the flow rate is inferred for the cerebrospinal fluid in the hydrocephalus shunt.

In one embodiment, the distance between at least one of the thermal elements and one of the sensors is no more than 5 mm.

In yet another embodiment, the one or more thermal elements and the one or more sensors can be packages in a package smaller than 10 cm along an axial direction of the tube of the hydrocephalus shunt. In a further embodiment, this package can be implanted in a person.

BRIEF DESCRIPTION OF THE DRAWINGS

For interpretation of the gray-scale in the drawings the reader is also referred to US63/189519 filed May, 17, 2021 to which this application claims the priority /benefit, and which includes color drawings.

FIG. 1 shows according to an exemplary embodiment of the invention a schematic of a device for measuring a flow rate of a cerebrospinal fluid in a tube of a hydrocephalus shunt. It is noted that there are temperature sensors for measuring temperature upstream (to the left of the heater or thermal element) and downstream (to the right of the heater or thermal element). It is further noted that the distance between at least one of the thermal elements and one of the sensors is no more than 5 mm. In other words, temperature can be measure very close to the thermal element or heater.

FIG. 2 shows according to an exemplary embodiment of the invention a circuit diagram of a benchtop prototype. FIG. 3 shows according to an exemplary embodiment of the invention (top) simulated steady-state temperature distribution along shunt when a constant heating power of 16 mW is applied, and (bottom) difference in temperature distribution from no-flow condition.

FIG. 4 shows according to an exemplary embodiment of the invention simulated temperature of the temperature sensors (thermistors, top) and the fluid in the tube (bottom) during heating with a single-material (mat.) and bi-material device

FIG. 5 shows according to an exemplary embodiment of the invention ambient temperature variation in a laboratory setting. Each thin irregular trace represents an independent measurement, and the thick center trace around zero and shade are the mean and standard deviation, respectively.

FIG. 6 shows according to an exemplary embodiment of the invention temperature readings with heating at a constant power. Top: temperature change as a function of time at different sensors and flow rates. The shaded areas represent the range of variation for each line. Bottom: steady-state temperature change probed at 6 minutes after the onset of the heating. The error bars indicate the range of one standard deviation.

FIG. 7 shows according to an exemplary embodiment of the invention temperature readings with sinusoidal heating. Top: temperature change as a function of time at different sensors and flow rates. The shaded areas represent the range of variation for each line. Bottom: peak-to-peak modulation depths of the sine waves in the steady state. The error bars indicate the range of one standard deviation. FIG. 8 shows according to an exemplary embodiment of the invention an assembly drawing of an implantable packaging and the placement of essential elements inside.

FIG. 9 shows according to an exemplary embodiment of the invention traces of temperature change over time measured by sensor #3 at different CSF flow rates, when heating is applied at a constant power. The traces not only suffer from variation across measurements (shown in FIG. 6), but also the overlapping between different flow rates, which compromises the reliability of flow rate measurements.

FIG. 10 shows according to an exemplary embodiment of the invention traces of temperature change over time measured by sensor #2 at different CSF flow rates, when heating is sinusoidally modulated at 10 mHz. The flow rate is inferred from the spectral intensity at 10 mHz after a Fourier transform of the time-domain signal, which is related to the peak-to-peak value in the zoom-in view

FIG. 11 shows according to an exemplary embodiment of the invention that the spectral intensity of the temperature measured by sensor #2 at 10 mHz is highly consistent over time, and well distinguished between different flow rates

FIG. 12 shows according to an exemplary embodiment of the invention the bi-material cross-sectional profile of the implant, where the heaters (H) and the thermistors (T) are coupled to the shunt with a thermal-conductive material such as thermal paste, while insulated from each other with a second material of low thermal conductivity, such as foam. DETAILED DESCRIPTION

In one embodiment, the present invention addresses the limitations in the art by designing an implantable thermal flow meter, which wraps around the shunt tubing to reduce the heat attenuation between the device and the CSF flow, and allow sensors to be placed in their ideal position with respect to the heated region. By designing the material profile to direct heat transfer toward the CSF flow, and by using sinusoidal modulation to differentiate signals from ambient temperature variation, accurate flow rate measurement is obtained while minimizing size and maximizing efficiency, with an improvement on the signal -to-noise ratio (SNR) by orders of magnitude. Sinusoidal modulation of the heating power

