Session B Apps & Diagnostics

Track 2

Session B

Apps & Diagnostics

60 BME SENIOR DESIGN PROJECTS

Development of Affordable Kidney Disease Diagnostic Device for Use in Low Resource Areas

Team 4: Lujain Khusheim, Ashley McFarlane, Ethan Strauther, Zakiah Tcheifa Technical Advisors: Catherine Klapperich, Sushrut Waikar (BU School of Medicine)

There is a huge need for more accessible options for the diagnosis of kidney disease in low-income areas. The device will be used to perform mass screening of individuals and provide a metric for additional testing and diagnosis from a licensed medical professional. It will return a semi-quantitative value based on low, slightly elevated, or high creatinine values using the results from a Jaffe reaction - a more affordable alternative to enzymatic reactions. The main component of our device is a small light proof cuvette holder and a cuvette with a lid to contain the blood sample (obtained from a finger-prick). The lid will hold a plasma separation membrane and a capillary tube to drain the serum. The blood will then mix with the reagent found in the cuvette and the reaction will occur. The reaction will run for five minutes and then the voltage drop will be recorded across the phototransistor at an absorption value of 500 nm. The patient will be categorized into a risk group based on the corresponding voltage reading, where higher readings indicate higher creatinine levels. The device will display this reading on a built-in LED screen. To achieve this functionality, 3-D printed components, simple spectrophotometer-like measurements, and a microprocessor (such as an Arduino) will be used to create the POC diagnostic device at an affordable price range.

Assessing Interstitial Fibrosis and Tubular Atrophy (IFTA) Using Minimally-Invasive Elastic Scattering Spectroscopy as an Indicator of Kidney Health

Team 6: Giulia Boccardo, Ji Young Chung, Meghan Howard. Emily Oros Technical Advisors: Ousama A’amar, Irving Bigio

Chronic Kidney Disease (CKD) is a serious illness characterized by diminished kidney function that affects about 1 in 7 US adults. It consistently results in required dialysis and eventually kidney transplant(s). Currently, assessing the health of kidneys for transplant involves reviewing the deceased’s medical record history and physically examining the organ ex vivo. The current gold standard of CKD diagnosis involves measuring the extent of IFTA through a kidney biopsy, a risky procedure involving a large needle with many drawbacks. Therefore, there is a pressing clinical need for a method to rapidly measure and assess IFTA for both the allocation of transplant kidneys and treatment of CKD without the guesswork or tissue damage. Our team has designed, built, and tested a minimally invasive fiber optic probe to measure the viability of kidney function. Our probe incorporates elastic-scattering spectroscopy (ESS) which measures backscattered light over the 300-900 nm range. We have tested our device on unlabeled tissue of an adenine rat model of CKD and were able to find correlation between disease progression and changes in the measured ESS spectra. The results were observed spectral trends of decreased slope of scattering intensity in the near-UV to short-visible region (300–500nm), relative to longer wavelengths, for fibrotic kidneys compared to normal kidneys. The results of this study will demonstrate the potential of ESS technology as a more accurate, rapid, and non-invasive assessment of IFTA for potential donor kidneys and its application for in vivo procedures.

Effect of Digital Intervention on Social and Verbal Skills in Individuals with Autism Spectrum Disorder

Team 9: Matthew Collins, Nathaniel Heitmann-Basoni, James Maher, Technical Advisor: Andrey Vyshedskiy (ImagiRation, LLC; BU MET)

Verbal communication is arguably the most common and effective method of communication between people. Children with cognitive impairments and learning disabilities, however, often face challenges with regards to the expression and understanding of ideas in a verbal form. Difficulty with such communication can have a very harmful effect on a person’s ability to be independent and fully integrated into society. As such, studies have been conducted on a variety of methods to help these children develop crucial conversation skills, such as understanding social cues and continuing a conversation. Direct intervention and one-on-one speech therapy have been widely proven to be the most impactful methods. Unfortunately, such methods are costly and potentially difficult to implement at a large scale, and many speech therapists face difficulty when trying to distinguish between providing feedback and participating in conversation, an issue which is exacerbated by echolalia. This project aims to improve upon this method of therapy by developing a smart device-compatible application which will assume the role of conversation partner, leaving the clinician to provide feedback. The application utilizes a cloud storage system to pull videos of actors which correspond to conversation topics or particular questions, and it provides methods by which clinicians may give feedback to the patient. Ultimately, this project aims to understand the effects, if any, of replacing human conversation partners with an application on the effectiveness of one-on-one speech therapy for children with Autism Spectrum Disorder (ASD).

Digital Biopsy for Glomerular Ultrastructural Measurement in TEM Images

Team 37: Connor Kim, Aksel Laudon, Grace Qian, Zhaoze Wang (ECE), Jackson Ye INTERDISCIPLINARY Technical Advisors: Weining Lu (BU School of Medicine), Vijaya Kolachalama (BU School of Medicine), Joel

M. Henderson (BU School of Medicine)

The glomerular filtration barrier is vital to normal kidney function. Transmission electron microscopy (TEM) images visualize its cross-sectional ultrastructure, which consists of the sieve-like glomerular basement membrane (GBM) and podocyte foot processes. Proteinuric kidney diseases (PKD) are associated with filtration barrier morphological changes, which are observed experimentally and clinically in terms of GBM width and foot process width (FPW) measurements. Current TEM image measurement is performed manually as no fully automated solution exists. This limits research into PKD mechanisms and therapeutics with its labor intensiveness and operator bias. We developed an automated tool to measure input TEM images from a dataset of wild type (WT, n=5) and PKD model (ILK cKO, n=5) mice. A U-Net semantic segmentation model was trained on the dataset and corresponding manually annotated GBM masks. First, the segmentation model identifies GBM in input images. Next, these GBM segmentations enter an image processing algorithm that estimates GBM width and FPW. In our validation study on the dataset, resulting segmentation accuracy was strong relative to existing literature. The estimated mean GBM width and FPW measurements closely matched manual measurements for WT but differed significantly for ILK cKO animals. Measured GBM width and FPW were significantly wider for ILK cKO than WT mice, which aligns with known morphology. These results suggest our tool performs comparably to manual measurement on healthy tissue and can distinguish healthy from pathological samples. Our tool provides high-throughput, objective morphological analysis for research, and in the future could potentially facilitate clinical PKD diagnosis.