Development of a "Cytology-on-Chip" Sensor for Monitoring Potentially Malignant Oral Lesions
| dc.contributor.advisor | McDevitt, John T | en_US |
| dc.creator | Abram, Timothy | en_US |
| dc.date.accessioned | 2017-08-03T15:20:17Z | en_US |
| dc.date.available | 2017-08-03T15:20:17Z | en_US |
| dc.date.created | 2016-05 | en_US |
| dc.date.issued | 2016-04-21 | en_US |
| dc.date.submitted | May 2016 | en_US |
| dc.date.updated | 2017-08-03T15:20:17Z | en_US |
| dc.description.abstract | The poor prognosis associated with oral cancer, which has progressed into a global epidemic, is attributed to late-stage diagnosis, multi-focal involvement, and a lack of prognostic indicators for identifying which lesions will undergo malignant transformation. Scalpel biopsy combined with histopathological evaluation remains the gold standard in oral cancer diagnosis and monitoring, but this invasive procedure suffers from severe limitations including poor inter-pathologist agreement and an inability to predict patient outcomes. Furthermore, surgical excision of oral lesions is not capable of completely eliminating the risk of malignant transformation; hence, patients are subjected to repeat biopsy as the sole means of monitoring disease progression and recurrence. Oral medicine clinicians urgently need new technologies that can afford non-invasive, sensitive, and quantitative risk assessments for monitoring the progression of pre-malignant oral lesions. This dissertation describes the development of a “cytology-on-chip” sensor combining non-invasive sampling, microfluidic sample processing and single-cell interrogation, and adaptive machine learning algorithms toward the ultimate goal of empowering oral medicine and dental practitioners to monitor subtle changes at the molecular and cellular levels of suspicious oral lesions for personalized disease management. The use of an “enhanced gold standard” to increase the confidence in histopathological grading for 775 prospectively-recruited patients with potentially malignant oral lesions is demonstrated, resulting in increased pathologist agreement from 69% to 100%. Single-cell measurements from a panel of molecular biomarker assays for each patient were then used to train and validate robust risk stratification models with an overall accuracy of 70%. Two strategies for improving future access to this approach are discussed, including the design of a scalable, pumpless microfluidic device to enable parallel assay processing and the adaptation of a fully-integrated lab-on-a-chip system for future chair-side testing. Equivalency to previous methods was indicated for both systems, ensuring consistent functionality. Finally, the realization of a continuous risk index is detailed, resulting in an improvement in overall predictive accuracy to 79.2% and with strong potential to monitor disease severity over time. A proof-of-concept of the ability to inform clinical decision making with such a tool was performed on a high-risk patient population that successfully identified high-risk oral lesions. | en_US |
| dc.format.mimetype | application/pdf | en_US |
| dc.identifier.citation | Abram, Timothy. "Development of a "Cytology-on-Chip" Sensor for Monitoring Potentially Malignant Oral Lesions." (2016) Diss., Rice University. <a href="https://hdl.handle.net/1911/96537">https://hdl.handle.net/1911/96537</a>. | en_US |
| dc.identifier.uri | https://hdl.handle.net/1911/96537 | en_US |
| dc.language.iso | eng | en_US |
| dc.rights | Copyright is held by the author, unless otherwise indicated. Permission to reuse, publish, or reproduce the work beyond the bounds of fair use or other exemptions to copyright law must be obtained from the copyright holder. | en_US |
| dc.subject | microfluidics | en_US |
| dc.subject | oral cancer | en_US |
| dc.subject | machine learning | en_US |
| dc.subject | lab-on-a-chip | en_US |
| dc.subject | cytology | en_US |
| dc.subject | image cytometry | en_US |
| dc.title | Development of a "Cytology-on-Chip" Sensor for Monitoring Potentially Malignant Oral Lesions | en_US |
| dc.type | Thesis | en_US |
| dc.type.material | Text | en_US |
| thesis.degree.department | Bioengineering | en_US |
| thesis.degree.discipline | Engineering | en_US |
| thesis.degree.grantor | Rice University | en_US |
| thesis.degree.level | Doctoral | en_US |
| thesis.degree.name | Doctor of Philosophy | en_US |
Files
Original bundle
1 - 1 of 1