Current thermal flow meters for the hydrocephalus shunt apply a step function of constant power on the temperature actuator for each measurement. To achieve more reliable measurements of the flow rate, embodiments of the invention apply a sinusoidal modulation to the heating power to differentiate between the useful signal that reflects the flow rate, and the inevitable noise caused by the variation of the ambient temperature. Because sinusoidal functions are the eigenfunctions of any linear time-invariant system, the temperature change is also sinusoidally modulated at the same frequency of the heating power, providing a clear spectral distinction between the signals that reflect the flow rate and the random variation of the ambient temperature. To differentiate the signals from the ambient temperature variation, a transform from the time domain to the frequency domain, such as a Fourier transform, is performed on the time sequence of temperature measurement, from which the spectral intensity at the frequency of modulation is extracted and used to infer the flow rate. The spectral intensity is directly related to the modulation depth of the time sequence (FIG. 7). To achieve time fidelity in the spectral analysis, a sliding window, such as the Blackman window, may be applied to the time sequences. With an average power of 16 mW and 10 mHz modulation, in an exemplary embodiment, the inventors were able to achieve an average SNR of 32.5 dB, better than the step-function (constant) heating with the same hardware by more than 2.5 orders of magnitude (300 times) with just half of the heating power. Assuming the signal amplitude linearly increases with the heating power, SNR of 5.9 dB, equivalent to that with the 32 mW step-function heating, is achievable with sinusoidal power of lmW. SNR of 3 dB, representing the signal is twice as strong as the noise, requires a heating power of 0.5 mW.

Implantable thermal flow meter for the hydrocephalus shunt

When the device is placed on the skin surface, the amplitude of the detected temperature change is attenuated by heat transfer through skin between the device and the shunt. Furthermore, for a device on the skin surface, to capture the effects related to the CSF flow in the shunt tube buried under the skin, instead of direct thermal conduction in the skin tissue, a lateral distance between the heated region and the temperature measurement position is required. The distance should be at least comparable to the thickness of the body tissues between the device and the shunt tube, which further attenuates the signal amplitude. The inventors aimed to improve the reliability of measurements by designing an implantable thermal flow meter which wraps around the shunt tubing to reduce the heat attenuation between the device and the CSF flow. The optimal site of temperature measurement that provides the greatest amplitude of signal is determined to be at the same position of the heater along the axis of the shunt tube (FIG. 3), which is only possible with an implantable device, where multiple components are placed around the circumference of the shunt tube at the same axial position (FIG. 12).

A Computer-Aided (CAD) model of the design was constructed based on sizes of internal electrical components, and a titanium enclosure (FIG. 8). This device has a comparable size to existing shunt valves, such as the Medtronic Strata valve (4.7 x 1.6 x 0.28 cm) or Codman Certas plus programmable valve (small: 3.19 x 1.3 x 0.7 cm; regular: 3.4 x 1.65 x 0.7 cm). A flow meter this size could be placed on the distal end of the shunt in the peritoneal cavity, which is an ideal location due to several reasons:

• Surgeons often make an incision in the upper abdomen when installing the shunt, and thus could easily slip the device onto the shunt here.

• The abundance of fatty tissue near the device would be more comfortable for the patient, as opposed to in the head or neck.

• The device would be close enough to the skin surface to allow wireless charging. The battery is the largest component of the flow meter device. Based on the 16 mW of heating power used in a prototype, a single fluid-flow measurement of 6 minutes would require ~6 Joules of energy. A 3 V, 3.4mAh battery allows six independent measurements. In principle, lower heat magnitudes could also be used, allowing more flow measurements per battery charge, or a smaller battery. The flow meter battery would be charged using inductive coupling with an external charging device. For example, the Qi standard is effective up to 4 cm. The device can be paired with a wireless charger, and provide two modes of operation:

• On-demand, in which the device is wirelessly powered and functions regardless of the battery status, and

• Battery-powered, in which the device measures independently up to a number of times limited by the battery life and the heating power.

