Fischell Department of Bioengineering Research Works

Permanent URI for this collectionhttp://hdl.handle.net/1903/6627

Browse

Start date: 2020Subject: search.filters.subject.breast cancer

Search Results

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Dual-Intended Deep Learning Model for Breast Cancer Diagnosis in Ultrasound Imaging
    (MDPI, 2022-05-27) Vigil, Nicolle; Barry, Madeline; Amini, Arya; Akhloufi, Moulay; Maldague, Xavier P. V.; Ma, Lan; Ren, Lei; Yousefi, Bardia
    Automated medical data analysis demonstrated a significant role in modern medicine, and cancer diagnosis/prognosis to achieve highly reliable and generalizable systems. In this study, an automated breast cancer screening method in ultrasound imaging is proposed. A convolutional deep autoencoder model is presented for simultaneous segmentation and radiomic extraction. The model segments the breast lesions while concurrently extracting radiomic features. With our deep model, we perform breast lesion segmentation, which is linked to low-dimensional deep-radiomic extraction (four features). Similarly, we used high dimensional conventional imaging throughputs and applied spectral embedding techniques to reduce its size from 354 to 12 radiomics. A total of 780 ultrasound images—437 benign, 210, malignant, and 133 normal—were used to train and validate the models in this study. To diagnose malignant lesions, we have performed training, hyperparameter tuning, cross-validation, and testing with a random forest model. This resulted in a binary classification accuracy of 78.5% (65.1–84.1%) for the maximal (full multivariate) cross-validated model for a combination of radiomic groups.
  • Thumbnail Image
    Item
    Development of an Endoscopic Auto-Fluorescent Sensing Device to Aid in the Detection of Breast Cancer and Inform Photodynamic Therapy
    (MDPI, 2022-11-11) Gaitan, Brandon; Inglut, Collin; Kanniyappan, Udayakumar; Xu, He N.; Conant, Emily F.; Frankle, Lucas; Li, Lin Z.; Chen, Yu; Huang, Huang-Chiao
    Breast cancer is the most diagnosed cancer type in women, with it being the second most deadly cancer in terms of total yearly mortality. Due to the prevalence of this disease, better methods are needed for both detection and treatment. Reduced nicotinamide adenine dinucleotide (NADH) and flavin adenine dinucleotide (FAD) are autofluorescent biomarkers that lend insight into cell and tissue metabolism. As such, we developed an endoscopic device to measure these metabolites in tissue to differentiate between malignant tumors and normal tissue. We performed initial validations in liquid phantoms as well as compared to a previously validated redox imaging system. We also imaged ex vivo tissue samples after modulation with carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone (FCCP) and a combination of rotenone and antimycin A. We then imaged the rim and the core of MDA-MB-231 breast cancer tumors, with our results showing that the core of a cancerous lesion has a significantly higher optical redox ratio ([FAD]/([FAD] + [NADH])) than the rim, which agrees with previously published results. The mouse muscle tissues exhibited a significantly lower FAD, higher NADH, and lower redox ratio compared to the tumor core or rim. We also used the endoscope to measure NADH and FAD after photodynamic therapy treatment, a light-activated treatment methodology. Our results found that the NADH signal increases in the malignancy rim and core, while the core of cancers demonstrated a significant increase in the FAD signal.