Deep orthogonal multi-wavelength fusion for tomogram-free diagnosis in diffuse optical imaging.

Novel portable diffuse optical tomography (DOT) devices for breast cancer lesions hold great promise for non-invasive, non-ionizing breast cancer screening. Critical to this capability is not just the identification of lesions but rather the complex problem of discriminating between malignant and benign lesions. To accurately reconstruct the highly heterogeneous tissue of a cancer lesion in healthy breast tissue using DOT, multiple wavelengths can be leveraged to maximize signal penetration while minimizing sensitivity to noise. However, these wavelength responses can overlap, capture common information, and correlate, potentially confounding reconstruction and downstream end tasks. We show that an orthogonal fusion loss regularizes multi-wavelength DOT leading to improved reconstruction and accuracy of end-to-end discrimination of malignant versus benign lesions. We further show that our raw-to-task model significantly reduces computational complexity without sacrificing accuracy, making it ideal for real-time throughput, desired in medical settings where handheld devices have severely restricted power budgets. Furthermore, our results indicate that image reconstruction is not necessary for unbiased classification of lesions with a balanced accuracy of 77% and 66% on the synthetic dataset and clinical dataset, respectively, using the raw-to-task model. Code is available at https://github.com/sfu-mial/FuseNet.

Multitask Deep Learning Reconstruction and Localization of Lesions in Limited Angle Diffuse Optical Tomography.

Diffuse optical tomography (DOT) leverages near-infrared light propagation through tissue to assess its optical properties and identify abnormalities. DOT image reconstruction is an ill-posed problem due to the highly scattered photons in the medium and the smaller number of measurements compared to the number of unknowns. Limited-angle DOT reduces probe complexity at the cost of increased reconstruction complexity. Reconstructions are thus commonly marred by artifacts and, as a result, it is difficult to obtain an accurate reconstruction of target objects, e.g., malignant lesions. Reconstruction does not always ensure good localization of small lesions. Furthermore, conventional optimization-based reconstruction methods are computationally expensive, rendering them too slow for real-time imaging applications. Our goal is to develop a fast and accurate image reconstruction method using deep learning, where multitask learning ensures accurate lesion localization in addition to improved reconstruction. We apply spatial-wise attention and a distance transform based loss function in a novel multitask learning formulation to improve localization and reconstruction compared to single-task optimized methods. Given the scarcity of real-world sensor-image pairs required for training supervised deep learning models, we leverage physics-based simulation to generate synthetic datasets and use a transfer learning module to align the sensor domain distribution between in silico and real-world data, while taking advantage of cross-domain learning. Applying our method, we find that we can reconstruct and localize lesions faithfully while allowing real-time reconstruction. We also demonstrate that the present algorithm can reconstruct multiple cancer lesions. The results demonstrate that multitask learning provides sharper and more accurate reconstruction.

Translation of a portable diffuse optical breast scanner probe for clinical application: a preliminary study.

Most breast cancer lesions absorb higher levels of near-infrared (NIR) radiation compared to healthy breast tissue due to its increased vascularity. Oxy-hemoglobin (HbO2) and deoxy-hemoglobin (Hb) primarily found in cancerous vascular lesions, absorbs higher levels of radiation in the 650 nm to 850 nm wavelength range than the surrounding fatty tissue and water in the human breast. NIR diffuse optical spectroscopy (DOS) provides real-time functional and compositional information based on the optical properties of biological tissues, which cannot be accomplished by other portable breast imaging modalities. Here we present the first set of clinical trials using a non-invasive, hand-held diffuse optical breast scanner (DOB-Scan probe3) to capture in vivo cross-sectional images of the breast. The scanner uses four NIR illuminating sources with different wavelengths, 690 nm, 750 nm, 800 nm, and 850 nm, to determine the concentrations of the four main constituents of breast tissue, oxy-hemoglobin (HbO2), deoxy-hemoglobin (Hb), water (H2O), and fat. In this paper, we briefly explain the hardware design and image reconstruction algorithm of the DOB-Scan probe, the data collection process, and the imaging results of four different participants, selected from twenty, all who are diagnosed with breast cancer. For each patient, images were scanned from two locations, the first over the cancerous lesion and the second over the same region on the contralateral healthy breast, as a means of establishing controls for comparison. During each scan, four cross-sectional images of the breast, corresponding to four different NIR wavelengths, are reconstructed and displayed on a user interface for reference. Clinical results confirm that the absorption coefficients of cancerous lesions are significantly higher than the normal surrounding tissue. We propose to deploy the probe to effectively identify cancerous breast tissue at an early stage in a primary care setting, which could increase the efficiency of screening programs.