Embodiments of the invention provide:

• An advanced hydrocephalus shunt system that integrates the implantable sensor described above,

• On-demand readings of the flow rate for hydrocephalus patients on their smart devices, • A hydrocephalus shunt system that continuously records the flow rate for personalized prognostics or scientific research of hydrocephalus,

• The system does not suffer from the heat attenuation in the body tissues between the device and the catheter, so the heating power required to run the measurement is reduced,

• With largely reduced power consumption enabled by the sinusoidal heating scheme, the measurement can be performed frequently to provide a continuous monitoring data of the

CSF flow. The variation of the environment temperature is an intrinsic source of noise, which cannot be reduced by high-quality electronics, but the sinusoidal modulation proposed herein is robust to it. The sinusoidally modulated heating power creates a unique signature frequency in the temperature change downstream, so one can reliably differentiate the signal from the ambient temperature variation, and hence achieve an SNR improved by over 300 times, compared with the traditional step-function heating.

Additional notes

The sinusoidal modulation should be performed at frequencies from ImHz to lHz. The lower limit is set by the time practically required to complete one modulation cycle (1000 second), and the upper limit is determined by the characteristic thermal diffusion length in water (0.4mm in Is).

The minimum average power of the sinusoidal heating is 0.5 mW, based on the discussion on SNR infra. The maximum heating power depends on the design of the apparatus, such that the maximum temperature change in the body tissue does not impose thermal damage to the patient. In an exemplary embodiment, this value should be 200mW, based on the data shown in Table 1 infra.

The heating duration for each measurement should be at least the period of the sinusoidal modulation, and can be as long as the battery/wireless charger allows.

The implant (as shown in FIG. 8) should be less than 10 cm and 3 cm in the axial (along the shunt tubing) and the lateral directions, respectively, for the safety and comfort of the patient. Benchtop Prototype

In an exemplary proof of concept embodiment, a Sophysa B905S distal catheter (Sophysa USA, Inc., IN) of the hydrocephalus shunt was placed in the groove of a polycarbonate breadboard with a height and width matching the diameter of the catheter. Four 2-kQ thermistors (NTCLE203E3202HB0, Vishay Intertechnology, Inc., Malvern, PA) and a 2-1<W resistor were placed on the catheter, which are used as the temperature sensors and the heater, respectively (FIG. 2). The spacing is 5 mm between components, except for Sensor #0, which is placed distally from the catheter for sensing the ambient temperature. Thermal paste (XTM50, Corsair Gaming, Inc., Fremont, CA) was applied to improve the thermal coupling between the electronic components and the catheter. The structure is encapsulated in a semi-cylindrically molded polyurethane foam (SOFF3, Smooth-On, Inc., Macungie, PA) of 12 mm in diameter, in order to minimize heat dissipation into the ambient environment. Auxiliary resistors (R0-R3, Rh) were used to monitor the current of each component (FIG. 2).

Computational modeling

COMSOL Multiphysics 5.3a (COMSOL, Inc., Stockholm, Sweden) was used for finite element modeling of the fluid and heat transfer dynamics of the thermal flow meter in vivo. A computer- aided design (CAD) geometry of the silicone shunt tubing, with a titanium enclosure, electrical surface mount components (microcontroller, battery, resistive heaters, thermistors), thermal insulation, and thermal conductive paste, was created in COMSOL or SolidWorks (Dassault Systemes, Velizy-Villacoublay, France). Restive heaters and thermistors were modeled as homogenous silicon dioxide objects. The entire device was embedded within muscle tissue. Geometry dimensions were altered based on mechanical design optimization. Material properties, specifically, density (p), thermal conductivity (k), and heat capacity at constant pressure (Cp), were determined from published data sheets, COMSOL’ s built-in library, or vendor specifications. Fluid flow within the shunt tubing was modeled as laminar flow, with no outlet pressure and backflow suppressed. Fluid velocities tested were 0, 0.1, and 0.89 ml/min. Heat transfer between all components was modeled through conduction. The entire configuration was set to 37 degrees Celsius at the simulation onset, with the outer boundary of the muscle tissue fixed at 37 degrees Celsius. Resistors were treated as heat sources with a constant or sinusoidal power output, and the average temperature of the modeled thermistors, or various points, were used for the temperature outputs.

Engineering goals and approach One of the goals was to optimize the method and design of an implantable device that can be slid onto a shunt and clipped at an arbitrary location. The inventors aimed to obtain accurate flow-rate measurements, with maximum efficiency, and minimum form-factor size. The target size was similar to currently used shunt valves that are implanted with shunts. Using a combination of experimental results from a benchtop prototype, finite element modeling using COMSOL Multiphysics, and patient feedback, the inventors optimized the algorithmic and mechanical design of such a device.

