Design Objectives

  1. Number of electrodes 

Previous iterations of the colonic monitoring device utilized anywhere from 8 to 128 electrodes. While ‘upgrading’ to 128 channels in a recent iteration increased the quality of the GI signals recorded, many current studies only have about 25-30 electrodes in use (Erickson, 2020). Although we wanted to maintain the device’s ability to collect enough data spatially map the active colonic regions, we also needed to stay within our budget and size constraints. Having more electrodes would result in a higher resolution spatial map, but it would also mean requiring more bioamplifiers, incurring a higher cost and requiring more abdominal real estate. One study successfully utilized an 8 x 8 array of electrodes with 2 cm spacing (Seo, 2023). With these considerations in mind, we decided that 32 electrodes with a spacing of 2.5 cm would be the best for our research purposes. 

  1. Predicate devices

    1. Gastric Alimetry System 

When considering upgrading the existing device, we investigated implementing existing colonic monitoring systems to modify. One system was the Gastric Alimetry System (GAS), shown in Figure 8. This system offers an all-in-one electrode patch, eliminating the need to place each electrode one-by-one. Additionally, this device seemed compact and looked as though it would minimally interfere with subjects’ daily tasks. Lastly, Alimetry’s use of an app to track real-time symptoms and their “cloud-based analysis” is an appealing feature. Current research tries to track symptoms throughout the study to align with any increases in colonic activity to characterize what activity matches with specific symptoms. The cloud-based analysis appeals to us, because while we will be storing our 24-hour colonic and cardiac data locally via a micro-SD card, we are also considering pairing the local saving method with a systematic upload via Bluetooth or Wi-Fi, primarily because of the benefits of not having to access the SD card to retrieve data.  

The system had its drawbacks, though. Because it focuses only on colonic monitoring, GAS samples at 4 Hz, which is insufficient for determining the heart rate via R-peak marking—a key tool of investigation in our IBS research, as we are looking in the colonic movements along with exploring hypersensitivity and BGA axis investigations. The device is also very costly, at around $10,000 for a kit that includes an amplifier module, an iPad for cloud integration, and 10 arrays. This would be over our estimated budget, prior to other modifications we would need. In addition to these drawbacks, GAS is marketed to conduct 4.5-hour studies, not 24-hour studies, so there are concerns over the duration of the battery life of up to 22 hours (Gharibans et al, 2022). Because of GAS’s low sampling frequency, high cost, and battery concerns, we decided to not use this device. 

Figure showing a large patch of electrodes across an abdomen

Figure 8: Gastric Alimetry System 

    1. OpenBCI Cyton system 

The second device we considered was OpenBCI’s Cyton+Daisy board (Figure 9). This device appealed to us because it offers its own MCU, onboard micro-SD module, and bioamplifier all in one device. It also comes with a dongle to set up BLE, which is a feature we are considering utilizing. The device only offers 16 channels though, and we would need two to achieve our 32-channel constraint. Given that we would need 2 modules, the total size would be 122.42 x 61.21 mm and the cost would be nearly $4,000. Therefore, we rejected this because of its high cost and large size, which would be uncomfortable for overnight studies, and opted to build our own device. 

Picture of OpenBCI’s Cyton+Daisy board, the Daisy module is mounted on top of the Cyton module. It looks like a white octagon with some small electronics surface mounted

Figure 9: OpenBCI’s Cyton+Daisy board, the Daisy module is mounted on top of the Cyton module 

The last device that we considered was the Sessantaquattro from Bioelettronica (Figure 10). This device was considered due to its thin size of 73 x 105 x 12mm, making it portable. In addition, the device offers 64 channels, which is more than we require, but will be more helpful when spatially mapping the colonic signals. However, when considering other factors, it became clear that we could not choose this option because it is extremely pricey, at around $5,000, and only offers Wi-Fi with no BLE, which is not a deal breaker for us. What is a deal breaker is that users have had poor experiences with the electrode pads compatible with their system, and because we are looking to prioritize user comfort, this is a major drawback for us.  

Another device similar to the one we are aiming to build. It is a small white rectangular device.

