טכניון מכון טכנולוגי לישראל
הטכניון מכון טכנולוגי לישראל - בית הספר ללימודי מוסמכים  
Ph.D Thesis
Ph.D StudentNisreen Shehada
SubjectGas Sensors Based Silicon Nanowire Field Effect Transistors
for Real-World Disease Diagnosis
DepartmentDepartment of Chemical Engineering
Supervisor Full Professor Haick Hossam
Full Thesis textFull thesis text - English Version


Abstract

Cancer is a devastating disease with several medicinal challenges including delayed diagnosis, low efficacy of the anti-cancer therapy and heterogeneity of the disease. In addition, cancer is often misdiagnosed as it has several common symptoms with non-cancerous diseases. Therefore, there is an urgent unmet need for inexpensive and non-invasive technology that would allow: efficient early detection of cancer and bed-side fast assessment of treatment efficacy in order to change the therapeutic approach accordingly. In the current research, we have developed molecularly modified silicon nanowire field effect transistor (Si NW FET) sensors for the detection of multiple diseases via volatile organic compounds (VOCs). These diseases include Gastric cancer (GC), Lung cancer (LC) and non-cancerous Lung diseases (represented by Asthma and COPD). Various features can be extracted from the characteristic curves of Si NW FET devices, supplying a wide assortment of independent signals to be used as virtual sensors. The study was conducted in two parts; first, detection of Gastric cancer from non-cancerous gastric conditions was achieved with high accuracy (~83%). In this part of the study 25% of the samples were used as a "blind" group for validation. In the second stage, both simulations and real breath of patients with of three types of diseases (LC, GC and Asthma/COPD) were analyzed. Analysis has indicated that the sensors differentiate between simulations of breath samples corresponding to each type of disease: LC, GC, and Asthma/COPD. The sensors have also shown strong ability to differentiate real breath samples collected from patients suffering from the different diseases, as well as healthy control patients. The accuracy of the classification is above 80% in most cases, making our device suitable for preliminary testing in the Cancer diagnostic process. In addition, the sensors were able to separate the early stages of the cancer from the advanced stages, adding the benefit of staging the disease in addition to diagnosing it, in the same test.      _____________________________________________________________________________