Position of heaters and temperature measurement

First the ideal position of the temperature sensors was determined along the axial position of the shunt, with respect to the heated region. Initial simulations were conducted with a constant 16 mW of uniform heating around the circumferential direction of the shunt tube. With no fluid flow, the temperature distribution along the axial direction of the shunt was symmetric with respect to the heated region. Increasing values of fluid flow transferred heat downstream, resulting in decreasing steady-state temperatures at the heated region, a non-monotonic trend downstream, and increasing temperatures further downstream (FIG. 3). Overall, measuring temperatures close to the heated region will maximize sensitivity, and result in a smaller form- factor, making this the ideal position for temperature measurement.

Directing heat transfer A potential drawback of heaters and sensors being placed closely is heat being conducted directly from heaters to neighboring temperature sensors, without transferring through the shunt fluid, and consequently not reflecting the flow rate. The inventors ran a simulation using two heating resistors and two thermistors for temperature measurements, placed at the same axial position of the shunt (FIG. 12). By adding thermally conductive paste (k = 5 /m-K, Cp = 1000 ¼g-K) between heaters or sensors and the shunt tube, and separating heaters and sensors with thermally insulative foam (k = 0.0285 ^rin-K, Cp = 1450 ¾g-K), the inventors directed the heat conduction path from heaters, through the shunt fluid, and to the temperature sensors. Simulations confirmed that using thermally conductive paste increased heat flow into the fluid (FIG. 4). A greater temperature difference was observed between flow and no-flow conditions when using a combination of thermally insulative and conductive materials, compared with a thermally intermediate single material (k = 1 ^rin-K, Cp = 1200 ¾g-K) (FIG. 4). Drawbacks of constant heating

Thermal flow meters for hydrocephalus shunts typically apply a step function of constant heating power. Portable devices powered by batteries rely on small steady-state temperature change to measure flow rates. For example, one device heats at 30mW, and the difference in the steady-state temperature between adjacent flow rates is about 0.2 degrees Celsius at best. However, FIG. 5 shows that even in a relatively well-controlled laboratory environment, the temperature variation within 6 minutes, the typical timespan of a measurement, can be as large as 0.3 degrees Celsius. Such variation of the ambient temperature affects the temperature of the inflow, and hence introduces a significant noise that compromises the precision of flow-rate measurements, regardless of the quality of the electronics. The susceptibility of the temperature reading to ambient variation is evidenced by the overlapping ranges of temperature change between adjacent flow rates in FIG. 6, where a step function of constant heating power is applied at 32mW. In addition, the traces of temperature change over time for different flow rates overlap each other, as shown in FIG. 9. The average SNR of the temperature reading, defined as the sum of variance divided by the square of difference between the measurements of two adjacent flow rates, is 5.9 dB. In a typical environment without strictly controlled ambient temperature, inference of flow rate from the temperature sensing using constant heating power is unreliable. Improvement of the SNR by increasing the heating power is possible, but it would also reduce the battery life and raise the risk of thermal damage to the patient. Sinusoidal heating improves signal-to-noise ratio

To achieve more reliable measurements of the flow rate with the same setup, the inventors applied a sinusoidal modulation to the heating power to differentiate between the useful signal that reflects the flow rate, and the inevitable noise caused by the variation of the ambient temperature. Because sinusoidal functions are the eigenfunctions of any linear time-invariant system, the temperature readings are also sinusoidal with no distortion in principle (FIG. 10), and the flow rate may be determined from the specific sinusoidal frequency, the spectral intensity at which is highly consistent over time (FIG. 11). As shown in FIG. 7, with an average power of 16 mW and 10 mHz modulation, the inventors were able to achieve an average SNR of 32.5 dB, better than the step-function heating by more than 2.5 orders of magnitude (300 times) with just half of the heating power. Assume the signal amplitude linearly increases with the heating power, SNR of 5.9 dB, equivalent to that with the 32 mW step-function heating, is achievable with sinusoidal power of 1 mW. SNR of 3 dB, representing the signal is twice as strong as the noise, requires a heating power of 0.5 mW.