Figure 10: Sessantaquattro by Bioelettronica  

  1. Electronics

Our device will include at least three of the electrical components employed by the current system, which are a microcontroller unit, one or several amplifiers, and a low voltage differential signal (LVDS) transceiver. Each of which includes several considerations. The specific units we have chosen or are considering for each component class are discussed latter in section 5b: Electrical Components 

    1. MCU 

We sought a MCU that is small, fast, and powerful. As previously stated, we need to sample at rate of at least 100 Hz, thus we need an MCU that is capable of processing at speeds fast enough to receive, transmit, and write, the required amounts of data within this timespan.  We wanted a microcontroller that contained multiple SPI ports, as SPI will be the electrical interface between all of the aforementioned components, and the interface that allows the MCU to write to the micro-SD card. Additionally, since our system is required to store the recorded data locally, we wanted an MCU that contained on-board SD module to eliminate the need for external, add-on shield module, which would require additional space.  Lastly, we considered the Bluetooth and Wi-Fi capabilities of the MCUs. Neither are strictly necessary for the required functions of our device; however, they may later be useful in implementing a user interface that allows real time display and monitoring of the data. 

    1. Amplifier 

Investigation into the bioamplifier also clarified a need for one with 32 channels and electrodes, and one that ideally has on-board adjustable filters that allow for a 0-100 Hz bandwidth, which technically speaking implies an effective anti-aliasing passband of 50 Hz, the Nyquist frequency for a sampling rate of 100 Hz, since any frequency above this sampled at 100 Hz will be aliased. We want the lower end of the bandwidth of the bioamplifier to be as close to 0 Hz as possible to record low-frequency colonic signals that may be as low as 1-2 cycles per minute, or 0.05 Hz (Sarna, 2010) with attenuation. The system need must be capable of sampling at a rate of at least 100 Hz to record colonic and cardiac QRS signal components (Huizinga, 2021). 

    1. LVDS transceiver

The only concern for the transceiver, other than its ability to perform its function of interfacing the MCU and amplifier, is minimizing its size. To do this, we will surface mount the LVDS-CMOS transceiver chip rather than using the headstage that the current design utilizes, which takes up more space. 

    1. Other electrical components

Our device is also intended to work remotely and continuously for at least a 24-hour period, so it will necessarily include a battery. We are also currently considering implementing an OLED screen that displays real-time data signal, either in place of or in conjunction with a Wi-Fi signal. 

  1. Mechanical Housing 

The aim is for the housing box to be smaller than the current box, which is 13.1 x 8.8 x 2.5 cm. Beyond this, the mechanical housing needs to be comfortable to wear and not disrupt day to day activities. Human survey work will help us determine how small the housing unit needs to be in order to achieve maximum comfort levels. When thinking of potential day to day activities, we must consider the subject walking, sitting, and sleeping. All of which have different movements that could cause the box to feel uncomfortable for the subject. The material of the housing unit will also be important to consider as we do not want the material itself to cause any discomfort.  

  1. Electrode Contacts

The current design uses 32 contacts and we will plan to do the same. Currently, research in our lab utilizes Ambu wet electrodes of dimensions 44.8 x 22/30 x 22 mm (Ambu Forever Forward, n.d.). The small size of the electrodes allows us to attach more on subjects’ abdomens, but there have been concerns from previous testing in our lab over the wet electrodes’ tendency to cause skin irritation, especially over longer periods of time than our current 3-4-hour studies. The gel on the wet electrodes contains AgCl, which some individuals show to have a sensitivity to (Erickson, 2023). Additionally, the wet electrodes tend to dry over time, which leads to “increased electrode-skin impedance thus lower quality recordings” (Erickson, 2023). An alternative could be to utilize dry electrodes, which have been tested to work in a comparable manner to the wet electrodes and does not require the AgCl gel. The major drawback in using the dry electrodes is that they are much larger than the Ambu wet electrodes, with a size of 55 x 40 mm. We will need to investigate into other options for dry electrodes and aim to find ones of a size comparable to the Ambu wet electrodes so we can ensure to cover most of the rectosigmoidal region.