Optimal heating parameters

The frequency of the sinusoidal modulation, the duration of each measurement and the heating power should be jointly optimized. With higher frequencies, the shorter duration of measurement is required to record the same number of cycles, but the sensors would also yield a weaker signal (smaller peak-to-peak amplitude) due to the diffusive nature of heat transfer. The heating power should be chosen to render the desired precision of flow-rate measurements, but not much higher than that to avoid excessive heating and conserve battery power. Based on empirically optimization, the inventors chose lOmHz, 6 min and 16 mW for the frequency, the duration of measurement and the heating power, respectively. The duration and the heating power are comparable to existing solutions. COMSOL simulations confirmed that these heating parameters provided enough difference of the temperature reading to not only predict flow and no-flow states within the shunt, but also infer the flow rate (Table 1). The inventors note these thermistor temperature measurements are taken at the same axial position as the heaters, and thus are greater than values obtained from the benchtop prototype, which measures temperature slightly downstream of the heater (FIG. 2). Temperatures outside the flow device increased by less than 1 degree Celsius, well under the temperature limit for biological tissue (Table 1).

Table 1: Simulated temperature measured at the thermistors of the same axial position of the heater, and at the external boundary of the device.

Battery capacity

Next, the inventors sought to determine the ideal battery capacity, which is the largest component of the flow meter device. Based on the investigation of the optimal temperature modulation scheme, a single fluid-flow measurement would require ~6 Joules of energy. Given the intermittent nature of fluid flow, the inventors anticipate up to three measurements need to be taken to accurately conclude a lack of fluid flow. Most patients and caregivers surveyed were fine with a charging frequency of once per week or more, and thus a 3.4 mAh battery size would be ideal. This would allow two independent fluid-flow measurements for a single battery charge, consisting of six heating cycles. As mentioned above, in principle lower heat magnitudes could also be used, allowing more flow measurements per battery charge, or a smaller battery. The flow meter battery would be charged using inductive coupling using an external charging device. For example, the Qi standard is effective up to 4 cm. It was noted that patients also expressed a preference for a continuous monitoring solution over an on-demand solution. However, this would drastically increase form-factor size due to a larger battery, and was not recommended by neurosurgeons.

Form-factor and device location

A computer-aided (CAD) model of the design was constructed based on sizes of internal electrical components, and a titanium enclosure (FIG. 8). This device has a comparable size to existing shunt valves, such as the Medtronic Strata valve (4.7 x 1.6 x 0.28 cm) or Codman Certas plus programmable valve (small: 3.19 x 1.3 x 0.7 cm; regular: 3.4 x 1.65 x 0.7 cm). A flow meter this size could be placed on the distal end of the shunt in the peritoneal cavity, which is an ideal location due to several reasons. (1) Surgeons often make an incision in the upper abdomen when installing the shunt, and thus could easily slip the device onto the shunt here. (2) The abundance of fatty tissue near the device would be more comfortable for the patient, as opposed to in the head or neck. (3) The device would be close enough to the skin surface to allow wireless charging.

Conclusion The inventors have designed a thermal flow meter for hydrocephalus shunts from both hardware and temperature modulation scheme. An implantable solution that wraps around the shunt will allow one to place the sensors closer to the thermal elements for higher sensitivity, and avoid heat loss through the skin. The sinusoidal temperature modulation scheme may readily be applied to other designs, such as the ones described in US8894584 and US20210022609, and the inventors expect a universal improvement in the SNR independent of the hardware. The variation of the environment temperature is an intrinsic source of noise which cannot be reduced by high-quality electronics, but the sinusoidal modulation presented herein is robust to it, because only signals at the specific frequency of temperature modulation (10 mHz in an exemplary embodiment) is utilized to infer the flow rate.

A design can be paired with a wireless charger, and provide two modes of operation: (1) on- demand, in which the device is wirelessly powered and functions regardless of the battery status, and (2) battery-powered, in which the device measures independently up to a number of times limited by the battery life and the power of the thermal element. To determine the optimal power of temperature modulation in vivo requires further experiments, but inventors’ ex-vivo experiment showed that without increasing the heating power, the measurement quality can be largely improved by sinusoidal modulation and the hardware design. Albeit the design is most valuable in pediatric patients, the same device can help adult patients monitor their shunt status as well. In addition, with the device providing reliable measurements of the CSF flow rate, the research community may find its relevance in the study of hydrocephalus - for example, prolonged study of the correlation between CSF production and behavioral patterns.

This application claims the benefit of US63/189519 filed May, 17, 2021 which is hereby incorporated by reference in its entirety.